Early last year, Senator Mitt Romney proposed a new approach to family policy that exposed some significant rifts among right-leaning policy wonks who care about fighting poverty and supporting family formation. This week, Romney (together with fellow Republicans Richard Burr and Steve Daines) has offered a revised version of the idea that might just have what it takes to patch up those differences and attract a broad coalition of backers.

The original proposal was, roughly speaking, to replace a variety of overlapping family supports with a simple, monthly cash benefit to parents, amounting to $4,200 a year per child beginning in pregnancy until age five, and then $3,000 a year per child from ages six to 17. This benefit was intended to replace the existing Child Tax Credit as well as the Child and Dependent Care Credit, the Temporary Assistance for Needy Families program, and the head-of-household tax filing status. It would also have reformed the Earned Income Tax Credit to eliminate marriage penalties and slow benefit cliffs. And its net cost, after replacing those existing benefits, would have been paid for by eliminating the state and local tax deduction.

It was a clever and creative way to simplify and expand the reach of federal support for parents, which would also have gone a long way toward reducing short-run child poverty. But it would have separated those child benefits from any requirement that parents work — particularly by eliminating the TANF program and the limited refundability of the child tax credit. In essence, it proposed a child allowance on the scale of what working parents now receive but without any connection to whether those parents have any earnings.

The arguments sparked by that approach forced champions of anti-poverty and pro-family policy on the right to more or less rank the importance of their priorities. Was the appeal of providing support and encouragement to more parents worth a reduced work incentive? Was the value of promoting and rewarding work as the path out of poverty worth curtailing the reach of support for family formation? Was greater help for married parents raising a family worth the risk of increasing the proportion of children raised by single parents?

On the one hand, work and marriage as the path out of poverty had been the core principle of the spectacularly successful welfare reform of the 1990s, and Romney’s proposal largely abandoned that approach and even scrapped the TANF program created by that reform. The history of welfare in America offered strong reasons for thinking that doing that would incentivize choices that would ultimately impede the very intergenerational mobility everyone wants to foster.

On the other hand, the face of poverty in America had changed dramatically over the past three decades (in no small part thanks to that successful welfare reform), while family formation had weakened and slowed. The balance of priorities, and therefore of policies and incentives, might well need to look different as a result.

The debate over this question (including an actual friendly debate among scholars at the American Enterprise Institute, where both authors of this piece work) helped a lot of conservatives clarify our priorities, goals, and arguments. The two of us were on opposite sides of the Romney proposal — Winship in opposition and Levin in support. Everyone involved supported both pro-family policy and the sustained reduction of poverty, but the nature of Romney’s approach forced hard choices and brought some differing assumptions to the surface.

At the time, the Democrats who controlled both houses of Congress and the presidency were uninterested in cross-partisan policy-making and fully immersed instead in their ill-fated “Build Back Better” fiasco. So, conservatives could let the intra-Right debates about Romney’s idea develop and take shape. Properly conducted, such disagreements among friends and allies can sometimes point toward mutually agreeable ways forward — means of balancing concerns and priorities that result in improvements of the ideas under debate. And that’s just what seems to have happened here.

The new Romney-Burr-Daines proposal tries to bridge the difference between critics and supporters of Romney’s original approach in a very straightforward way: The key difference between this new version and last year’s is the introduction of a modest work requirement. To qualify for the full child benefit, the proposal requires families to have met a $10,000 earnings test in the prior year — a target that not only incentivizes work but also incentivizes marriage (since it is easier for married-couple families to meet that test). Families earning less than that amount would receive a benefit reduced in proportion to their earnings.

Up to that $10,000 earnings level, the work incentive in this new proposal is actually stronger than current policy for most families with children. But the requirement is not so demanding that it will deny substantial support to most single parents. Roughly two-thirds of single parents would qualify for the full benefit if no one changes how much they now work, and of course, if the policy encourages some parents to work more, then that number will be higher.

Like any compromise, this isn’t anyone’s ideal approach. The earnings requirement would reduce the reach of the benefit, and therefore reduce its potential to support family formation and reduce short-run poverty. Yet it’s a modest requirement, meant to mitigate perverse incentives but not to fully replace the emphasis on work and marriage that defines the TANF program (which would be retained under this new proposal). To us, the balance seems roughly right. This proposal will appeal to reformers who actually want to both support parents and encourage work and marriage — which means it could unite most conservatives, though it is also likely to have less appeal than Romney’s original proposal to some progressives who mostly liked that original version because they saw it as a path toward a universal basic income (UBI).

Through this revised approach, Romney, Burr, and Daines seem plainly to be distancing themselves from any case for a UBI. One of us (Winship) continues to think that, to further demonstrate this, the proposal should also avoid providing the benefit on a monthly basis and instead provide it annually as a refundable tax credit. A monthly benefit is not practically necessary, is not widely desired among potential recipients (as the experience of the EITC program suggests), and would create an administrative infrastructure that could be used in the future to advance a UBI — obliterating the incentives for employment and thus for the only truly durable path out of poverty.

But on balance, the new proposal advances the two core goals shared by all on the right who have been engaged in this debate: support for Americans looking to build families and support for Americans looking to rise out of poverty. It helps to show that those two goals need not be in tension; they can reinforce each other and be pursued together.