As the Washington DC City Council prepares to vote on a bill that would provide workers in Washington, DC up to 11 weeks of paid family leave upon the birth of a child, a fundamental question remains unanswered: how much should government intervene in how employers compensate workers?

The federal government does so quite a bit at present. By exempting employer-provided health insurance from income taxes, our tax law is responsible for the fact that a majority of Americans get their health insurance from their employer. The exemption is also largely responsible for the fact that so many of these employers have what can only be described as overly generous health insurance plans, which can cover health care expenses both routine and exceptional.

The tax code also nudged American businesses to provide pensions as well, since the money set aside in a defined benefit plan generally isn’t taxed. When pension law created the tax breaks for employer-provided health insurance these spouted up instead.

In the heyday of unionism, labor union leaders pushed for more fringe benefits for their workers, often more fervently than they sought out wage increases. They did so in part because of the tax break—why not get a tax-free benefit for workers rather than have workers pay for the same benefit with after-tax wages, they reasoned—and partly because such benefits could be made more durable than other forms of compensation. For instance, the UAW contracts in the 1970s-1990s typically provided health insurance for laid-off workers for up to a year after they were let go, and sometimes longer.

This wasn’t necessarily a good thing for the U.S. economy. Rigid compensation meant that companies resorted to overtime when demand picked up rather than hire more workers. While it also deterred them from laying off workers when there was a downturn in demand since the attendant cost savings would be slight, the short-term stability was an ephemeral benefit to workers. There were fewer jobs available as a result and it did nothing to encourage employment growth in such industries.

Some of these benefits were jettisoned—or at least scaled back—after the bankruptcy of GM and Chrysler in 2008—fifteen years after at Caterpillar—and today the manufacturing workers in a union are much more likely to have a 401k, health insurance with co-pays, deductibles and a monthly contribution, and modest ancillary benefits.

Unions changed course only in part because of their reduced leverage after the diminution of manufacturing in the U.S. economy. They also perceived that their workers would rather have money than an additional benefit. Also, the realization that many workers’ manufacturing jobs may be less permanent than a generation ago also helped change demand. A long-term benefit means little for someone who worries that their job may not exist after the next recession.

More flexible compensation that is directly tied to a worker benefits the economy in the long run. Firms find it less expensive to contract and expand, which should increase employment in the long run. It should also increase wages, if and when we return to a full-employment economy—the tremendous wage gains in the bottom quintiles of the income distribution in the late 1990s should be the goal of every administration of both parties.

Now the DC government is countering this trend by providing a new benefit, financed not by the companies directly but via a tax of .62 percent on corporate profits. To its credit, there has been some thought put into how to efficiently create this benefit, and the Council concluded that by having the government finance it rather than having the employer pay for it directly removes any disincentive that such a benefit would have towards hiring expectant parents. It also reduced the cost, along with the attendant tax increase, by capping the benefits at roughly $1,000 a week, so it’s a little more egalitarian than it would otherwise be.

But the new benefit is still a mistake. It will end up increasing the cost of doing business in DC and will likely end up pushing a few businesses that are trying to decide between DC or Virginia or Maryland to head to one of the other states. To suggest that it’s too small of a tax to matter may sound intuitively appealing, but the notion that costs do not really matter is becoming a tired trope. It is the reason why the left says the minimum wage will not decrease employment, land-use restrictions don’t increase housing costs, or that unemployment insurance benefits don’t lengthen unemployment spells. One can argue about the degree of the impact, but to pretend firms can “swallow” costs ad infinitum is just facile.

If the DC City Council wants to do more about the plight of the working poor, it should focus more on how to encourage them to acquire a better education. DC’s aggressive embrace of charter schools has paid dividends in that respect, as I have observed firsthand. Additionally, the federal government should take steps to reduce the disincentives to work that exist in the tax code. University of Chicago economist Casey Mulligan has found in his research that most working Americans earning between $30,000 and $50,000 face an implicit marginal tax rate in excess of 50 percent.

A narrowly-focused benefit such as paid family leave will be a costly solution with unintended consequences, the least of which is serving as a harbinger for other taxes for other benefits.