This week, Congress faces a deadline to reauthorize federal highway funding and prevent the bankruptcy of the Federal Highway Trust Fund. Congress has been looking under the federal government’s couch for pennies, passing dozens of short-term extensions since 2009 using a combination of budget gimmicks and one-off payments.

In one of the bigger deals, it cut back federal pension contributions that will eventually need to be made anyway. Further proposals include selling oil out of the Strategic Petroleum Reserve, cutting dividends paid on Federal Reserve stock and creating a “repatriation” holiday for corporate profits held offshore to return at a favorable tax rate.

Corporate tax reform is certainly desirable, but one-off gimmicks are distracting us from sound principles and longer-term thinking. The focus should be on putting the costs of infrastructure onto the users and beneficiaries of infrastructure. Lawmakers should consider a deal to raise gas taxes back to their inflation-adjusted 1993 level for Democrats, but block-grant highway funding to states and localities for Republicans.

Eisenhower’s gas tax has always been a way to toll the interstates and federally-aided state and local roads without having to install and operate toll barriers. When Ronald Reagan doubled the gas tax in 1982, he reminded the nation that it was a user fee. Funding infrastructure through user fees is both the fairest and most efficient means available for government, short of privatizing highways.

That said, it’s not clear that the Feds should control as much gas tax revenue as they currently do. States and localities know their infrastructure needs better than Washington does. And, as a forthcoming paper by my Manhattan Institute colleague Aaron Renn will show, not all local roads in need of repair actually get federal aid under the current system.

That’s why the feds ought to block-grant the majority of the money back to states and localities, reserving the federal role only for maintenance of important cross-state projects that no one state would have an incentive to pay for.

A “Race-to-the-Top” type of prize system for rewarding particularly innovative and efficient state infrastructure projects, conceptually similar to the Department of Education’s program to encourage state innovation in K-12 education, might also be appropriate. Such a system would incentivize states to compete for a pot of federal prize money for the best projects. Even if a state doesn’t win a national prize, the competitive effort will still have yielded engineering plans for a project that would likely merit a state’s own funds.

Maintaining a federal gas tax — which is really a “user fee on fuel” — while block granting the money back to states and localities has the advantage of reducing wasteful cross-border differences in gas taxes that encourage “tax-shopping” across state lines. For instance, if one state funds its highways through sales taxes, while its neighbor funds with user fees on fuel, it creates an incentive to buy gasoline on one side of the border and do retail shopping on the other side. But this is just wasting shoppers’ and drivers’ time and fuel, compared to flat user fees on fuel on both sides of the border.

Microeconomics is, in large part, a science that explains how tying the beneficiary of an action to the cost of an action leads to more efficient outcomes. Politics is the art of getting somebody else to pay for your benefits. It’s therefore quite natural that Congress would struggle with developing an economically efficient mechanism for paying for infrastructure.

But no matter what lawmakers choose to do next, we should generally judge them by how closely they hew to the “beneficiary pays” principle of infrastructure funding.

This article originally appeared in Washington Examiner