In the twenty-first century United States, social mobility is highly dependent on personal mobility. And, in much of the country, personal mobility also means automobility. Low-income Americans, who neither own nor have access to a car, often are, consequently, locked out of various opportunities.

Transit-oriented development, a favorite program of progressives, can help only at the margins. Two pilot programs operated by the Department of Housing and Urban Development—Moving to Opportunity and Welfare to Work vouchers—have demonstrated that low-income workers with access to cars were twice as likely to get jobs and four times as likely to stay employed. And a recent Brookings Institution study showed that only 22 percent of potential metro area jobs for low-income workers were accessible to them by means of mass transit in less than an hour and a half—each way.

Unfortunately, as the Urban Institute observed in its 2014 report “Driving to Opportunity,” “there are few federal programs aimed at helping low-income families gain access to automobiles and some programs actually act as barriers to gaining such access.”

Is the answer, therefore, some kind of means-tested program to subsidize car access or ownership for the poor—and only the poor? Such an approach would be misguided, because, as policymakers have long recognized, “programs for the poor are poor programs.” Middle- and upper-income voters resent means-tested welfare programs. The most successful social programs are universal programs that help all income groups at least somewhat, while disproportionately helping the poorest.

Federal tax subsidies to predominantly middle-class commuters already exist. Since 1993, tax-free, employer-provided commuter benefits, including carpooling and parking benefits, have been added to previously existing, employer-provided fringe benefits. But the goal of U.S. social policy should be to move away from employer-provided benefits like these that encourage “job lock” and toward a fairer and more efficient system of universal, portable, individual-based benefits.

To increase access to cars by low-income commuters, I propose a Universal Mobility Account (UMA). Every American working-age adult would receive a UMA in the form of a flat, refundable federal income tax credit. If the credit exceeded the worker’s tax liability, the IRS would provide a direct subsidy.

UMAs could be used to pay for monthly passes from transit agencies of the kind that are already routinely used for many forms of transit, including rail, bus and subway. Many metro transit authorities have expanded the options to also include both bike-sharing and car-sharing programs.

But since, as the Brookings study showed, most low-income workers can’t get to the jobs via mass transit, UMAs would provide them with access to cars, through transit agency car-sharing programs. This would have a number of benefits. It would allow participants to use cars to commute to work without having to undertake the burden of car payments and insurance. And it would not provoke the kind of political backlash that a means-tested car-buying or car-sharing program might inspire.

Most of the beneficiaries of the Universal Mobility Accounts would be middle-class commuters. The program might be designed so that the transit agency, or a contractor on its behalf, could arrange, along with the IRS, to subsidize low-income commuters who qualify for tax credit refunds. Nobody, except the IRS and the transit authority (including its agents), would know which transit pass-holders were being subsidized because of their income level. Poor workers would be empowered to use the same metro-wide multimodal transit system that was predominantly used by the middle class—including car-sharing—without any shame or stigma attached.

Another advantage of Universal Mobility Accounts is that they would not advantage any one mode of transportation over others. Low-income commuters could use their subsidized accounts for bus, rail or subway transit, whichever was most convenient. At the same time, the ability to use UMAs in public, or use publicly sponsored car-sharing systems, would greatly expand economic opportunities for low-income workers who do not own cars, particularly in neighborhoods and regions where most potential work sites are accessible only by automobile.

Because they are flexible, Universal Mobility Accounts could easily accommodate radical technology driven changes in transportation in the future. For example, many believe that the future of passenger transportation will involve new private and public sector models based on “rides” rather than individual car ownership. UMAS could help ease the transition from today’s mix of personal commuting and mass transit to this new one.

Universal Mobility Accounts that expand social mobility by simultaneously expanding geographic mobility exemplify inclusive, “middle-out” reform. Their goal is to allow low-income Americans to enjoy the opportunities and amenities that define a middle-class life, and not to create a separate, means-tested, rule-bound, punitive and stigmatizing set of supports limited to the poor. By benefiting middle-class and poor American workers alike, Universal Mobility Accounts might successfully obtain the broad political support that could make them a success.