On Both Politics and Policy, "For All" Beats "For Some"

Medicare for All or Medicare for All Who Want It? Free college for all or free college for just the non-rich? The debate between universal (available to everyone) and means-tested (available only to those who meet certain criteria) programs has defined the Democratic primary. Bernie Sanders, often joined by Elizabeth Warren, argues for universalism, declaring education and health care to be basic human rights. Amy Klobuchar, Pete Buttigieg, and Joe Biden argue against, contending that government resources must be targeted only to those in need, rather than wasted on the rich and/or on those who ostensibly don't want them.

On the most commonly cited rationale for each position--sustainability for universalists and resource constraints for means testers--proponents of universalism have the upper hand. Medicare and Social Security, two of the United States's largest, most successful, and most popular programs, are as close to universal as we've got. By giving everyone a stake in these programs, proponents argue, their near-universality has insulated them from attack. Bob Greenstein (the President of the Center on Budget and Policy Priorities, where I used to work) points out both that these programs have been cut and that their popularity could conceivably be due to the perception that they're tied to work rather than to their quasi-universal nature, but the Alaska Permanent Fund, a state-level universal program not tied to work, also enjoys overwhelming public support. So do universal programs that aren't tied to work in other countries - other countries' universal health care systems, for instance, are way more popular than our means-tested approach. It's reasonable to expect a universal program to be more sustainable than a means-tested alternative over time.

The Buttigieges of the world counter that universal programs are too expensive; in December, for instance, Buttigieg said they would require "the kind of taxation that economists tell us could hurt the economy." But even if you reject the notion that government spending can be substantially increased without raising taxes, concerns about higher taxes are entirely without merit. Research has consistently (and predictably) failed to support such concerns, the United States has significantly lower taxes than the rest of the developed world, and scores of reputable economists support tax proposals, like those Sanders and Warren have released, that can fund the universal programs on offer. When Buttigieg says he'd prefer to "save those dollars [that would otherwise be spent on free college] for something else," he is presenting a false choice. It is only his and others' political preferences, not actual resource constraints, that stand between us and full funding of all the priorities he listed: education, infrastructure, child care, housing, and health care.

Still, the most compelling case for universal programs isn't political. It is, ironically, that they're better at achieving two of means testing's major goals: helping people in need and doing so efficiently. They reduce stigma, arbitrariness, usage barriers, and administrative costs.

Universal programs help people in need by reducing stigma

Most low-income people work incredibly hard to put roofs over their heads and food on their tables. Yet they're constantly accused of being unskilled, lazy, good-for-nothing loafers in search of government handouts. Afraid of being perceived that way and/or ashamed of their economic situation, many people who are struggling to get by decide not to access the means-tested benefits to which they're entitled. They'd rather go hungry than risk someone catching them using food stamps in the checkout line.

Correcting false stereotypes is a top priority, with universal programs a useful complement for improving the experience of people in need. If everyone received SNAP benefits (SNAP, which stands for Supplemental Nutrition Assistance Program, is the contemporary name for the food stamp program), for example, using them would no longer identify someone as low-income. We would thus expect higher rates of SNAP usage among low-income people.

That's exactly what we've seen with the school meals program following the introduction of a program called "community eligibility," which enables schools and school districts with a certain percentage of low-income students to offer free school meals to all students--regardless of their income levels--free of charge. Research suggests that reduced stigma is at least part of the reason students at schools that have adopted this program are more likely to take advantage of school breakfast and lunch programs.

Universal programs help people in need by eliminating arbitrary cutoffs

For SNAP, the income eligibility threshold is 130% of the poverty line, or about $27,700 annually for a family of three. People who make less than that amount (provided they meet other requirements - SNAP also has an asset test and restrictive eligibility rules for various groups of people including immigrants, individuals aged 18 to 49 who don't have children, and students) can access benefits; people who make more than that amount cannot. Under Buttigieg's higher education plan, college is free only for families making less than $100,000 a year (and discounted for families making between $100,000 and $150,000).

Means-tested benefits typically phase out slowly - that is, benefits get gradually smaller as beneficiary income gets higher - to ensure that the sum of pay plus benefits continues to increase when people pass eligibility thresholds. But why shouldn't a family of three making $30,000 a year get food assistance? Why should $100,001 be the level at which a family starts having to pay for college? Eligibility thresholds in means-tested programs are arbitrary and inevitably create strange, difficult-to-justify divides between people right above and right below them. Universal programs avoid this problem completely by providing the same benefit to everyone.

Universal programs help people in need by reducing usage barriers

Means testing requires some form of testing, as the name implies, to determine whether or not someone is eligible for benefits. Depending on the complexity of a program's eligibility rules, that testing might require a form of identification, proof of residence, proof of income, or any number of other things. Eligible beneficiaries may need to mail, hand-deliver, or electronically submit one or more forms, which, as Sanders accurately observed during the December debate, "people are sick and tired of filling out.

Filling out forms and proving eligibility is much more than an annoyance for many eligible people in need. Some may not know how to read or write. Some may move and/or change jobs frequently. Some may lack an official ID. The more hoops people have to jump through to access benefits, the fewer eligible people will actually end up receiving benefits.

Government agencies can mitigate this problem with outreach efforts and assistance programs, of course. But even well-administered means-tested programs like SNAP that continue to improve in these areas don't catch everyone they should, in part because of the access barriers means testing inherently creates--in 2016, the most recent year for which we have data, about 15% of people eligible for SNAP did not participate in the program.

Universal programs improve efficiency by reducing administrative costs

In addition to creating an obstacle for eligible beneficiaries, the complexity introduced by means testing presents a challenge for efficient government. Every form that needs to be filled out has to be processed. Eligibility has to be verified. Complex rules have to be actively managed. Means-tested programs spend a larger share of their money on administrative overhead than universal programs do.

Administrative costs for Social Security, for example, are only 0.7% of total expenses. For SNAP, one of the most efficient and effective means-tested government programs, administrative spending comprises 7.7% of its total budget. Over three-quarters of those administrative costs are "certification-related," meaning they're "associated with determining household eligibility."

To be clear, the overall cost of SNAP and other means-tested programs would be many times higher, even with substantially reduced overhead costs, if they were more universal. Increased overall cost is the only real potential downside of universality. And if one were forced to choose between increasing benefits for people in need and extending benefits to higher-income people who don't currently receive them, increasing benefits for people in need would be the clearly correct choice.

But as noted above, that choice is a false one. There is no question that the US government has the money to offer increased benefits through universal programs. The only question is whether we will choose to spend it on the worthy goals of helping people in need and improving government efficiency for everyone.