Growing up during the Depression, Sam Marchesi had to drop out of school after eighth grade—soon after his father died—to work and help support his mother and his seven younger siblings. When World War II began, he enlisted in the Army and served in the Pacific. Upon his return, he took advantage of the educational and training benefits of the G.I. Bill, joining the 52 percent of fifteen million returning veterans who did so. He acquired vocational training in architectural drawing and on-the-job training as an apprentice carpenter, skills that enabled him to become a successful custom builder. When I interviewed Marchesi in the late 1990s for a study of the G.I. Bill, he reflected, “I think it was a great thing that the government did, to give us this opportunity to pick up where we left off. We had to face the world. We had to make a living. Thank God the government had the doors open for us.”

The G.I. Bill’s transformative effects on the lives of men like Marchesi have become legendary, but just as striking in hindsight is the clearly visible role that government played as the source of those opportunities. In more recent decades the federal government has expanded its efforts to provide college aid to all Americans. But instead of delivering a straight benefit, like the original G.I. Bill, most of that aid has come through roundabout means, like payments to banks to provide students with loans, or loopholes in the tax code to subsidize families to save for or pay for college. Generations of Americans have now graduated with the help of these costly-though-indirect programs. Yet over the years, in conversations with my own students, I’ve noticed that, unlike Marchesi, few of them recognize that they’ve received benefits from government. It’s hard to imagine them reflecting on their HOPE Tax Credits, or their 529 and Coverdell college savings plans and saying, “Thank God the government had the door open for us.”

Photo: Mark Thomas

And it’s not just my students. In 2008, I conducted a survey to gauge the degree to which Americans who had received various government social benefits recognized them as such. Not surprisingly, most beneficiaries of the G.I. Bill who took part in the survey acknowledged that they had been given a leg up by the government. But of the respondents who made use of tax-advantaged Coverdell or 529 education savings accounts, 64 percent said they had “not used a government social program,” as did 59.6 percent of those who used HOPE and Lifelong Learning Tax Credits.

This disparity has far less to do with some inherent difference in character between the Greatest Generation and their grandchildren than it does with a fundamental change that has taken place in the relationship between citizens and the welfare state. Over the past few decades, while many standard social benefits have atrophied in real value, those packaged as “tax expenditures”—the formal name in federal budgeting parlance for subsidies provided through the tax code—have flourished, growing rapidly in value and number. These tax expenditures for individuals and families represented 7.4 percent of GDP in 2008, up from 4.2 percent in 1976. (Tax expenditures for business, such as those for the oil and gas industry, made up another 1 percent.) By way of comparison, Social Security amounted to 4.3 percent of GDP in 2008; Medicare and Medicaid, 4.1 percent.

These social tax expenditures comprise a major part of what I call the “submerged state.” By that I mean that they are public policies designed in a manner that channels resources to citizens indirectly, through subsidies for private activities, rather than directly through payments or services from government. As a result, they are largely hidden from the public: through them, government benefits people, providing them with opportunities and relieving their financial burdens, often without them even knowing it. Appearing to emanate from the private sector, such policies obscure the role of the government and exaggerate that of the market.

What’s more, the vast majority of Americans garner only modest assistance, if any, from the submerged state. In the case of social tax expenditures, that’s because the most expensive of these subsidies shower their largest benefits on the most affluent Americans.

The great drama now unfolding in Washington over how to deal with the government’s deficits and growing debt tends to be framed in conventional ways. Conservatives aim to use this moment to reduce the size of “big government” while liberals find themselves on the defensive, hoping to limit the damage and furious at the president and Democratic congressional leaders for not fighting harder. But these negotiations can actually be an important opportunity to advance progressive goals, if—as the Bowles-Simpson Commission and others have recommended— we scale back tax expenditures. Doing so could improve the nation’s balance sheet and restore some fairness to the tax code. Even more, it could address the real if inchoate sense many Americans have that government has been “growing,” as measured by deficits and new programs, but in ways that don’t benefit them. Saying good-bye to the submerged state could reconnect citizens with government and reinvigorate our democracy.

The clarion call of the conservative approach to governance that has dominated American politics for much of the past thirty years has been the demand to rein in the welfare state. Although few provisions have suffered outright termination, average benefit rates for several traditional and longstanding policies—such as welfare, unemployment insurance, Pell grants, and food stamps—have deteriorated in real terms, and in some cases the scope of coverage has atrophied. As deficit hawks continually remind us, costs have grown for the “entitlement” programs—Social Security, Medicare, and Medicaid— owing to inflation-protected benefits, soaring health care costs, and the sheer numbers of Americans aging into eligibility. Generally ignored, however, have been the rapidly escalating costs of tax expenditures for social welfare purposes—the sine qua non of our submerged state.

Known in informal parlance as “tax breaks” or “tax loopholes,”
these policies permit households to pay less in taxes if they are involved in some kind of activity or belong to a class that policymakers deem worthy of public support. From the time Ronald Reagan took office in 1981 until 2010, the number of such tax subsidies had increased by 86 percent, from 81 to 151. As of 2011, the federal government annually doles out more than $1 trillion in these tax expenditures.

Understandably, to many people tax breaks may seem substantively different from traditional social benefits. The latter are funded by tax revenues collected from the public and delivered through checks or services to particular citizens, whereas tax breaks function by allowing recipients themselves simply to keep more money, reducing the amount that they would otherwise owe. Traditional social programs also require the development of a bureaucracy to determine eligibility and deliver benefits, whereas the tax expenditures do not. For these reasons, many libertarians and conservatives object to the term “tax expenditures.” While conceding that tax loopholes constitute government intervention in the market, such thinkers equate closing them with raising taxes, unless the changes are offset by lower rates.

As a matter of budgeting, however, there is no difference between a tax break and a social program: both have to be paid for, either by raising tax rates or by adding to the deficit. Eugene Steuerle, a tax economist and political appointee in the Reagan administration, said of the distinction between tax expenditures and direct social spending, “One looks like smaller government; one looks like bigger government. In fact, they both do exactly the same thing.” Certainly their status has not eluded the policymakers who crafted them; the Louisiana senator Russell Long, chair of the Senate Finance Committee from 1966 to 1981 and the father of the Earned Income Tax Credit, said of the term “tax expenditures,” “That label don’t bother me.… I’ve never been confused about it. I’ve always known that what we’re doing was giving government money away.”

The largest tax expenditures have been around at least half a century, each the product of inadvertent origins. As political scientist Christopher Howard of the College of William & Mary has observed, policymakers involved in the haphazard array of decisions that generated these policies could not have imagined that they were establishing benefits that by the late twentieth century would gain quasi-entitlement status and cost the nation burgeoning amounts. Of the three most expensive ones, the Home Mortgage Interest Deduction was created first, as part of the original tax code in 1913; the preferential tax treatment of employer pensions was established through a hodgepodge of administrative rulings and congressional statues between 1914 and 1926; and the tax-privileged status of employer-provided health benefits resulted from a similar conglomeration of policies during World War II and in the 1950s. In 2011, these three pillars of the submerged state are expected to cost the nation $104.5 billion, $67.1 billion, and $177 billion, respectively. The cost of each component has ballooned, owing not only to market forces but also to the incentives the policies themselves offer that promote consumption in particular forms—such as the purchase of bigger homes, pricier college educations, or more expensive health care—thus inflating their value.

Remarkably, despite the vast drain such provisions impose on federal resources, policymakers have mostly allowed them to grow unchecked. Unlike direct social spending, they are not subject to the annual appropriations process in Congress, and thus they have been able to snowball while sheltered from the public glare.

Whereas mainstream Democrats have traditionally taken the lead in creating our landmark direct social programs, it was originally Republicans and conservative Democrats who initiated the benefits that operated through the tax code. Doing so enabled them to court their favored constituencies and channel resources toward them, but without creating or enlarging government bureaucracies to distribute the funds.

Over the past three decades, however, tax expenditures have evolved into the template of choice for anyone designing new social benefits. Conservatives have actively promoted and protected them, and moderate and liberal Democrats have realized that it is far easier to build a coalition of support for social provision through the tax system rather than through direct spending. Thus they have become willing accomplices. President Clinton promoted and signed into law higher education tax credits, a policy favored by Republicans as an alternative to direct spending ever since the creation of the Higher Education Act of 1965; he also expanded dramatically the Earned Income Tax Credit. President Obama has treated tax expenditures as a policy tool for achieving a broad array of objectives; tax benefits accounted for 37 percent of the $787 billion stimulus he signed into law in 2009, including everything from expansions of existing programs to credits for first-time home buyers and those who purchased energy-efficient doors, windows, and appliances.

But the broader goals of progressive politics are undermined by tax expenditures. Reducing them is a goal we should embrace. The problems start with their redistributive impact.

Most Americans assume that U.S. government social programs aid primarily the poor and the middle class, but tax expenditures generally shower their most generous benefits on those in the upper reaches of the income spectrum. To be sure, there are exceptions—most notably the EITC, which genuinely aids the working poor, and Clinton’s HOPE credit, which targeted the middle class. But in the main, such policies are upwardly redistributive, despite rhetoric to the contrary.

On the rare occasions when policymakers do actually speak about the Home Mortgage Interest Deduction, they portray it as a middle-class benefit that helps to increase home ownership, a pillar of the American dream. Yet countries such as Canada and Australia manage to have U.S.-level rates of home ownership without offering a home mortgage interest deduction in their tax codes. Moreover, in 2004, 69 percent of the benefits of America’s home mortgage interest deduction were claimed by households with incomes of $100,000 or above—the top 15 percent of the income distribution. That same group also reaped 55 percent of the benefits emanating from the tax-free status of retirement benefits and 30 percent of those from employer-provided health benefits. This is because most tax expenditures reward activities that people with greater resources are better poised to take part in: buying more expensive homes and qualifying for mortgages far bigger than those of the typical home buyer; or obtaining generous employer-provided benefits, whose previously broad coverage has declined sharply, particularly among those with low to moderate incomes. Tax expenditures also exacerbate economic inequality by dramatically reducing the revenues government collects, leaving considerably fewer resources available for the programs like Head Start and Pell grants that benefit lower-income Americans.

Even more harmful to the United States than the economic
effects of these submerged state policies are the effects of the politics they generate. The vested interests that profit from these policies—ranging from the real estate and health care industries to the nonprofit foundation world—are keenly aware of them and invest heavily in their political capacity to preserve and defend core policies. For example, the amount the real estate sector contributed to campaigns more than tripled between 1992 and 2008, growing from $43 million to $138 million in 2010 dollars. Its spending on lobbying escalated far more quickly, increasing by 73 percent between 1998 and 2009.

Ordinary Americans, by contrast, have little awareness of the very existence of such policies, even if they are beneficiaries themselves. In the 2008 survey I mentioned above, respondents were asked whether they had “ever used a government social program, or not” and later queried on whether they had ever utilized any of nineteen different social benefits. Those who had benefited from tax expenditures were most likely to deny having used a government social program—for example, 60 percent of those who had used the Home Mortgage Interest Deduction gave that answer.

To make sure respondents weren’t just reacting negatively to the term “government social program”—which for some Americans may connote welfare for low-income people—the survey also asked participants whether they agreed with the statement “Government has given me opportunities to improve my standard of living.” Among people with the same income, level of education, age, race or ethnicity, and sex, the greater the number of direct, visible policies they had ever used—from a list including Social Security, food stamps, Pell grants, unemployment insurance, and several others—the more likely they were to agree with this statement. Yet, controlling for the same factors, the more tax expenditures the individual had used, the less likely he or she was to agree. Those who used the visible programs could see government improving their life chances, but those who used the hidden ones failed to observe such effects.

This finding is ironic, because policymakers of both parties routinely justify tax expenditures by claiming that they either provide people with opportunities—for example, to pursue education or home ownership—or they improve their standard of living. Such policy effects appear to be lost on the beneficiaries themselves—who, in fact, seem fairly convinced that government did not assist them. The submerged design of tax benefits appears to mistakenly convey to people that they have gained whatever measure of economic security or opportunity they have strictly through their own merits, unaided by government help.

We might expect that even if recipients of tax expenditures do not recognize them as public social benefits, then at least the lower tax bills they enjoy as a result would generate more positive views about the tax system. I examined this possibility in the same survey by asking people whether they felt that the amount they were asked to pay in federal income taxes was “more than [their] fair share,” their “fair share,” or “less than [their] fair share,” once again controlling for the factors noted above. All else equal, the greater the number of visible policies individuals had used, the more likely they were to feel that they paid “less than [their] fair share.” By contrast, however, tax expenditures seemed to have no discernible impact on their beneficiaries’ attitudes about taxes, regardless of the enormous drain these policies impose on federal budgets. Evidently, beneficiaries of visible policies understood that taxes help to pay for such programs, whereas those who prospered from the policies of the submerged state failed to grasp that connection.

Not surprisingly, given the invisibility of tax expenditures to most Americans, they generate a passive public. If people are barely aware that such policies emanate from government, then they are obviously unlikely to have opinions about them that reflect their interests and values, or to engage in political participation related to them. The same 2008 survey also asked beneficiaries of specific policies who had indicated that they participated in political activities whether they had ever done so with that policy in mind. For example, beneficiaries who had voted were asked if they had ever “taken into account the position of a candidate on the [program used] in deciding either how or whether to vote,” and, if they had ever made campaign contributions, whether they did so “at least in part, because of [their] concern about [the program used].”

Beneficiaries of visible programs like Social Security and Medicare reported high rates of action to influence the policies they rely on—far more than beneficiaries of tax expenditures. Certainly part of the problem is that, whereas seniors are mobilized by the AARP and the political parties, no broad-based citizens’ groups advocate on behalf of the public’s interest in tax expenditures, leaving the vested interests’ power unchecked. But even beneficiaries of the food stamps program, who lack a group to mobilize them, targeted their political activity at higher rates than the tax break beneficiaries: among those who had voted, 21 percent reported taking the policy into account when doing so, compared to only 14 percent of Home Mortgage Interest Deduction beneficiaries; the rates for campaign contributing among those same groups were 17.7 percent and 8.1 percent, respectively. In short, while the policies of the submerged state engender activism among the powerful groups that benefit most from their existence, they inculcate only passivity among ordinary citizens. This means not only that their interests are routinely circumvented through these upwardly distributive policies, but also that democracy itself is undermined by their existence.

Given how unaware the average citizen is of the submerged state, it’s no surprise that decisions to cut or expand it typically happen with little public airing, usually as part of some larger debate over budgets and taxes. In his first budget to Congress, President Obama proposed to rein in the tax breaks given to the most affluent Americans via deductions for charitable contributions and home mortgage interest by capping their value at a rate lower than the marginal tax rate assigned to those in the upper income brackets. Instead of being able to deduct 39 or 36 percent of the value of their mortgages, in other words, single Americans making more than $174,400 per year would only be able to deduct 28 percent—the same as those who earn $83,600 a year. (This change would actually have reinstated restrictions on tax breaks that existed during the 1990s, signed into law by President George H. W. Bush and ended by his son George W. Bush in 2001. During the decade in which these restrictions were in place, housing prices and charitable contributions soared.) Obama’s proposal was projected to save the federal government $267 billion over ten years—45 percent of the funds needed to finance health care reform.

Instantly, the proposal set in motion the typical politics of the submerged state, as vested interests rallied to defend their pet policies while the public remained out of the loop. Each of the organizational “likely suspects” in the real estate industry stormed Capitol Hill, circulated letters to all elected officials, and placed ads in prominent newspapers in order to express their unequivocal opposition to the changes. Charles McMillan, president of the National Association of Realtors, said, “Diminishing or eliminating [the Home Mortgage Interest Deduction] would hurt all families, the housing market, and our national economy”— language that cloaked the fact that the proposed changes would affect only the wealthy, and would simply reinstate prior policies. Perhaps more surprisingly, the philanthropic, foundation, and nonprofit sector mobilized just as quickly—and in some ways more effectively than the real estate sector. They committed two-thirds as much to lobbying—$44 million—but as the presumed beacon of altruism, they were able to stir even more opposition to Obama’s plan. Claiming the moral high ground, the Council on Foundations and others predicted sharp declines in charitable giving if the tax benefit was reduced. Even Democratic leaders in Congress quickly distanced themselves from the president’s proposal.

At one press conference, reporters pressed Obama on whether he regretted putting forth the idea, to which he answered forthrightly,

People are still going to be able to make charitable contributions. It just means if you give $100 and you’re in this tax bracket … instead of being able to write off 36 or 39 percent, you’re writing off 28 percent. Now, if it’s really a charitable contribution, I’m assuming that shouldn’t be the determining factor as to whether you’re giving that hundred dollars to the homeless shelter.… What it would do is it would equalize. When I give $100, I get the same amount of deduction as … a bus driver who’s making $50,000 or $40,000 a year [who] give[s] that same hundred dollars…. [H]e gets to write off 28 percent, I get to write off 39 percent. I don’t think that’s fair.

This explanation—a rare instance in which a public official clearly articulated how tax breaks function and who benefits— failed to be heard amid the fray, as executive directors of numerous philanthropic organizations wrote to Congress insisting that the proposal would “impose a tax on charitable giving.” The proposal never had a chance. Rent-seeking crony capitalism reigned, with interest groups protecting the subsidies that benefited themselves and the affluent, and ordinary Americans had no inkling of what had transpired.

Would ordinary citizens be more likely to support efforts like Obama’s if they were better informed about them? A survey-based experiment I conducted with Matt Guardino of Syracuse University in 2008 suggests the answer is yes. After respondents were provided with basic information about the upwardly redistributive effect of the Home Mortgage Interest Deduction— the fact that it benefits mostly affluent people—opposition grew sharply, particularly among those with low to moderate incomes and among liberals and Democrats. By contrast, after being informed that the EITC benefited mostly those in low-to-moderate-income groups, support grew among respondents generally, regardless of income.

These findings are consistent with what we know about contemporary American attitudes toward taxes in general. Countless polls in recent years have shown that strong majorities of voters favor higher taxes on the wealthy as the way to close budget deficits and shore up entitlements like Social Security. Similar majorities would likely favor cutting tax expenditures for the well-off if the issue were explained to them.

At present, however, few opinion leaders, even on the left, have taken up the cause of tax expenditure reform, except on a few issues, such as ending tax subsidies for oil companies. Indeed, some of the most sophisticated and influential seem confused about the subject. Comedian-commentator Jon Stewart, for instance, has mercilessly skewered Republicans for defending the Bush tax cuts for the wealthy. But when Obama gave a speech in April calling for “spending reductions in the tax code,” Stewart took the position—common among conservatives— that the president was engaging in doublespeak:

What? The tax code isn’t where we spend. It’s where we collect.… You managed to talk about a tax hike as a spending reduction. Can we afford that and the royalty checks you’ll have to send to George Orwell?

American politics today is ensnared in the paradox of the submerged state. Our government is integrally intertwined with everyday life from health care to housing, but in a form that eludes our vision because it makes governance invisible. As a result, many Americans express disdain for government social spending, incognizant even when they themselves benefit from it. They are disturbed by growing deficits, but do not realize that policies whose benefits flow disproportionately to the affluent consume a large portion of growing federal entitlements. They are easily seduced by calls for small government, not realizing that champions of that philosophy brought us the most costly policies of the submerged state.

Political and opinion leaders permit this deception to persist by failing to talk openly and honestly about tax expenditures. A minor exception to the rule, Obama has shown that he is quite capable of speaking clearly and candidly about how such policies work and who benefits from them, but the number of occasions when he has done so have been few and far between. His own party in Congress has yet to join him in a full-throated manner; in fact, its leaders themselves torpedoed his efforts in 2009. Without discussion of what is actually at stake in spending through the tax code, citizens are not told what these subsidies really are: special provisions for particular groups of people, especially the wealthy, that are paid for through either higher taxes on other Americans, reductions in spending, or expansions of our deficit.

For too long, progressives have accepted the conservative playbook, creating and expanding tax expenditures on the assumption that they can tilt some of their benefits to low- and middle-income Americans. As long as this cornerstone of the submerged state is left intact, it fosters the delusion that governance is generally ineffective and unhelpful to most Americans, and it prompts people to attribute to markets more credit than they are due.

Fortunately, and rather ironically, the coming showdown over raising the debt ceiling presents a golden opportunity to substantially scale back the submerged state—and to advance progressive goals in the process. Republican lawmakers, obliged by their base to cut federal spending but fearing an electoral backlash if popular social programs like Medicare are decimated, are increasingly open to the suggestion of the Democratic deficit hawks in the administration and Congress that if the budget ax must fall, it ought to fall most heavily on tax expenditures. It is a negotiating strategy that liberals would be wise to encourage. For those who care about reducing inequality, making American governance more transparent, and reinvigorating democratic citizenship, this is a chance that should not be missed.

Suzanne Mettler

Suzanne Mettler is Clinton Rossiter Professor of American Institutions at Cornell University. Her latest book, The Submerged State: How Invisible Government Policies Undermine American Democracy, will be published this fall by the University of Chicago Press.