One of the regular patterns in public opinion polling is that approval for proposed programs is higher if a survey question omits any reference to the cost in higher taxes. The same phenomenon has probably also been at work in the Democratic primary race this year. Bernie Sanders has been calling for the federal government to pay for free health care, free college tuition, and other programs, but neither Hillary Clinton nor the media have focused much attention on the tax implications.
If Sanders were to become the Democratic presidential candidate, the silence surrounding the tax issue would surely end. Republicans may be holding fire on Sanders’s proposed taxes in the confident expectation that they could use them against his candidacy in the general election. That’s all the more reason Democrats themselves should want to get the issue out on the table now.
The gulf separating the tax proposals of presidential candidates this year—between Republicans and Democrats, and between the two Democrats—is an example of how radically polarized much of American politics has become. On one side, the Republicans are proposing staggering tax cuts: Ted Cruz would cut federal revenues over the next decade by $8.6 trillion, Donald Trump by $9.5 trillion. On the other side, Sanders’s proposed tax increases are even larger than the Republican tax cuts—$15.3 trillion. The distributive effects go in opposite directions: While the rich get the biggest benefit of the Republican cuts, they bear the biggest burden of Sanders’s increases.
Like Sanders, Hillary Clinton also proposes to increase taxes on the rich, and in any other recent election, her proposals would have been seen as distinctly progressive. She calls for a new surtax of 4 percent on those with incomes over $5 million, a new minimum tax of 30 percent (the “Buffett rule”) on those with pre-deduction incomes of over $1 million, increases in capital gains taxes for assets held less than six years, a new tax incentive for employee profit-sharing, and other changes in tax policy to discourage high-frequency trading and shifts of corporations, jobs, and investment abroad.
But while the direction of tax changes proposed by Clinton and Sanders is similar (and some specific provisions overlap), there is a huge difference in magnitude. Clinton’s proposed tax increases amount to $1.1 trillion over a decade, an amount that seems comparatively modest since Sanders’s tax increases exceed hers by a ratio of 14 to 1.
Sanders and his supporters reject analyses that emphasize the tax side of his proposals alone on the grounds that the funded government programs will relieve Americans of other costs, notably private health insurance premiums and out-of-pocket health expenses. The force of that argument depends, however, on the true cost of Sanders’s single-payer health plan. According to his campaign, the plan would require raising $1.38 trillion annually in additional federal revenue, nearly as much as the federal income and estate taxes raise today. But other analysts calculate that even that amount would be insufficient. Kenneth Thorpe, an economist at Emory University, puts the cost (over current federal spending) at $2.43 trillion annually (“Why Sanders’s Single-Payer Plan Would Cost More Than His Campaign Says”). Once the added taxes needed to pay for the plan are factored in, it would no longer produce the net savings Sanders is promising to the majority of the currently insured.
The most reliable, independent analysis of the candidates’ tax proposals comes from the Tax Policy Center (TPC), a joint effort by economists at the Urban Institute and the Brookings Institution who are widely respected for their professionalism. This is the source of the projected revenue effects for the Trump, Cruz, Clinton, and Sanders proposals I have already cited. (The TPC projections are “static” estimates that assume no behavioral change in response to the taxes but provide a reasonable baseline for comparisons.) Unless otherwise noted, I use the TPC numbers.
Senate Budget ranking member Bernie Sanders holds a news conference to introduce legislation "to make millionaires and billionaires pay their fair share of estate taxes and close loopholes that have allowed billionaires to avoid billions in taxes" on Thursday, June 25, 2015.
For journalistic analysis of the tax issues in the Democratic campaign, the best source has been Vox. The major newspaper, broadcast, and other media have given the issue surprisingly little attention. Part of the explanation for that inattention may be that Clinton has chosen not to attack Sanders on taxes, perhaps out of fear of alienating his supporters, whom she hopes to win over for the general election. Some journalists may also believe that Sanders’s proposals are so unlikely to be adopted that there is no point reporting on them. Others in the media may have avoided the tax issue because of the sheer complexity of what Sanders is proposing. If he was calling for just one tax increase, that might have become a focus of discussion. But he is proposing so many different increases for so many different programs that the subject is dauntingly complex.
The Sanders Tax Proposals
The TPC report on Sanders’s tax proposals distinguishes more than 30 proposed changes, nearly all of them raising taxes. In an earlier article, I grouped many of these changes together and referred to Sanders as having proposed 11 different tax increases. I was counting six of these as coming from his single-payer health plan (as the Sanders campaign itself presents them):
1) a new payroll tax of 6.2 percent;
2) an additional, across-the-board 2.2 percent tax on income;
3) higher estate taxes;
4) taxing capital gains and interest as ordinary income for individuals with incomes over $200,000 (and for married couples with income over $250,000);
5) limiting tax deductions for those upper-income taxpayers; and
6) higher income-tax rates in the form of new surtaxes on the same group.
Then there are five other tax increases linked to different programs:
7) an additional increase in payroll taxes of 0.2 percent for paid family leave;
8) lifting the income cap on Social Security taxes to pay for higher Social Security benefits;
9) a financial transaction tax to pay for free public college tuition and student debt relief;
10) a tax on carbon fuels, which would fund alternative energy projects and rebates to Americans making less than $100,000 a year; and
11) a package of increases in corporate taxes to finance new infrastructure investment.
Although these tax increases are heavily oriented toward the top brackets, they are not insignificant for middle-income households. Middle- and lower-income people would be mainly affected by the 6.4 percent rise in payroll taxes and the across-the-board, 2.2 percent income-tax rise—increases that together account for about two-fifths of the additional revenue Sanders raises. In the short run, some employers may bear the cost of additional payroll taxes, but economists generally agree the long-term impact falls on employees even if the employer nominally pays the tax. On that assumption, households with incomes in the middle fifth of the income distribution would see their taxes rise by nearly $4,700, or 8.5 percent of their average after-tax income.
As a result of the new surtaxes on income and other measures, upper-income households would pay a lot more. The top fifth would see their taxes increase by an average of about $45,000 (17 percent of their after-tax income), while those in the highest bracket, the top .1 percent, would pay $3.1 million more in taxes (45 percent of their average after-tax income).
Taking both payroll and income taxes into account, marginal tax rates—the tax on the next dollar earned—would go up for middle- and upper-income households, but a lot more for those with high incomes. For example, individuals making between $78,300 and $118,500 would see their marginal rate rise from 37.4 percent to 43 percent. Marginal rates for those with incomes between $500,000 and $2 million would go from 42.8 percent to 59.4 percent, and the rates in Sanders’s top bracket (more than $10 million) would hit 67.3 percent. These marginal rates on top incomes would be as high as Sweden’s (67 percent), somewhat higher than Denmark’s (55.6 percent), and much higher than the top marginal rates in Germany and the United Kingdom (both 45 percent).
While raising tax rates on upper-income taxpayers, Sanders also holds down the value of deductions for them. Here’s how he does it: For individuals with up to $200,000 income, Sanders retains the current income tax brackets up to the 28 percent level. The surtaxes on higher incomes—ranging from an additional 9 percent on income between $200,000 and $500,000 to an additional 24 percent on income over $10 million—fall on income before deductions (“adjusted gross income”). As a result, individuals with incomes above $200,000 get only 28 percent of the value of a deduction, even though they are in higher tax brackets. Clinton also caps deductions for the rich, except that her cap does not apply to charitable donations—a difference of some importance to the nonprofit sector.
Putting together his different proposals, Sanders would raise the top capital gains tax rate to 64.2 percent. That would be the highest of any developed country by a wide margin—more even than the two Scandinavian countries Sanders often mentions as models. Denmark, which has the highest rates in Europe, has a 42 percent rate; Sweden, 30 percent. In the United States, the federal rate on capital gains has risen under Barack Obama from 15 percent to 23.8 percent. Sanders’s proposed 64.2 percent rate on capital gains does not include his new financial transaction tax, which would be levied on the total value of transactions, not the net gain. Counting state capital gains taxes averaging about 4 percent, the total top marginal rate in the United States would approach 70 percent.
The major European countries, including those with the most progressive welfare-state policies, have generally avoided raising taxes on capital to such high levels out of concern about the consequences for investment and growth. To finance their social programs, they have instead relied heavily on value-added taxes, which fall on consumption. Sanders has been oblivious to the prudential concerns that have led socialist as well as other parties in the economically advanced democracies to limit their taxation of capital. A broad-based consumption tax like a VAT would require more direct sacrifice from the public at large. But taxes focused on the rich, corporations, and finance seem to promise no sacrifice at all from the electorate.
Why Democrats Should Be Concerned
Considered one at a time, many of Sanders’s tax proposals have genuine merit. A carbon tax would promote alternative energy, a financial transaction tax would penalize high-frequency trading, and more progressive income and estate taxes would help counteract the trend toward rising economic inequality. If we need more funds for public purposes—and I believe we do—taxes on carbon fuels and financial transactions and more progressive income and estate taxes are legitimate ways to raise that revenue.
More than 3,000 supporters listen as Democratic presidential candidate, Senator Bernie Sanders addresses the crowd Saturday, April 2, 2016, a a campaign rally at the University of Wisconsin-Eau Claire.
But it is not possible to evaluate all the tax increases Sanders has proposed in isolation from one another and without considering their magnitude, likely economic and political effects, and public needs that his proposals leave unaddressed. Adopting these increases in taxation together, at the levels Sanders is talking about, and for the purposes he proposes to use them would be not only politically impossible but also unwise and dangerous.
The magnitude of the changes ought to raise both political and economic concerns. Americans have limited tax tolerance. The introduction of major new forms of taxation is a rare event in American history. When new federal taxes have been adopted—such as the income tax in 1913, the estate tax in 1916, and the payroll tax in 1935—the initial rates have been extremely low. The income and estate taxes also originally applied to a very small fraction of the population; they became forms of mass taxation only as a result of the two world wars. The payroll tax began at 1 percent on both employer and employee and was raised in small increments, typically less than half a percent at a time. There is no peacetime precedent in American history for anything like the package of tax increases Sanders has proposed: a 6.4 percent increase in payroll taxes, a dramatic increase in income taxes on top earners, increases in corporate taxes, and two entirely new taxes on carbon fuels and financial transactions.
If Sanders’s whole program of new taxes and expenditures were put before the Senate today, I doubt there would be a single vote for it besides that of the Vermont senator himself. If he were elected president, Democrats in Congress would beg him to back off these proposals before their certain rejection. Hypothetically, however, let us imagine the legislation passed, and the United States increased these taxes to Sanders’s proposed levels. The political dangers would be even greater as a result of the likely backlash.
Major increases in taxation have taken place during major wars for a good reason. Wartime emergency not only creates the political basis for raising taxes in a big way; it also fosters the popular, patriotic support for complying with them. A massive tax increase without that foundation stands the risk of provoking massive resistance.
Our tax system depends on voluntary compliance. The Internal Revenue Service ought to be better funded and to conduct more audits. But even when it has been relatively well funded, tax collections have rested on a “tax-paying culture.” This is not something to take for granted; many other countries without that kind of culture see their taxes flouted with impunity, their governments unreliably financed, and both the rule of law and the fabric of social trust weakened as a result.
Earlier I noted that the TPC provides only static estimates of the revenue from the changes in tax policy that Sanders and other candidates have proposed. Dynamic estimates take into account behavioral changes and are always ideologically contentious. Conservative economists such as those at the Tax Foundation project ill effects from more taxing and spending, while left-of-center economists tend to estimate more benefit than cost. But when the proposed changes are far outside any recent historical experience, there is the additional problem that relevant data are unavailable.
Supporters cheer as Democratic presidential candidate, Senator Bernie Sanders speaks at a rally Monday, March 21, 2016, in Salt Lake City.
This is one source of the difficulty with estimates of “optimal” taxation based on historical data. The United States did have high progressive tax rates on income during World War II and the Cold War, but compliance with the rates in that era may not be a reliable basis for estimating compliance now. The federal government in the mid-20th century enjoyed a degree of trust and confidence it no longer has; federal expenditures were heavily oriented to defense. Tax evasion today does not carry the stigma of disloyalty and lack of patriotism. Global markets and the emergence of currencies like Bitcoin and Ethereum also provide new avenues for hidden transactions. Taxes certainly can be increased on people with top incomes. Using historical data, Thomas Piketty, Emmanuel Saez, and Stephanie Stantcheva argue the top marginal rates could be raised to levels even higher than Sanders proposes. But, under the new economic and political conditions that exist today, both tax avoidance and tax evasion could limit the revenue that high marginal rates raise in practice.
Two of Sanders’s proposed tax changes—the increases in capital gains taxes and the financial transaction tax—are very likely at counterproductive levels. The Congressional Budget Office estimates that the revenue-maximizing level for capital-gains taxes to be around 32 percent. The CBO estimate may be too low, but at 64.2 percent Sanders’s proposed rate seems likely to backfire. Faced with such high taxes, people with appreciated assets have strong incentives to hold onto them rather than sell them, reducing the revenue the tax generates. The danger here is also a “lock in” effect—capital gets locked into current uses, limiting new investment and potentially damaging innovation and economic growth.
Sanders’s financial transaction tax—from which he expects to raise $75 billion annually to pay for free public college tuition and relief of student debt—raises a similar problem. It is not clear that the tax can be set at a rate that would generate $75 billion. The higher the tax rate, the more likely investors are to shift transactions to foreign markets or to change the form of assets and transactions to escape the tax. According to a TPC analysis of a financial transaction tax similar but not identical to Sanders’s, the revenue-maximizing rate on stock transactions would be .34 percent and little revenue would be lost at .10 percent; Sanders’s proposed rate is .50 percent. The TPC report adds that “an FTT at the rates being proposed by Senator Sanders would discourage all trading, not just speculation and rent seeking. An FTT appears as likely to increase market volatility as to curb it, as it would create new distortions among asset classes and across industries.”
If the taxes Sanders proposes don’t generate as much revenue as the programs require, he would need other taxes—or debt—to make up the difference. And that may well happen anyway because of the real cost of the programs, particularly for health care.
Paying for Single-Payer Health Care
The idea that a single-payer health plan could reduce net costs to consumers despite a tax increase has an intuitive appeal. Sanders points to other countries with national insurance programs that have lower costs than the United States. If only the United States adopted a single-payer system, the argument goes, Americans would save more in insurance premiums and out-of-pocket costs than they would pay in new taxes. Insurance companies are so little loved that doing away with them would hardly be a matter of deep popular regret.
I don’t doubt that nationalizing the insurance industry and adopting a uniform federal system could reduce administrative costs significantly and provide the government with the monopoly power to reduce prices for drugs and other health-care goods and services. At the same time, however, the Sanders health plan increases health costs in other ways, depends on economically and politically implausible developments, and provides no incentive for providers, patients, employers, or state governments to control costs. It would also confront intense resistance from Americans with employer-provided private insurance, the majority of whom would see themselves as losing out under the plan.
Although Sanders refers to his plan as “Medicare for All,” it is far more generous than Medicare. It covers every sort of health care—inpatient, outpatient, long-term care, home care, dental care, vision, and so on—with no patient cost sharing. The only limitation appears to be cosmetic surgery. The economist who estimated costs for the plan, Gerald Friedman, assumes it would cover 98 percent of all national health expenditures. Health care currently represents 18 percent of GDP, of which the federal government pays for just over half (counting tax expenditures for private health insurance and medical deductions). If the federal government is to become the single payer, it would have to pay for the remaining half, though the additional cost in federal expenditures may be lower if there are net savings and higher if there are net costs.
As Thorpe has detailed in an exchange with Friedman in the Prospect, Friedman’s estimate of the additional cost to the federal government—$1.38 trillion annually—is too low for several principal reasons:
· First, although health-care goods and services would be available at zero price to consumers, Friedman sees only a 3 percent increase in health-care utilization. Using standard estimates for the effects of reducing out-of-pocket costs to consumers, Thorpe finds a much larger impact.
· Second, the Sanders plan counts on the states continuing to pay their share of Medicaid and the Children’s Health Insurance Program, even though both programs would no longer exist and Supreme Court rulings wouldn’t allow Congress to require states to make such payments for the new federal program.
· Third, in one sector after another, Friedman projects implausible savings—for example, $241 billion out of a projected $350 billion in annual spending on prescription drugs. And he projects an overall growth rate in costs lower than the recent experience of single-payer countries.
Once Thorpe plugs in more realistic numbers, 70 percent of those with private insurance today come out losers under the single-payer plan.
We don’t need to rely on Thorpe’s numbers, however, to predict plenty of resistance to single-payer from the privately insured. People with employer-provided insurance generally see their monthly premium contribution as the price they pay for coverage; they may be only dimly aware of what their employer contributes, and in any event they may not believe they will get that money in higher pay if the government takes over the insurance function. So when they compare their monthly premiums with the tax costs under single-payer, the latter looks worse than if they had full information and knew with certainty that their wages would increase.
Americans with employer-provided insurance may also not be confident of what a government monopoly over insurance would mean for them in the long run. The great appeal of the “public option” proposal in the run-up to the Affordable Care Act was precisely the idea that it would be optional; people wouldn’t be forced into a governmental program. Single-payer implies a federal monopoly on insurance; the government as single payer would become the single decider of all important decisions about technologies, treatments, investments, and so on.
For all its problems, the current system has interested a number of parties—employers, insurers, providers, state governments, and others—in the effort to control health costs and improve the quality of care. Nationalizing the finance of health care implies nationalizing the interest in limiting health-care costs; it eliminates the government’s partners in cost control.
If the United States had adopted national health insurance in the 1940s, when health care represented only about 4 percent of GDP, the health-care system would probably have grown more slowly than it has, primarily because it would have competed for budgetary appropriations with the Defense Department and other claimants on the Treasury. Instead, the United States opted for a mixed, public-private system with high deference to physicians and other health-care interests. That pattern of accommodating private interests led not just to enormous increases in costs but also to a steep barrier for anyone wanting to change the system now. If health care represented as small a percentage of the economy as it did in the 1940s, it would be a lot easier to substitute taxation for private payments. But because health care is 18 percent of GDP today, bringing it all into the federal government requires an enormous increase in taxes.
Even if we could raise taxes to that extent, there would be a question about political opportunity costs. Sanders’s health-care plan (as well as his college-tuition plan) would substitute federal revenue for a great deal of current private and state-government spending. If the federal government used so much revenue for substitution purposes, it would not be able to do other things that aren’t being done now at all. As big as it is, the Sanders agenda leaves a lot out. It puts billions into higher education, but day care, pre-K, and K-12 education don’t get the same priority. Neither do other public needs such as housing. Although his program is highly redistributive, it isn’t particularly concerned with the poor. But once Sanders had raised taxes to Scandinavian levels to nationalize health insurance and much of higher education finance, what could be done about other problems? What would happen if those taxes aren’t even enough to cover the costs of the health-care plan? There is a limit to how much even Bernie Sanders can squeeze out of the rich.
Some of Sanders’s supporters are thrilled with his willingness to back a big expansion of government. If Sanders becomes the Democratic nominee, Republicans may be thrilled too because his tax proposals offer so rich a target in the fall. Democrats owe it to themselves to have a thorough debate about taxes before they find themselves with a presidential candidate whose platform is impossible for most of the party’s candidates for Congress to back.
Note: An earlier version of this article mistakenly attributed figures on marginal income rates in Sanders's tax plan to James Galbraith.