Senator Jeanne Shaheen (D-NH) and a score of Democratic cosponsors want to use the tax code to discourage direct-to-consumer advertising by drug companies. Their bill, the End Taxpayer Subsidies for Drug Ads Act, would prohibit firms from taking tax deductions for any consumer advertising of prescription drugs.

Limiting tax deductions is a blunt and arbitrary way of approaching a legitimate concern. Consumer drug ads play an important role in debates about the costs of prescription drugs, the risks of misuse and overuse of some medications, the balance of authority between doctors and patients, the limits of commercial speech, and a host of other issues. For overviews, see here, here, and here.

But the bill is not well crafted to address those issues. The problem starts with the legislation’s name: Allowing drug companies to deduct advertising costs is not a subsidy. Many other deductions are: The charitable deduction in the personal income tax, for example, subsidizes charitable giving. And the mortgage interest deduction subsidizes borrowing to buy a home.

But the business deduction for advertising costs is not a subsidy. The corporate income tax is a tax on corporate income. To calculate income properly, businesses tote up their revenues and deduct their expenses. Those expenses may include wages for workers, rent for office space, and yes, the costs of advertising.

Under an income tax, companies deduct those expenses because they incur them in pursuit of the profits we have chosen to tax. One can debate how rapidly companies in any industry should write-off their advertising costs. But in an income tax, there is no question that they should write them off.

Rhetoric aside, the broader question is whether limiting deductibility is a good way to discourage consumer drug advertising. Using the corporate income tax to impose penalties this way has the same strengths and weaknesses as much more common efforts to offer rewards.

On the plus side, the tax code provides ready infrastructure for creating a financial penalty. With little legislative effort (the bill is less than three pages), lawmakers can design a meaningful deterrent to consumer ads.

But limiting deductibility is a blunt and arbitrary instrument. In principle, lawmakers should discourage ads based on the harms they want to reduce. Congress should impose large deterrents against the most damaging forms of consumer ads, smaller disincentives to less damaging ads, and rewards for beneficial ads (there is evidence some consumer drug ads create benefits).

Ending the tax deduction does not allow such careful calibration. Instead, it creates a single financial penalty based on the corporate tax rate. Recent tax changes illustrate how arbitrary this can be.

When proposals to eliminate tax deductibility for drug ads were floated in 2009, 2015, and 2016, the corporate tax rate was 35 percent. Eliminating the tax deduction would have increased the effective cost of drug ads by more than half. Without deductibility, a $100,000 ad would have cost as much as a $154,000 ad with the deduction.

But the 2017 Tax Cuts and Jobs Act (TCJA) lowered the corporate rate to 21 percent. Now, eliminating tax deductibility would increase ad costs only by about a quarter. A non-deductible $100,000 ad would cost as much as a deductible $127,000 one.

To those not steeped in tax policy, the Shaheen bill has the same rhetorical power as earlier proposals to eliminate tax deductibility for consumer ads. Indeed, it may have even more rhetorical power – a similar billgarnered only four sponsors last year. In practical terms, however, the bill has lost half its economic effect since passage of the TCJA.

For better or worse, advocates for limiting the tax deductibility of drug ads have lowered their ambition. Such are the perils of basing policy on arbitrary parameters of the tax code, rather than focusing on the real costs and benefits of drug advertising.

So what’s a better approach? Well, as much as I enjoy talking tax, regulation should be the first line of attack here. The Food and Drug Administration should weigh the pros and cons of consumer ads and how they vary across different conditions, therapies, and advertising media.

If taxes are the only game in town, lawmakers should do the hard of work of deciding how bad consumer ads are. They do that when they impose taxes on alcohol and tobacco. They do that when they propose taxes on carbon dioxide. And they do that (for goods not bads) when they decide how big tax credits should be for electric cars and research and development. Arbitrary tweaks to the tax code are not the way to go.


Designing Carbon Dividends

Carbon dividends are the hottest idea in climate policy. A diverse mix of progressive and conservative voices are backing the idea of returning carbon tax revenues to households in the form of regular “dividend” payments. So are a range of businesses and environmental groups. Two weeks ago, six House members—three Democrats and three Republicans—introduced carbon dividend legislation.

Here is the idea: A robust carbon tax would cut emissions of carbon dioxide and other gases that are threatening our climate. It also would indirectly increase taxes on consumers and raise significant revenue. Carbon dividends would distribute that revenue back to households through regular payments, thus softening the financial blow of the tax while still reducing emissions. (Of course, the revenue also could be directed to other purposes.)

While the premise is simple, the details of implementing carbon dividends are complex. Policymakers face a range of philosophical, political, and practical issues. In a new report, How to Design Carbon Dividends, my Tax Policy Center colleague Elaine Maag and I explore those issues. Our work was funded by the Climate Leadership Council, an advocate for carbon dividends (full disclosure: I am a senior research fellow with the organization).

Two distinct philosophic views animate carbon dividend proposals. One sees dividends as shared income from a communal property right. Just as Alaskans share in income from the state’s oil resources, so could Americans share in income from use of atmospheric resources.

The second sees dividends as a way to rebate carbon tax revenues back to the consumers who ultimately pay them.

Though these ideas can be complementary, they have different implications for designing carbon dividends. The communal property view, for example, implies people should receive identical dividends that would be taxed as a new source of income. But those who see dividends as a tax rebate would link the payment to a person’s carbon tax burden and make the payment tax-free. Similar differences arise for dividend eligibility, the treatment of children, and other issues.

Elaine and I show how dividends should be designed under each of these views. But we do not recommend either pure approach. Political and practical concerns also matter.

Instead, we suggest a hybrid model that combines the best of both ideas. As a starting point, we suggest US residents with Social Security numbers (a group that could be expanded) would be eligible for dividends that would be paid out quarterly.  Adults would receive equal dividends. Children would receive half as much as adults. Dividends would not be taxed as income, nor would they be treated as income in means-tested benefit programs such as SNAP (food stamps).

Our proposal strikes a balance among the philosophical, political, and practical issues. Paying identical dividends to adults, for example, reflects the communal property view – but halving the benefit for children nods in the direction of a rebate. Paying dividends quarterly reflects both practical and political concerns. They’d be less complicated, less costly to administer, and more visible than smaller, monthly payments.

Unfortunately, the dividend would be somewhat less than the amount of revenue collected by the tax. If life were simple, outgoing dividends would equal incoming carbon tax receipts. But rebating all the revenue would significantly increase budget deficits.

Why? Because it would cost the government money to manage the tax and dividend program. And the government itself would bear some of the cost of the carbon tax. To keep from adding to the deficit, the government must keep enough revenue to cover those costs.

To illustrate, suppose a carbon tax of $43 per metric ton goes into effect in 2021, as the Climate Leadership Council has proposed. Gross revenues would be about $200 billion. It would cost the government about $6 billion to operate the dividend program and cuts in other revenues and higher spending would total more than $40 billion. Thus, to avoid deficit increases, the dividend pool would be about $150 billion.

Even with these offsets, we estimate that eligible adults would receive an annual dividend of $570 in 2021 while a family of two adults and two children would get $1,710.

Different design choices could raise or lower these amounts. Collecting income tax on dividends, for example, would allow larger dividends. But the after-tax dividend would be the same, on average, with more going to people in lower income tax brackets and less to those in higher ones. Different treatment of children also would change dividends. Giving children full dividends, for example, would reduce dividends for adults, but increase them for families with children. Dividend amounts could also vary depending on eligibility, participation rates, the size of the carbon tax, and other factors.

In short, details matter in designing carbon dividends. We hope our paper provides a useful guide to those details and the choices and tradeoffs that carbon dividends pose.

Record stock buybacks—driven in part by the corporate tax changes in the Tax Cuts and Jobs Act (TCJA)—have sparked a media and political furor. Unfortunately, they’ve also created a great deal of confusion. To help elevate the debate, here are three things you should know.

1. Repatriated overseas profits are the main way TCJA is boosting buybacks

By slashing corporate taxes, TCJA will boost after-tax profits and cash flow. Companies will use some of that cash to buy back shares. But that is not the main way TCJA is fueling today’s record buybacks.

The big reason is the “liberation” of around $3 trillion in overseas profits. Our old system taxed the earnings of foreign affiliates only when the domestic parent company made use of them. To avoid that tax, many companies left those earnings in their affiliates. They could reinvest them in their foreign operations or hold them in U.S. financial institutions and securities, but they couldn’t use them for dividends to parent company shareholders or stock buybacks.

By imposing a one-time tax on those accumulated profits, the TCJA freed companies to use the money wherever they wanted, including in the United States. And multinational firms are leaping at the chance. Cisco, for example, says they are repatriating $67 billion and buying back more than $25 billion in stock.

Cisco’s response reflects a broader trend. Repatriated profits will account for two-thirds of this year’s increase in stock buybacks, according to JP Morgan. Stronger earnings, due to both improved before-tax profits and lower taxes, make up only one-third.

2. Buybacks do not mechanically increase stock prices

Buybacks can help shareholders. But it’s not as simple as much commentary suggests.

The common story is that buybacks boost stock prices by reducing the number of shares outstanding. That sounds like basic economics: Reduce supply, increase price. But buybacks have a second effect that pushes the other way.

When companies pay out cash, their value falls. This effect is easy to see when companies pay dividends. On the morning after a dividend, a company’s stock price usually drops. One day the stock price reflects the company’ s operating value plus the cash it will use for the dividend. The next day it’s just the operating value.

That also happens when a company buys back stock. It spends its cash, so the value of each share declines. But this decline isn’t conspicuous. It doesn’t happen on an announced day (as with a dividend payment), and it happens at the same time the supply effect is pushing the other way. So commentators tend to overlook the stock price hit from cash going out the door.

How do the two effects net out? It depends. For starters, suppose we live in a frictionless (and mythical) world. There are no taxes, everybody is well informed, executives are perfect agents of shareholders, and it’s costless to raise new capital. In that world, the two effects exactly offset. The decline in the stock price from paying out cash matches the increase in the stock price from having fewer shares outstanding. And the stock price does not change.

In this world, a dollar on the company’s balance sheet has the same value for investors—one dollar—as a dollar on the investors’ own balance sheets. Stock buybacks simply move some dollars from the company to its investors. But they don’t create or destroy any value.

Of course, the real world isn’t frictionless. Taxes matter, as do imperfect information, misaligned incentives, and the cost of raising capital. The question is how do they matter. Is a dollar on a company’s balance sheet worth more than an investor’s dollar? Less? The same?

It depends. Suppose a company has many promising investment opportunities. If raising capital is expensive, a dollar inside the company may be worth more than a dollar outside. Investors value the ability to pursue good opportunities without the burden of raising new capital. In this case, a (misguided) stock buyback would drive the value of the stock down, not up.

Another company might have few promising investment opportunities. Shareholders might worry executives will squander extra cash, perhaps through empire building, executive perks, or just taking their eye off the ball. Investors might therefore treat a dollar of cash inside this company as worth less than a dollar, perhaps 90 cents or even as little as 40 cents. In these cases, stock buybacks would drive the value of the stock up.

On average, stock buybacks are more common in the second scenario than the first. So buybacks do tend to lift stock prices. Not because of the mechanical link between supply and price, but because shareholders value a dollar on their own balance sheets more than a dollar inside the companies buying back stock.

3. Today’s buyback furor reflects a much larger debate about shareholder capitalism.

Concern about buybacks didn’t start with the TCJA. Some commentators have long worried that shareholders can be shortsighted. They may focus too much on next quarter and too little on next decade. If so (there is much debate), shareholders may undervalue cash inside companies. Buybacks may thus be too large, hurting shareholders themselves and the broader economy.

Others object to shareholder capitalism more fundamentally. Comparing the value of dollars inside and outside a company presumes those dollars belong to shareholders. Our system largely works that way today. But what if workers and other stakeholders also have a claim? If so (again there is much debate), buybacks may allow shareholders to capture all the value from what should be shared resources. This concern animates Senator Cory Booker’s recent proposal to limit stock buybacks.

Evaluating these concerns is a job for another day. For now, remember that the stock buyback debate boils down to three basic questions. When do investors think money inside a company is less valuable than outside? How well do investors make that judgment? Should investors be the only ones with a claim on a company’s money? Debate over those questions will continue long after today’s focus on the TCJA.

This post originally appeared on TaxVox, the blog of the Tax Policy Center.

For your weekend listening pleasure (?): I visit the ReConsider podcast to chat money, inflation, fiat currencies, gold, Bitcoin, & Uncle Sam’s balance sheet. Starts at 4:59.


The tax treatment of employee stock and options raises a classic Goldilocks problem. We want to tax this compensation neither too much or too little. In a recent policy brief, I consider three questions about how to strike that balance.

Do companies get excessive tax deductions for employee stock and options?

This concern rocketed to prominence in 2012 when Facebook went public. Its employees earned billions from their stock options and restricted stock units. The company, in turn, got billions in tax deductions, reducing its income taxes for years.

Those deductions outraged some observers who asked how Facebook could get billions in tax write-offs when its financial statements showed much lower compensation costs. Lawmakers on both sides of the aisle denounced the “stock option loophole” and proposed limiting these deductions.

While there are good reasons for Congress to worry about companies gaming the tax code, this is not one. The tax deductions that companies receive for employee stock and options are, with few exceptions, just like those for cash wages, salaries, and bonuses.

Here is an example: Suppose Esther has 1,000 options from her employer, Acme, Inc. Each allows her to buy a share of Acme stock for $10. If Esther exercises her options when Acme stock is worth $15, she pays $10,000 for stock worth $15,000 and thus has a $5,000 gain. Esther pays ordinary income taxes on the $5,000 while Acme deducts $5,000 as compensation. And both Esther and Acme pay payroll taxes on the $5,000.

Esther’s stock options are taxed just as if she received a $5,000 cash bonus. The deduction for Acme, just like the deduction for Facebook, is an integral part of our income tax system. Employees pay income taxes on their compensation, and businesses get a corresponding deduction. Lawmakers should maintain the parallel treatment of cash and equity compensation in any tax reform.

Does taxing options at exercise pose a special challenge for employees of private companies?

Employees owe income and payroll taxes when they exercise most options. That’s not a problem for employees who have easy access to the cash they need to pay their tax bill. But it can be a problem for employees who find themselves option rich and cash poor.  Employees of publicly traded companies can always cover their tax bill by selling stock. But employees of private companies often can’t. This problem is rare, but has reportedly become more common as successful start-ups stay private longer.

A bipartisan group of lawmakers has proposed to solve the problem with the Empowering Employees through Stock Ownership Act. The bill allows employees of privately-held firms to defer their taxes from exercising options for up to seven years or until their stock becomes liquid, whichever comes first.

Deferral of tax payments would help solve the liquidity problem. But it also would give equity compensation a notable advantage over cash compensation. Deferral of tax payments would be valuable for all employees, not only those with limited liquidity. Even taxpayers who have available cash would welcome the chance to defer their taxes for several years. Charging interest on the amount of deferred taxes would be one way to maintain some balance between cash and equity compensation.  (In essence, this approach would treat the deferral of tax payments as a loan from the IRS.)

Does the AMT pose a special burden for employees who receive incentive stock options?

Employees who get a special type of stock option—known as incentive stock options—face a different tax structure. ISO gains are taxed at capital gains tax rates, not ordinary income rates. And taxes aren’t due until the employee sells their stock, which could be long after they first exercise their options. That’s a big advantage for employees. But businesses don’t get a tax deduction for the compensation. Since the loss to firms is usually larger than the gain to employees, ISOs are rare. (This could change if Congress cuts the corporate rate much more than individual rates in a coming tax bill.)

But there’s a catch. Under the Alternative Minimum Tax, ISO gains are taxed when exercised. This creates an unwelcome surprise for employees who are unaware of the AMT and its accelerated tax on ISOs. Indeed, it can create financial hardship when stock prices fall, leaving taxpayers with a big tax bill on gains that have since evaporated.

This problem was especially severe during the financial crisis. Congress responded by temporarily exempting ISO gains from the AMT. That exemption has long since lapsed. Congress could permanently fix the problem by repealing the AMT, as the GOP framework and many other plans propose.

Stock options create some unusual challenges for tax writers, and they’ll have to be careful to be sure they are treated fairly relative to other forms of compensation to avoid creating incentives for tax-motivated compensation schemes.

On Tuesday, I had the chance to testify before the Senate Finance Committee on business tax reform. Here are my opening remarks. They are a bit on the glum side, emphasizing challenges and constraints lawmakers face.  Moving from optimistic rhetoric about tax reform to legislative reality is hard. You can find my full testimony here.

America’s business tax system is needlessly complex and economically harmful. Thoughtful reform can make our tax code simpler. It can boost American competitiveness. It can create better jobs. And it can promote shared prosperity.

But tax reform is hard. Meaningful reforms create winners and losers. And you likely hear more complaints from the latter than praise from the former. I feel your pain. At the risk of adding to it, my testimony makes eight points about business tax reform.

  1. Thoughtful reform can promote economic growth, but we should be realistic about how much.

More and better investment boosts economic activity over time. The largest effects will occur beyond the 10-year budget window. If reform is revenue neutral, revenue raisers may temper future growth. If reform turns into tax cuts, deficits may crowd out private investment. Either way, the boost to near-term growth may be modest. Dynamic scoring will thus play only a small role in paying for tax reform.

  1. The corporate income tax makes our tax system more progressive.

The corporate income tax falls on shareholders, investors more generally, and workers. Economists debate how much each group bears. Workers are the most economically diverse. But they include highly paid executives, professionals, and managers as well as rank-and-file employees. The bulk of the corporate tax burden thus falls on people with high incomes even if workers bear a substantial portion.

  1. Workers would benefit from reforms that encourage more and better investment in the United States.

In the long run, wages, salaries, and benefits depend on worker productivity. Reforms that encourage investment and boost productivity would thus do more to help workers than those that merely increase shareholder profits.

  1. Taxing pass-through business income at preferential rates would inspire new tax avoidance.

When taxpayers can switch from a high tax rate to a lower one, they often do. Kansans did so when their state stopped taxing pass-through income. Professionals use S corporations to avoid payroll taxes. Investment managers convert labor income into long-term capital gains. Congress and the IRS can try to limit tax avoidance. But the cost will be new complexities, arbitrary distinctions, and new administrative burdens.

  1. Capping the top rate on pass-through business income would benefit only high-income people.

To benefit, taxpayers must have qualifying business income and be in a high tax bracket. Creating a complete schedule of pass-through rates could reduce this inequity. But it would expand the pool of taxpayers tempted by tax avoidance.

  1. Taxing pass-through business income at the corporate rate would not create a level playing field.

Pass-through income faces one layer of tax. But corporate income faces two, at the company and again at taxable shareholders. Taxing pass-throughs and corporations at the same rate would favor pass-throughs over corporations. To get true tax parity, you could apply a higher tax rate on pass-through business income. You could levy a new tax on pass-through distributions. Or you could get rid of shareholder taxes.

  1. It is difficult to pay for large cuts in business tax rates by limiting business tax breaks and deductions.

Eliminating all corporate tax expenditures except for deferral, for example, could get the corporate rate down to 26 percent. You could try to go lower by cutting other business deductions, such as interest payments. But deductions lose their value as tax rates fall. To pay for large rate reductions, you will need to raise other taxes or introduce new ones. Options include raising taxes on shareholders, a value-added tax or close variant like the destination-based cash flow tax, or a carbon tax.

  1. Finally, making business tax cuts retroactive to January 1, 2017 would not promote growth.

Retroactive tax cuts would give a windfall to profitable businesses. That does little or nothing to encourage productive investment. Indeed, it could weaken growth by leaving less budget room for more pro-growth reforms. Another downside is that all the benefits would go to shareholders, not workers.

How fast will the US economy grow? When mainstream forecasters consult their crystal balls, they typically see real economic growth around 2 percent annually over the next decade. The Congressional Budget Office (CBO) and midpoint estimates of Federal Reserve officials and private forecasters cluster in that neighborhood.

When President Trump looks in his glowing orb, he sees a happier answer: 3 percent.

That percentage point difference is a big deal. Office of Management and Budget director Mick Mulvaney recently estimated the extra growth could add $16 trillion in economic activity over the next decade and almost $3 trillion in federal revenues.

But could our economy really grow that fast? Maybe, but we’d need to be both lucky and good. We’ve grown that fast before. But it’s harder now because of slower population growth and an aging workforce. And there are signs that productivity growth has slowed in recent years.

To illustrate the challenge, I’ve divvied up past and projected economic growth (measured as the annual growth rate in real gross domestic product) into three components: the growth rates of population, average working hours, and productivity.


The link between population and growth is simple: more people means more workers generating output and more consumers buying it. Increased working hours have a similar effect: more hours mean more output and larger incomes. Hours go up when more people enter the labor force, when more workers find jobs, and when folks with jobs work more.

Productivity measures how much a worker produces in an hour. Productivity depends on worker skills, the amount and quality of capital they use, managerial and organizational capability, technology, regulatory policy, and other factors.

As the first column illustrates, the US economy averaged 3 percent annual growth over more than six decades. Healthy growth in population and productivity offset a slight decline in average hours. Of course, that six-decade average includes many ups and downs. The Great Recession and its aftermath dragged growth down to only 1.4 percent over the past decade. In the half century before, the United States grew faster than 3 percent.

Mainstream forecasters like the CBO and the Federal Reserve expect slower future growth along all three dimensions. People are having fewer children, and more adults are moving beyond their child-rearing years, so population growth has slowed. Our workforce is aging. Baby boomers are cutting back hours and retiring, and younger workers aren’t fully replacing them, so average working hours will decline. Productivity growth has slowed sharply in recent years, for reasons that are not completely clear. Productivity is notoriously difficult to forecast, but recent weakness has inspired many forecasters to expect only moderate growth in the years to come.

Proponents of President Trump’s economic agenda offer a rosier view. Four prominent Republican economic advisers—John F. Cogan, Glenn Hubbard, John B. Taylor, and Kevin Warsh—recently argued that policy, not just demographic forces, has brought down recent growth. They claim supply-side policy reforms—cutting tax rates, trimming regulation, and reducing unproductive spending—can bring it back up. They argue that encouraging investment, reinvigorating productivity growth, and drawing enough people into the labor force to offset the demographic drag would generate persistent 3 percent growth.

Many analysts doubt such supply-side efforts can get us to 3 percent growth (e.g., here, here, here, and here). Encouraging investment and bringing more people into the labor force could certainly help, but finding a full percentage point of extra growth from supply-side reforms seems like a stretch. Especially if you plan to do it without boosting population growth.

The most direct supply-side policy would be expanding immigration, especially among working-age adults (reducing our exceptional rates of incarceration could also boost the noninstitutional population). But the Trump administration’s antipathy to immigration, and that of some Republicans in Congress, pushes the other way. Cutting legal immigration in half over the next decade could easily take 0.2 percentage points off future growth (see this nifty interactive tool from ProPublica and Moody’s Analytics). Three percent growth would then be even more of a stretch.

Another group of economists believes that demand-side policies—higher spending and supportive monetary policy—could lift growth above mainstream forecasts.

One trio of economists took a critical look at past efforts to forecast potential GDP growth, a key driver of long-run growth forecasts. They conclude that forecasters, including those at the Federal Reserve and the CBO, have overreacted to temporary economic shocks, overstating potential growth when times are good and understating it when times are bad. We’ve recently had bad times, so forecasters might be underestimating potential GDP almost 10 percent. If so, policies that boost demand could push up growth substantially in coming years. (For a related argument, see here.)

So where does that leave us?

Well, every crystal ball (and glowing orb) is cloudy. We should all be humble about our ability to forecast the economy over the next decade. Scarred by the Great Recession and its aftermath, forecasters may be inadvertently lowballing potential growth. Good luck and good supply- and demand-side policies might deliver more robust growth than they anticipate. But those scars remind us we can’t always count on good policy, and luck sometimes runs bad.

We can hope that luck and good policy lift growth to 3 percent. But it’s prudent to plan for 2 percent, and pray we don’t fall to 1 percent.