Just two brief housekeeping notes before this edition of The Raspberry Patch gets underway. First, sorry for not posting on Friday—our usual day—as promised on Tuesday of last week. Other stuff just got in my way.
Second, after posting the SCOTUS essay on March 5 I went back and added a clarifying note. I deliberately avoided reciting the history of public election financing in the United States, which dates only from the mid-1970s, because it is complicated and would have distracted from the main points of the essay. But then, a few hours after posting it, I realized that I needed to say something, particularly for those unaware of what a recent development it is—younger folks in particular. More, of course, can and ought to be said about it, since it engages important questions about liberal democratic theory and practice. So, instead of making you look up the revised post, let me only say, very briefly:
In liberal democratic orders, especially in a Tudor republic like ours, law is sovereign and parties or factions vie to serve within constitutional-legal parameters. In a party-state arrangement, as the comparative politics texts tell us, the party dominates the state and its constitutional framework, and uses its dominance to guarantee its tenure against all non-party challengers. Historical examples of party-state arrangements are too numerous even to list but they range from fully totalitarian examples (the Soviet Communist Party, the German Nazi Party) to mere authoritarian/corporatist examples (Franco’s Spain, Perón’s Argentina, and, at the tender loose end of the type, Singapore’s Peoples Action Party). After the creation of the Federal Election Commission in 1975 and the first uses of public money in federal election mechanics, the United States created for itself a lesser and milder hybrid case of a party-state that protects not a single tenured party from honest electoral competition but a two-party biumvirate. A good case can be made that this innovation was and remains unconstitutional in our system, most certainly in the case of state party primary elections, and arguably also in the case of presidential elections. Each of the fifty states obviously needs to spend some public money to hold an election for president, despite the wise customary uses of volunteer poll workers. But when the Federal government offers public money to candidates, even if it is voluntary, it prejudicially excludes those who have been unfairly denied ballot access. We need to find other ways of limiting the influence of huge wads of corporate money skewing election results. One option is to limit by law the duration of the campaign period, as most liberal democracies do.
More on that later, but now, without further delay, to the advertising business.1
A Dilapidated Assumption
Beyond the purchase of goods, government at various levels within the U.S. federal system taxes a great many services: payment for airline tickets, rental cars, toll roads, utilities, hotel rooms, lottery purchases, financial transactions, cell phone usage, and much more. Some states, like Pennsylvania, also tax funeral expenses, a low blow if ever there was one. Others tax and are proposing to tax digital streaming services. But at no level of American government is advertising taxed.
On the contrary: The Federal tax code treats advertising as a deductible business expense, meaning that in effect taxpayers subsidize it. U.S. tax law also allows for the quick amortization of advertising expenses, so we probably get even more of it then we otherwise would. And we get a whole damned lot: U.S. corporate advertising buys, which exceed the combined totals of ad buys in the United Kingdom, Germany, Japan, and China, are by far the largest in the world. At least we’re still number one in something…..
Why is this? We do not avoid taxing advertising because it so hard to do that it would be counterproductive. It would be as easy to tax advertising as it is to tax any standard point-of-purchase contract arrangement. Rather, various national leaderships—not just the American one—have come to believe that advertising is good for economies. Since, as everyone knows, you get less of what is taxed and more of what is subsidized, a consensus grew that it would be unwise to tax advertising because, all else equal, it would conduce to a smaller economy. And in the enduring if increasingly patina-tarnished age of market fundamentalism—usually called neoliberalism in Europe—questioning this shibboleth is virtually taboo.
It shouldn’t be. The problem with this understanding of why advertising buys are not taxed is that its main premise is, or rather has become, mainly wrong. All else is no longer equal. Mainly wrong instead of entirely wrong because this is, in truth, a bit complicated. Still, the essence is easy to state. Advertising is or can be good for the economy and for the public weal generally under two conditions: when it provides information about the function, quality, or price of a good or service; and when it is not intrusive. For example, nearly all economists agree that advertising with clear and accurate price signals tends to be anti-inflationary, so more of it could be good.
Most advertising used to fall into these categories a century or so ago, the snake-oil factor obviously excluded. But with the professionalization of advertising and its increasingly sophisticated use of cognitive psychological tricks and deceptions to get to the bottom of consumer brain stems, especially in the still-young digital age, this is no longer the case. Alas, businesses—particularly large businesses—now do things they did not do, because they could not do, in the past thanks to this frankly disturbing trend.
For a minor but cloying example, when costs increase for processed or packaged food, whether in real terms or nominally because of inflation, some large purveyors have taken to keeping the price of packages the same while reducing their volume. Consumers are paying more per unit of what they are buying but the marketers understand that while consumers are sensitive to price rises on products they buy repeatedly, they are much less sensitive to incremental changes in the size of containers. For consumers to grasp what is being done they have to look for the volume indicator on the package, while the price is stated prominently on the shelf margin in nearly all stores. Some manufacturers engaged in this deceptive gimmick have slightly redesigned their packages to put the volume indicator in smaller print, have printed it in harder-to-find places, and/or have printed it in a color that contrasts poorly with the package background color to make it harder to find and to read.
There has always been marketing, of course, for as long as people have worn clothing. and the warning caveat emptor is justifiably an old one. It may even date all the way back to just after the ribbon-cutting ceremony that opened the Silk Road thousands of years ago, when spice merchants and middlemen soon began adulterating spices to enhance their appearance and add weight.2 Contemporary companies that do these kinds of things tell consumers exactly what they think of them: You, the buyer, are not half of a fair-minded contractual, positive-sum transaction with a seller, but are rather clueless prey to be tricked and exploited. And these companies still want us to be all right with allowing them to deduct advertising costs as a business expense. We are saps and suckers—and that’s nice language—to be all right with that.
A Simple Two-by-Two Matrix
These two elemental distinctions—informative or not, intrusive or not—enable construction of a simple two-by-two matrix. On one axis we distinguish between advertising that provides useful market signals and advertising that may be slick but is functionally useless. On the other we can distinguish between intrusive forms of advertising and forms that we may even choose to expose ourselves to for one reason or another. Indeed, sometimes it makes sense to seek out advertising in the process of shopping for some good or service because of the market-relevant information it can provide.
So, for example, a television ad that tries to sell an expensive car to rich men by posing a sexy woman in a slinky black dress near the car—suggesting subliminally that if a rich guy buys the car he gets the woman, too, but otherwise telling us nothing about the car—is worse than useless. It is a diseconomy in economic terms because it disorganizes via cognitive clutter the stock of knowledge consumers have about that class of products. It is also arguably in bad taste, but then most television advertising tends either to be in bad taste or in bad humor (read: those annoying Geico and Progressive auto insurance ads that tell us nothing of substance about the service being advertised).
On the other hand, if advertising revenue handed over to a television station brings us a baseball game to watch for free, then we may submit to viewing the advertising in order to get something we want, even if we are not the least bit interested in plastic shaving gadgets, crappy beer, or erectile dysfunction medications. A highway billboard, contrarily, is ugly and grabs our attention whether we want it to or not thanks to our evolved cognitive novelty bias. If it happens to deliver a small bit of useful information as we zoom by it, that’s one thing—it then becomes, clingingly at least, an intermediate case on the matrix as neither economically very positive nor wholly negative. But if it’s garden-variety ugly, intrusive by its very nature, and provides no useful information about any product, service, or price point, it certainly qualifies as a diseconomy and an aesthetic pain in the ass to boot.
Any scheme to tax advertising buys, both digital and more conventional forms, would therefore have to discriminate among these four varieties of advertising, one per quadrant of the matrix. It would not tax much or perhaps at all advertising that furnishes usable market signals and is not intrusive. It would tax heavily advertising that furnishes no useful market signals and that is intrusive. It would tax mixed cases in-between. And it would tax to the sky intrusive forms of advertising that are demonstrably unhealthy and unsafe.
Unsafe at Any Speed
How can advertising be unsafe, you ask? Not just by advertising unhealthy products—remember the Marlboro Man? The technology of advertising itself can be unhealthy.
As most readers are aware, both highway billboards and urban marquee displays are increasingly moving to digital technology and away from merely large print. Why? They attract more and more sustained attention that way, being a kind of novelty tease within what already touches off our evolved novelty bias. They can be rigged to advertise more than one thing in rotational array, too. They are certainly much more visible in the dark even than illuminated conventional billboards, so more eyes alight on them in the total lifespan of the advertising buy.
But digitized ads in public spaces also exact a higher cost on viewers in terms of neural attention calories expended. Additionally, the wavelengths and intensities of the light they typically choose increase eye strain, stress, and induce fatigue if you see enough of them on a long highway drive. Seeing too many bright lights at night is especially problematic for those predisposed to suffer from high blood pressure, migraines, glaucoma, and nasal allergies.3 LED illuminated billboards also flicker, adding extra stress and imperiling those prone to epileptic seizures.
Digital billboards are even more dangerous for city driving. They influence glance duration and hence driving behavior in areas crowded with pedestrians, cyclists, and of course other vehicles, and the more intense the light the more they distract drivers from looking at where they’re going at 25, 35, 45 miles per hour. Nighttime crash rates around digital billboards and marquee displays are as much as 30% higher than when they are not present; cover them up and the rates drop; let them loose again and the crash rates go back up. They are unsafe, in other words, because of how and how much they distract us—all of us, really, passengers as well as drivers—who cannot not look at them because of our ever vigilant novelty bias.
Indeed, trying not to look exacts an even higher price in neural caloric expenditure, for it becomes a cognitive task requiring agency and focus. Commercial distractions of all kinds thus function a little like second-hand smoke in that their mere presence affects us. For that reason, “shiny object” digital billboards—especially in cities but even outside cities—should be taxed out the wazoo if they cannot be banned, as most should be, for safety reasons.
That would not be an exceptional precaution. The Federal government has for a long time monitored and banned television advertising directly aimed at children, it being obvious that young kids cannot readily distinguish show content from ad content, and cannot suss out the motivations behind cleverly engrossing advertising. But no such ban exists for ads aimed at kids on the internet; advertisers have been having a rich field day—make that a rich field decade—tricking kids into spending their parents’ money. This practice is obscene and should be banned.4 But if our plutocratized, zombified politics makes that impossible, the least we should do is to tax those ad buys out the wazoo, too—just as with digital billboards and displays in densely trafficked areas.
Calculating a Pigouvian Tax on the Back of an Envelope
Taxing advertising would thus be part revenue raiser and part Pigouvian tax (named after British economist Arthur C. Pigou about a century ago) designed to pay for externalities and, possibly, to re-shape market behavior for the better. So if in order to reduce their tax bills advertisers moved from deceptive, information-free gimmicks to ads that actually tell us something useful and truthful about a given good or service, would you feel slighted? If nearly all of the useless, ugly billboards on our highways were to disappear as a consequence, I for one would not be the slightest bit put out. Would you?
To contemplate a tax on advertising we would need to know how much U.S. businesses spend on advertising in a given year, and what range of revenue taxing it at various rates might bring in. The basic data are not hard to find. In 2024 the best estimates are that U.S. businesses will spend just shy of $400 billion on advertising, of which about $300 billion will be digital and the rest print, billboards, and other non-screen-based media. That will be up about 10% from 2023, which is a typical boost for recent years. Spending on advertising has been rising fairly sharply as business sectors on balance have become more concentrated and professional techniques of manipulating would-be buyers, more easily afforded by larger corporations, have become more effective.
To set tax rates we would be wise to study the very few efforts already undertaken, one in Austria some many years ago, and a few in U.S. states, and examine why they did not work. We would need to model several variants of a Federal tax on advertising to see which one comes closest to doing what we want it to do without overreaching and so become politically toxic or economically harmful. In doing so, we might consider variants that set a threshold for taxation—maybe $50,000 per year in gross advertising buys, and/or $250,000 in gross company revenue—since we don’t want to hurt small businesses and small ad jobbers and cause them either to disgorge employees or go broke. Above that threshold rates could be prorated for every $250,000 increment, just to keep the process as simple and hence as inexpensive to administer as possible. Obviously, all taxes cost administrative money to collect, and in this case we would need people to scan and classify ads according to our 2 x 2 matrix. All the more reason to keep the process as a whole as simple as possible.
States might also design advertising taxes of their own, and the states could undertake experiments over time to see what works and what does not that could inform tax policy in other states and at the Federal level. Obviously, this is complicated: Taxing ads could be initiated either without changing Federal and state tax codes or with eliminating advertising as a deductible business expense, or, alternatively, by keeping the deduction but changing the amortization schedule to make the deduction less attractive.
With all this in mind, if we start from the $400 billion figure, then exclude small ad buyers and ad jobbers, and set a basic Federal tax rate at 5% for ads that are information-free and intrusive, 1% for ads that contain useful information and are not intrusive, and 2.5% for the two intermediate categories on the matrix—and say 20% for ads that are unsafe if they cannot be banned—we might end up with something like $10 billion. If we then subtract the costs of managing the tax at a robust 10%, we’d end up with perhaps $9 billion. Just for comparison’s sake, note that the entire National Science Foundation budget for FY 2024 is about $9 billion. Of course, if instead of 5%/1%/2.5% we decided on 10%/1%/5% we would raise perhaps double that amount. $18 billion is already a pretty nice piece of change, even in a government as large and expensive as ours.
Note again that a tax constructed in this manner might over time persuade advertisers to move away from glitzy but informationally vacuous forms of advertising to more economically useful forms, and away from intrusive toward non-intrusive forms. That would be a good thing in and of itself as far as I’m concerned. I would shed no tears if gasoline pumps at self-serve filling stations (in other words, gasoline pumps everywhere in the United States except in New Jersey) stop squawking irritating noises at me as I’m pumping my gas. Having what I would describe as an average male temper fuse, I make sure not to carry a baseball bat or a sledgehammer with me in the car…..for obvious reasons.
Moving the Overton Window
Speaking of the obvious, at a time of sprawling Federal debt one would think that so-called—not entirely unjustly—tax-and-spend Democrats would have proposed taxing advertising long ago. They haven’t: Apparently, obvious is a synonym for common sense, and we all know what Voltaire said about that. At the Federal level no one has raised this idea in recent memory and, to the best of my knowledge, neither has anyone in the states. In Maryland, with a new Democratic governor in place, all sorts of proposals to raise more revenue are being bandied about these days since the state surplus bequeathed by Governor Larry Hogan’s administration has been turned into a large deficit in record time. Not one of the proposals in circulation suggests taxing advertising.
This needs to change. The Overton Window needs to be pushed down the street so that the concept of taxing advertising at least starts to register between people’s ears. We can do this. Taxing advertising is not even that bold an idea, at least not compared to any of the four parts of The Quadrivium Fix that will be coming your way in due course in this space. It would just take intelligent leadership with a bit of creativity and political courage to stand up to the corporate opposition likely to arise in the face of any proposal to tax ads.
Indeed, many sensible and useful ways to defang corporate opposition to taxing advertising would become available if we were to open up the issue of tax reform writ large. Maybe putting the possibility of taxing advertising into the public’s consciousness would stimulate energy to finally tackle long-overdue major tax reform. For example, we could shrink the IRS and boost the economy by junking the expensive-because-difficult-to collect corporate income tax, and handsomely reward corporations that reinvest earnings in domestic capital formation. To make up for forgone revenue we could do several things, among them on a short list: institute a flexible VAT tax on consumption; rejigger personal income tax rates so that only individuals in the top two and a half quintiles of income would have to pay it; get rid of the carried interest giveaway; and tax non-reinvested capital gains at a higher rate.5 But to fry that fish, which some of us would love to do, we’d need a much larger pan than the one we can use merely to tax advertising.
Oh, you say, with crooked smile and a snide impression on your face……just some intelligent, creative, and courageous leadership……is that all we need? Answer: Don’t be a cynic until you have good reason to be, and on this subject you don’t yet have a good reason. Hence Yogi: “They say you can’t do it, but sometimes that ain’t always true.”
This essay is an expanded, updated, and refined version of my brief note, “In Search of an Advertising Tax: An Editor’s Frustrations,” The American Interest, June 7, 2019.
Regrettably, such practices have not ceased even though the routes have changed; see Coral Beach, “FDA confirms contamination of children’s applesauce; theorizes it was intentional,” FSN (Food Safety News), March 1, 2024. The cinnamon additive to the applesauce, from Sri Lanka, was adulterated in Ecuador in a processing plant before it was mixed into the applesauce and shipped to the United States. Pace the article’s possibly misleading headline, no one intended to harm children, only to cheat buyers.
Noted in Richard Cytowic, Digital Distractions, chapter 15, forthcoming in October from MIT Press.
Cytowic recites the truly outrageous details in chapter 14 of Digital Distractions.
This is obviously a complex subject, with dozens of major moving parts. The best proposal for overall structural tax reform—and it’s not new, merely ignored—is Michael J. Graetz, “Tax Reform: How to Shrink the IRS and Grow the Economy,” The American Interest VII:2 (November-December 2011).