Key Findings

• Digital advertising, social media, and data tax proposals have been introduced in nine states following enactment of Maryland’s digital advertising tax, which has since been postponed a year due to administrative and legal challenges.
• Motivations for these taxes vary, from misperceptions that there is currently a tax loophole to a belief that technology and social media companies merit an extra layer of taxation, either because of their large profits or as a response to their content moderation policies.
• These taxes are administratively complex, constitutionally dubious, and likely violative of the Permanent Internet Tax Freedom Act.
• Much of the cost of these taxes will be borne by in-state companies and individuals.
• Lawmakers should be extremely cautious about using taxation to target specific industries over disagreements about those industries’ policies or practices and should reflect on the justifications for these taxes.

Introduction

Across the country, policymakers of both parties have developed a strong interest in taxing large technology companies through digital advertising taxes, social media taxes, and data taxes. What is notably absent is any consistent rationale for these new proposals. When such rationales are articulated, they are varied and sometimes inconsistent. This is not to say that those motivations are a pretense; while lawmakers are undeniably interested in such taxes as a source of revenue, something else does lie behind the specific targeting of digital activity. But the motivations vary from state to state and from legislator to legislator, and sometimes the proposals have taken on a life of their own, independent of their initial appeal.

These taxes are fraught with legal difficulties, administrative challenges, and economic inefficiencies. They tax activity that is already taxed by other means, and rationales for this additional level of taxation tend to be as weak as they are inconsistent. With enactment of Maryland’s first-in-the-nation digital advertising tax, however, some lawmakers in other states are eager to get on the bandwagon. As Maryland lawmakers are postponing implementation as they sort out administrative complexities and respond to legal challenges, now is a good time for lawmakers elsewhere to revisit these proposals as well.

Current State Digital Tax Proposals

Twelve digital advertising or data tax bills have been introduced across nine states thus far this year, though they have made little progress, as many lawmakers have adopted a wait-and-see approach regarding Maryland’s digital advertising tax. Maryland lawmakers enacted the tax with a veto override,[1] then delayed a year[2] as the state faces multiple lawsuits challenging its constitutionality and its legality under the Permanent Internet Tax Freedom Act.[3] Two bills (in Montana and West Virginia) died for lack of action in committee, and another—in Arkansas—was referred for interim study. Nine other bills, across six states, are still active, but none have received a floor vote to date.

Three states—Arkansas, Connecticut, and Indiana—have explored specifically taxing advertising on social media, with per-account levies as well. Seven proposals in six states target digital advertising more broadly, though in some cases lawmakers have articulated a particular interest in taxing social media. Finally, policymakers in two states have introduced data taxes, one (New York’s) designed essentially as an intangible property tax, while the other (Washington’s) is structured as a gross receipts tax on the sale of consumer data. While some digital advertising tax bills feature multiple rates, they are not traditional graduated rate schedules. Rather, all taxed advertising is subject to a certain rate based on the state’s amount of in-state or global revenues.

Justifications and Analysis

The origin of digital advertising taxes is twofold. First, there is a borrowing from Europe, where digital services taxes came into vogue to address perceived under-taxation of cross-border digital activity due to the structure of tax codes within the European Union. Two, interest was raised by an op-ed by economist Paul Roemer,[4] who proposed high taxes on targeted digital advertising not as a source of revenue but as a way to disincentivize or even functionally prohibit targeted online advertising. It is a backdoor way to regulate digital platforms where misinformation and hateful content flourished, and where consumers enjoyed little privacy.

The European tax model, however, is not the model of U.S. states, which use formulary apportionment to allocate corporate income to states for tax purposes. The U.S. system of apportionment is complex, inconsistent, and often flawed, but the failure to tax income derived from online advertising (or from digital services more generally) is not among its shortcomings. What in Europe could be defended as an effort to close a tax loophole—even if many of the actual proposals did much more than that—is, in the United States, pure double taxation inasmuch as the income derived from advertising is already subject to state tax.

Whatever legitimate concerns there may be about misinformation on social media or online privacy, moreover, digital advertising taxes as they have been proposed at the state level do nothing to improve it. Indeed, lawmakers clearly do not want an end to online advertising, which would deprive them of the revenue they intend to generate under digital advertising taxes. Furthermore, nothing in the structure of these taxes treats advertising differently based on how it is targeted. For tax purposes, all that matters is that an advertisement is served into the state, whether this was the result of microtargeting of 18-29-year-old Annapolis men who like baseball’s Baltimore Orioles and listen to sports podcasts, or whether it’s a completely untargeted ad served to any visitor to a major news website, or whether content on the platform is divisive or above reproach.

There is, often implicitly and sometimes explicitly, a third motivation: a desire to tax the technology industry in particular, with digital advertising—or now social media—just a convenient vehicle for doing so. Sometimes lawmakers tax tech for the same reason Willie Sutton robbed banks: that’s where the money is, or at least, that’s the perception. Particularly in the early months of the pandemic, when brick-and-mortar businesses struggled and technology companies looked to benefit from a cooped up population, many lawmakers found taxes on tech companies attractive just because they were doing well.

Separately, some wish to tax tech companies in the name of combating inequality, not that tech companies cause poverty—if anything, their services can help lift people out of poverty and undeniably improve the lives of people across the income spectrum—but because they have produced such fantastic wealth for some entrepreneurs and investors. They make an inviting target in this regard, but merely that a sector is doing well is not a reason to expose it to unique and punitive taxation. Often, moreover, such taxes backfire, ultimately hurting people other than its intended targets.

Finally, some tech companies—particularly social media companies—now find themselves in lawmakers’ crosshairs due to disputes regarding content moderation. This new development has turned what was primarily an interest among Democratic lawmakers into one with significant appeal to some Republican lawmakers who have grievances with major social media platforms and want to use the tax code to influence the companies’ policies or exact a measure of revenge.

A Republican criticism of some (not all) Democrats’ motivations for such taxes is that it inappropriately uses the tax code as a cudgel against certain disfavored businesses. A Democratic criticism of some (not all) Republicans’ motivations for such taxes would be exactly the same.

A well-structured tax code strives for neutrality. Digital advertising and social media are already taxed through corporate income and other taxes, and there is little justification for exposing them to additional special taxes exclusively on their activities. Using the tax code for social engineering is, moreover, a dangerous path to go down.

Policymakers already use the tax code to influence business decision-making, often to the detriment of the economy even when their intention is to spur growth. Targeted incentives, while lowering tax costs for select businesses and activities, by their very nature pick winners and losers—and more than that, they drive investment into politically favored activities even when such allocations are inefficient compared to the alternative. No matter how insightful they may be, lawmakers cannot improve upon the allocational efficiency of the market.

This does not mean that markets are everything, or that economic efficiency is all that matters. There is broad political support for a wide range of interventions designed to advance different public policy purposes. But if the goal is economic growth, policymakers designing an incentives program will not be “smarter” than the market mechanisms they are distorting.

Here we have a different type of non-neutrality, one designed as an intentional penalty, either for the offense of making some people too rich or for adopting corporate policies that some lawmakers dislike. Neither approach is healthy; neither is what the tax code is for. The tax code is already replete with ways to tax high-income earners without arbitrarily taxing some people more based on the source of that income, and it is a dangerous path to begin targeting certain industries or businesses because some lawmakers disapprove of them or their policies.

Data taxes have a similar impetus but are much broader. Some, like one proposed in Washington, would be imposed on transactions: on the sale of data. Others, like one introduced in New York, would be imposed on intangible property: on the possession of consumer data. The notion is that companies should have to pay to use information about residents of a state, though earlier notions that residents themselves should benefit from this (through a “data dividend”)[5] seem to have lost favor with legislators.

But such a tax is potentially incredibly broad and almost impossible to administer. Consumer data is not just an online advertising network’s tracking of a person’s online activities. It is also a retailer’s list of customer names and addresses, or a grocery store’s loyalty card database, or really any identifying information a company has on its customers, down to details of their transactions directly with that company. It could even include aggregate data about sales which companies use to adjust their offerings. Assessing the value of such data, moreover, is extremely difficult; in practice, its value depends a great deal on what companies do with it.

Ultimately, such taxes rarely accomplish what they are intended to do. Digital advertising and social media taxes will increase costs for local businesses and consumers. There is no reason to believe that the big technology companies will have to take the hit on their own. A local restaurant advertising on a search engine or social media platform will pay more for its advertisements than it did previously, and that restaurateur is neither a wealthy tech titan nor responsible for whatever policies were adopted by the advertising platform.

Consumers might pay more as well, not just because some businesses might be able to pass along a share of their advertising costs (depending on the degree to which their competitors bear the same costs) but because such taxes could change their own access to online services. If an advertising-supported model is less profitable, then more of the internet might go behind a paywall. If social media companies are taxed both on their revenue-generating ads and per account, then they might restrict account creation or even, in some instances, adopt subscription fees. Even if this doesn’t happen, innovation in the space might be reduced if the model becomes less profitable.

These taxes face many legal and practical hurdles. Consider social media taxes, the newest entrant. If a social media tax is imposed not only on advertising revenues but also on in-state accounts, for instance, will all social media networks now be required to obtain full addresses of their account holders? Would they be required to shut down the existing accounts of anyone who fails to provide a physical address? Or would it be based on IP addresses, which are already imperfect, but would also be complicated by travel or by people splitting time across multiple addresses? If someone regularly checks a social media account at their home in one state and their office in another, is the tax owed in both?

With digital advertising taxes more generally, accurately pinpointing a viewer’s location is likewise a problem, as is the Permanent Internet Tax Freedom Act, a federal law which prohibits discriminating against digital activity in taxation. And defining what constitutes digital advertising is no easy task, either: a banner ad or paid link is obvious enough, but what about underwriting a newsletter or podcast, or placing sponsored content, conducting email marketing, making sales calls using Voice-over-IP phone systems, or rebroadcasting a radio station that includes local ads? Maryland lawmakers and regulators are sorting through that thicket right now, during the delayed implementation of their digital advertising tax.

Finally, data taxes create extraordinary implementational headaches. How should a company value its own consumer data? Some companies might leverage this data quite successfully to market to its customer base, or to design or feature new products; others might do next to nothing with a theoretical wealth of customer data beyond sending out invoices. How do you assess the innate value of that consumer data when it is not being sold (which would give it a price), independent of how it is used? Who would do so? And would people be better off if it became too expensive, or too much of a hassle, for retailers to offer loyalty cards or member perks?

Yet even above and beyond these legal and practical concerns—of which there are many more besides—there is a more fundamental question. That question is, simply, “why?”

Why tax this industry and this activity in particular, given that it is already subject to general taxes? What purpose is served? What cost is offset, which externality internalized? And if first digital advertising is taxed, then social media, then data—what next?

Conclusion

Thus far, Maryland is the only state to adopt any such tax, and its digital advertising tax is already the subject of litigation. Before other states follow in their footsteps, policymakers should ask themselves not only whether their proposal is practical, and not just whether it is legal, but also whether it is justified. They should also think very carefully before using the tax code to target industries or specific businesses over objections to their policies or business practices. Policymakers should consider what they are trying to accomplish with the tax, and whether it will actually advance those aims.

[1] Tony Romm, “U.S. eyes flurry of new taxes on Amazon, Facebook and Google, trying to force tech to pay its ‘fair share,’” The Washington Post, Feb. 12, 2021, https://www.washingtonpost.com/technology/2021/02/12/amazon-facebook-google-maryland-tax/.

[2] Liz Farmer, “Maryland To Delay Controversial Digital Advertising Tax As The Lawsuits Keep Coming,” Forbes, Apr. 17, 2021, https://www.forbes.com/sites/lizfarmer/2021/04/17/maryland-delays-digital-advertising-tax-as-the-lawsuits-keep-coming/.

[3] Jared Walczak, “Worse Than Advertised: The Legal and Economic Pitfalls of Maryland’s Digital Advertising Tax,” Mar. 16, 2020, https://www.taxfoundation.org/maryland-digital-advertising-tax/.

[4] Paul Roemer, “A Tax That Could Fix Big Tech,” The New York Times, May 6, 2019, https://www.nytimes.com/2019/05/06/opinion/tax-facebook-google.html.

[5] See, for example, CBS News, “California governor: Consumers should get paid a ‘digital dividend,’” Feb. 14, 2019, https://www.cbsnews.com/news/california-governor-consumers-should-get-paid-a-digital-dividend/.