Tax Policy – Worse Than Advertised: The Legal and Economic Pitfalls of Maryland’s Digital Advertising Tax
- The Maryland General Assembly is on the verge of adopting a vaguely worded, legally dubious tax on digital advertising in the final days of this session—now paired with new tobacco taxes.
- Although lawmakers intend to tax large out-of-state businesses, much of the tax burden would be borne by Maryland-based businesses and Maryland consumers. High taxes on digital advertisements also undercut consumer choice by pushing more content toward subscription models.
- The definition of digital advertising is vague, leaving open to interpretation whether sponsored content, email marketing, or rebroadcasting of content that did not originate online could incur tax liability.
- After concerns were raised about how advertisements would be determined to have been served to Maryland users, lawmakers amended the bill to punt on the question, leaving it up to the state comptroller.
- The tax likely violates federal law by imposing discriminatory taxes on electronic commerce and faces serious challenges under the Dormant Commerce Clause of the U.S. Constitution, as well as other constitutional provisions, and will almost certainly lead to extensive litigation.
- Business activity is already taxed, and justifications for a standalone digital advertising tax are weak.
An old adage counsels against picking a fight with those who buy ink by the barrel—but it has nothing to say about pixels. In Maryland, in a legislative process foreshortened and closed to the public due to the present public health crisis, the legislature is on the verge of adopting a proposal originally conceived as a punitive measure against targeted advertising that quickly became a broader tax on digital advertising. The bill, fast-tracked as the legislature seeks to wrap up its work, raises serious legal and economic questions. It represents the first major foray of a U.S. state into the global debate over the standalone taxation of digital advertising and digital services more broadly, but it is unlikely to be the last.
Maryland’s Senate Bill 2 (and the cross-filed House Bill 965) is concerning in its own right, but also as a harbinger of action in other states. No tax operates in a vacuum, but the taxation of digital advertising is notable for the breadth of its potential impact. It implicates, among other things, the U.S. position in international trade negotiations, the reach of the Permanent Internet Tax Freedom Act (PITFA), the restrictions imposed under the Dormant Commerce Clause of the U.S. Constitution, First Amendment protections, the technical limitations of geotargeting, and even the very definition of advertising.
Although rationalized as a tax on some of the world’s largest technology companies, most of which are headquartered outside Maryland, a significant share of the economic incidence of a digital advertising tax would be borne by in-state businesses in the form of higher advertising costs. Furthermore, some portion of the tax will be absorbed by Maryland consumers, both directly (for instance, in the form of new or higher subscription fees for formerly ad-supported or partially ad-supported digital products) and indirectly (in cases where some of a business’ higher advertising costs can be passed along to consumers).
Were a digital advertising tax enacted in Maryland or elsewhere, legal challenges are certain and likely to succeed, while the economic impact of any tax would be substantially concentrated in the state.
Outline of the Proposal
Maryland lawmakers were looking for revenue, but the original inspiration for S.B. 2 was never about revenue; it was a proposal for the punitive federal taxation of targeted advertising as a response to perceived exploitation of customer data and invasions of privacy, especially when used nefariously to influence U.S. politics or elections. The proposal that caught Maryland lawmakers’ eye, developed by distinguished economist Paul Romer, was not intended to tax all digital advertising, just a subset deemed exploitative. Testifying in Maryland, Romer indicated that he would be satisfied if a tax designed to his specifications raised no revenue at all because it had succeeded in eliminating that class of advertising.
Setting aside the constitutional implications of prohibitive taxation of a particular class of speech, this is not an apt description of the legislation that actually emerged. The proposed tax falls on the annual Maryland gross revenues derived from digital advertising services by all companies with at least $1 million in gross advertising revenues attributable to Maryland and $100 million or more in global gross revenues.
The Maryland taxable revenues would be subject to different rates based on global annual gross revenues for the broader business, not just advertising revenues, let alone just revenues attributable to Maryland. Rates range from 2.5 to 10.0 percent. However, since the largest global digital advertising networks (like Google AdWords or the Facebook or Amazon advertising platforms) are used by everyone from the largest corporations to the smallest “mom-and-pop” stores, the 10 percent rate is likely to apply to a large share of all advertising in Maryland, regardless of the size of the company placing the ad.
Notably, Maryland’s rates are not graduated or progressive, meaning that the higher rates do not just fall on marginal revenue but all revenue. This is inherent in the unusual structure by which all global revenues, and not just Maryland taxable revenues, determine the taxable rate. If a company with $5 billion in global sales has $5 million in Maryland-based advertising, it will face a higher rate than a company with slightly less in global sales but $10 million in Maryland advertising.
|Rate||Global Annual Gross Revenues|
|2.5%||$100 million to $1.0 billion|
|5.0%||$1.0 billion to $5.0 billion|
|7.5%||$5.0 billion to $15.0 billion|
|10.0%||More than $15.0 billion|
Source: Maryland Digital Advertising Tax Rate Schedule Under S.B. 2 (2020)
Under the bill as introduced, digital advertising services were deemed to be provided in the state of Maryland, and thus taxable, if they were served on the device of a user with an Internet Protocol Address (IP) indicating that user’s device is located in the state, or “who is known or reasonably suspected to be using the device in the State.” Perhaps recognizing the shortcomings of this approach, lawmakers adopted substitute language which fails to address the complexities of geofencing advertising activity (that is, setting or determining the location where the advertising is served to users), but rather punts on the question, leaving the design and implementation up to the state comptroller. What constitutes digital advertising may, if anything, be even less clear, though it would certainly include common forms of digital advertising like banner ads.
In a last-minute effort to broaden the bill’s appeal, the digital advertising tax proposal bas been combined with new tobacco taxes, but this paper focuses exclusively on digital advertising taxation.
Seven Issues with Maryland’s Proposed Digital Advertising Tax
With its digital advertising tax, Maryland is attempting something new in state taxation. While a few states include certain digital goods and services in their sales tax base, at present no state imposes a standalone tax on digital services, which is considerably more complex. Somewhat analogous proposals in New York and West Virginia have made no headway, but Maryland’s proposal is receiving serious consideration.
The potential pitfalls of Maryland’s proposed digital advertising tax are numerous, and the novelty of the proposal means that tax policy experts are still grappling with its implications. At least seven points, however, bear making.
1. Much of the tax will be borne by in-state companies and individuals.
The bill’s revenue thresholds are designed to target large, out-of-state companies (which raises legal issues), but legal and economic incidence are not the same thing. The tax is imposed on revenue from digital advertising served to people in Maryland, which will drive up the cost of advertising to Marylanders. The advertisers will bear most of the additional cost, and many of those advertising to Maryland residents are going to be Maryland-based businesses. The revenue thresholds may be designed around companies like Google and Facebook, but Maryland businesses—including many small businesses—are the ones actually paying.
Because the tax is constituted as a gross receipts tax, moreover, it is not based on ability to pay. Nor are the rates dependent upon the size or revenues of the business purchasing advertising and thus likely bearing the burden of the tax. Instead, rates are dependent upon the size of the advertising network, meaning that a large multinational business advertising on a local news site may face no tax, while a local restaurant purchasing keyword advertising on a prominent search engine would bear the costs of a tax imposed at the 10 percent maximum rate.
Industry research suggests that advertising generates about 16 percent of all U.S. sales. Such estimates are innately difficult even setting aside the incentive structure behind such findings, but lawmakers need not rely on industry-sponsored research to recognize that advertising drives sales, and that their local businesses would not be going to that time and expense were it not so.
A study of the French digital advertising tax concluded that 55 percent of these advertising costs would ultimately be passed along to end consumers, which, if applied to the Maryland tax, would result in more than $135 million of the revenue coming from Maryland consumers. The French tax and Maryland’s proposed tax are not identical, but lawmakers must accept that, whichever out-of-state businesses may be the intended targets of the tax, much of the economic burden will fall on Maryland businesses and consumers.
2. The proposal will put more of the internet behind a paywall.
The world wide web, in its infancy, was termed the “information superhighway.” It is easy to forget, now that the internet is so firmly embedded in our daily lives, just how much of an information revolution the internet represented, putting vast knowledge at our fingertips. Idealists proclaim that information wants to be free, and dream of an internet where all content is available to all people, but, necessarily, given that much of the information and content online is produced by companies and individuals with profit motives, vast swaths of online content is only available through subscription services or behind paywalls.
That more of the internet is not walled off, however, owes to the other significant means of monetizing online content: advertising. When advertising becomes more expensive, and the gains from advertising commensurately smaller, more content must be subscription-based to become viable.
There is a place for both ad-supported and subscription content online, and probably always will be. A substantial tax on digital advertising, however, favors the subscription model over ad-supported models, to the detriment of many users whose revealed preferences clearly favor free (or reduced-price) ad-supported content and services. Consequently, a digital advertising tax leads to a decline in consumer choice and likely a reduction in overall digital content.
3. The bill does not clearly define what constitutes advertising.
At first blush, defining digital advertising does not seem particularly daunting. In the bill, the definition of digital advertising services “includes advertisement services on a digital interface, including banner advertising, search engine advertising, interstitial advertising, and other comparable advertising services,” where a digital interface is defined as “any type of software, including a website, part of a website, or application, that a user is able to access.” (Interstitial advertising is advertising served within or between videos, or which briefly covers the full screen of an app or website.)
There are, clearly, the relatively easy cases, like a banner ad served to a user with an IP address in Central Maryland whose linked account shows a Maryland mailing address. In other cases, however, determining what constitutes advertising is considerably more complex. Sponsored and branded content, email marketing, and even Voice-over-IP (VoIP) sales calls could all theoretically constitute digital advertising, with obvious complexities in assigning value or allocating them to Maryland users.
An advertisement or promoted content on a microblogging platform like Twitter or Instagram plainly meets the definition for purposes of S.B. 2, but what of sponsored content on a non-corporate account, such as when a company pays a prominent account to promote their product? Interstitial advertising on a video platform or audio streaming site is relatively straightforward, but is it meaningfully distinguishable from a podcast with a corporate sponsor? Or consider product placement in digital content streamed to Maryland users. Either Maryland’s tax will have significant loopholes which discriminate against more traditional forms of advertising, or it will be an administrative nightmare which reaches almost every form of online content.
No consideration, to date, has been given to whether the tax would apply to advertising that originated in an untaxed medium but was later transmitted digitally. There are, for instance, subscription services which digitally transmit terrestrial (over the airwaves) radio and television broadcasting with the original advertising intact. These services were not a party to these ad contracts, and the companies involved did not directly enter into an agreement to serve their advertisements digitally. The original advertisements, broadcast over the airwaves or through cable or satellite transmission, would not be subject to the proposed tax. Do these advertisements become taxable when simulcast online, and who would be liable?
The potential breadth and uncertainty of S.B. 2’s definitions of digital advertising leaves much to be desired. These concerns only grow in magnitude when combined with the bill’s ambiguity on how it will be determined that advertising revenue is associated with Maryland.
4. The bill punts on important questions about how the tax would be imposed.
As introduced, the legislation would have taxed the revenue generated from ads served to individuals who have Maryland IP addresses or are reasonably believed to be in Maryland, but the imprecision of geofencing gave rise to both legal and practical objections, since the use of Virtual Private Networks, ambiguity surrounding IP addresses near state lines, and the potential for people known to live or work using a device outside the state all create scenarios where ads are improperly determined to have been served into Maryland. Responding to these concerns, lawmakers adopted alternative language which ostensibly implements an apportionment formula, but really just punts the all-important geolocation decisions to the state comptroller’s office. It also fails to answer practical questions like whether to count advertisements which were intended for a Maryland user but were prevented from being served by an ad blocker.
Formulary apportionment is a regular feature of corporate taxation, where it makes sense. It makes no sense whatsoever here. Take traditional corporate income tax apportionment, and the problem it is meant to solve. A business may have economic activity in several states, but its profits aren’t clearly allocated among them. It’s necessary, based on factors like where the business’ workforce, investments, and transactions (or, in apportionment terminology, the payroll, property, and sales factors) take place, to divvy profits across states.
Here, though, the tax base is gross revenue from advertising into Maryland. This is not an apportionment problem, where a factor like the location of payroll is used as a proxy for determining how to allocate the tax base (net income); it is a definitional problem, focused on determining whether advertising is really being displayed on Maryland devices. Consequently, the apportionment formula devised here is an exercise in circular reasoning. The apportionment factor used (gross advertising revenue) is identical with the tax base.
The apportionment formula adopted in the S.B. 2 substitute simply takes the percentage of U.S. gross revenue that is attributable to Maryland, and then multiplies it by all U.S. gross revenue. If we set x to gross advertising revenue in Maryland and y to gross advertising revenue nationwide, the formula is (x/y) * y, which is to say: x. We end exactly where we started, with the taxation of gross advertising revenue in Maryland. And, in lieu of a definition that could pass legal muster, the legislation simply instructs the state comptroller to come up with one.
5. The tax likely violates a federal law, the Permanent Internet Tax Freedom Act.
The Internet Tax Freedom Act, enacted in 1998 and, after several extensions, expanded and made permanent by the Permanent Internet Tax Freedom Act (PITFA) in 2016, prohibits discriminatory taxes on electronic commerce. The federal law does not ban state taxes on transactions just because they take place online, but it does require that taxes not single out e-commerce. A tax on all advertising, including digital advertising, would not offend federal law, but a tax which only targets digital advertising, without being imposed on similar offline activity, runs up against PITFA.
The Maryland attorney general’s office has acknowledged the seeming conflict between S.B. 2 and PITFA. The only comfort the attorney general’s office could offer to proponents of the bill was the possibility, however slight, that a court would choose to save the legislation by concluding that the tax in S.B. 2 was not really on digital advertising, but on contracts for digital advertising—contracts that could be entered into online or through other means.
This, however, is a weak argument, and the attorney general’s office made no attempt to hide that fact. A letter responding to a legislator’s inquiry noted the likely conflict, noting only that “[i]t is possible, however,” that a court might adopt a saving interpretation. However, a court previously invalidated a law requiring tax collections by businesses entering into contracts for online marketing even if the contracts themselves were not conducted online, since the taxable activity was electronic in nature. It is difficult to imagine that the courts would draw a different conclusion here.
6. The tax faces serious constitutional challenges.
The U.S. Constitution prohibits states from imposing taxes which discriminate against interstate commerce and requires that taxes on multistate businesses be reasonably related to their in-state activity. States have broad latitude to design their own tax codes, but that authority is not unlimited—and S.B. 2 tests those limits in multiple ways.
Under Dormant Commerce Clause analysis, states may not enact legislation “inimical to the national commerce” or which “impede the flow of interstate commerce.” Outright discrimination against interstate commerce is forbidden, as are undue burdens on such commerce. Here, the tax’s rate schedule, which is based on global annual gross revenues, uses extraterritorial receipts to discriminate against interstate commerce. The tax rate is higher, not based on greater activity in Maryland, but based on worldwide revenues that have nothing to do with Maryland. The U.S. Supreme Court has held such discrimination unconstitutional in the past.
The $100 million threshold before the tax applies may also discriminate against interstate commerce, even apart from the inclusion of non-Maryland revenue in the calculation. Taxes (and other laws) can violate the Commerce Clause if they have the effect of discriminating against foreign commerce even if they are not written to do so expressly, so if a threshold is designed in such a way as to only apply to out of state firms, the courts may strike it down.
The courts have also held that states can only tax “that portion of the revenues from the interstate activity which reasonably reflects the in-state component of the activity being taxed,” which serves as a working definition for what is known as the “external consistency test”—that taxable activity must be apportioned to states in some reasonable and defensible manner based on the company’s activity. Here, the choice of tax base (gross revenues from advertising in Maryland) appears to meet the test, but only so long as the courts conclude that the mechanism chosen for determining when an advertisement is served to someone in Maryland is reasonable and accurate. If IP addresses or some other proxy chosen by the comptroller’s office does not reasonably reflect in-state activity, or would likely lead to double taxation were other states to adopt similar regimes (which would also violate what is known as the “internal consistency test”), the courts could nullify the tax on those grounds as well.
Furthermore, since the United States has threatened retaliatory tariffs over the adoption of digital services taxes in other countries, and these taxes constitute a major element of ongoing trade negotiations as they continue to crop up in Europe, Maryland’s proposal for a similar tax undercuts American foreign policy and may prevent the U.S. from “speaking with one voice” in regulating foreign trade. The U.S. Supreme Court has ruled that “a state tax on the instrumentalities of foreign commerce must not prevent the Federal government from speaking with one voice when regulating commercial relations with foreign governments,” raising the possibility that S.B. 2 could conflict with Congress’ power to regulate commerce with foreign nations.
This list, moreover, is far from exhaustive. The taxation of some forms of advertising but not others—either by design, due to the focus on digital advertising, or due to structural limitations as to which forms of digital advertising can be taxed—raises Equal Protection and even First Amendment questions. The tax would, for instance, apply to online marketplaces (which now also have remote sales tax collections and remittance obligations) but not to the online storefronts of single sellers, with seemingly little or no justification. If Maryland lawmakers adopted a digital advertising tax, they would be inviting a lengthy constitutional fight—or fights.
In weighing these legal risks, lawmakers can take little solace from a lengthy letter from the Attorney General’s office responding to legislators’ questions about the legality of the proposal. The office’s job is to defend lawmakers’ aims in court, and its official standard in evaluating the constitutionality of proposed legislation is to decide not whether it is likely constitutional, but only whether it is definitely unconstitutional—a “not clearly unconstitutional” standard that has been misunderstood, by some, as an affirmation of constitutionality which it by no means represents. Throughout the letter, the Attorney General’s office acknowledges significant concerns under both the PITFA and the U.S. Constitution and casts strong doubt on the legality and constitutionality of the proposed legislation.
7. The argument for taxing digital advertising is weak.
In Europe, digital advertising and digital services taxes have been justified as attempts to tax substantial economic activity that is otherwise out of countries’ reach. In the United States, however, with formulary apportionment for corporate income, the rationale for a separate tax for digital advertising is far weaker. States can already tax companies based on their location or the location of their sales or both, making a tax on one discrete business activity—advertising, and not all advertising, but digital advertising—seem superfluous at best.
A digital advertising tax is nontransparent and nonneutral. Those bearing the actual costs (advertisers, many of them local, as well as in-state consumers) will have those tax costs hidden from them, passed on in the form of higher prices while, on the margin, depressing in-state wages and investment. The tax would influence business decision-making, encouraging companies to reduce overall advertising or to shift into other forms of marketing, either analog or structured in such a way as to avoid the digital advertising tax.
Some have justified a tax on digital advertising as a punitive measure to discourage targeted advertising, or as recompense for using Marylanders’ data. But even setting aside First Amendment concerns over such targeting, the Maryland tax does not distinguish between ads specifically targeted at Maryland residents using their data and ads served to all (or a broad swath of) users, including Marylanders. Ultimately, a digital advertising tax is a tax on a business input within a broader tax regime that already taxes the resulting business profits. It is widely recognized that business inputs should not be taxed, as the result is tax pyramiding, where these costs get embedded in the final price of the good or service being produced.
Although proponents sometimes analogize digital advertising taxes to expanding sales tax obligations to online sellers following the U.S. Supreme Court’s South Dakota v. Wayfair decision, the reality is very different. There, the internet represented a largely untaxed extension of traditionally taxed markets. Here, a traditionally untaxed activity would be subject to taxation just because it occurs digitally.
A state tax on digital advertising would represent a stark departure from traditional forms of state taxation—one which raises a host of serious legal issues. Even the state attorney general’s office, in a 16-page letter, could offer little more assurance than it was not an absolute certainty that Maryland would lose in court. The bill leaves crucial and legally fraught details to the state comptroller and tax administrators, while many vital questions remain unanswered. If lawmakers rush S.B. 2 through in the waning days of the shortened 2020 legislative session, it may take years to sort out the legal and administrative details, leaving the state short of its budgeted revenue, generating uncertainty for businesses, and tying up both businesses and the state in extensive litigation.
 Eversheds Sutherland, “Legal Alert: Maryland State Senate’s Budget and Taxation Committee Hears Testimony on Proposed Digital Advertising Tax,” Jan 30, 2020, https://us.eversheds-sutherland.com/NewsCommentary/Legal-Alerts/228356/Legal-Alert-Maryland-State-Senates-Budget-and-Taxation-Committee-Hears-Testimony-on-Proposed-Digital-Advertising-Tax.
 M.D. Senate Bill 2 (2020).
 Dave Grimaldi, “Opinion: The (Misguided) Push for a Digital Advertising Tax,” AdAge, May 17, 2019, https://adage.com/article/opinion/opinion-misguided-push-digital-advertising-tax/2172521.
 Julien Pellefigue, “The French Digital Service Tax: An Economic Impact Assessment,” Deloitte/Taj, Mar. 22, 2019, https://taj-strategie.fr/content/uploads/2020/03/dst-impact-assessment-march-2019.pdf.
 Maryland S.B. 2 (2020).
 The latter is less likely, as it is not served over a digital interface, but not impossible, as the legislative language is so broad as to “include” but not necessarily be limited to advertising services on a digital interface.
 A plausible conclusion is that the company responsible for its digital transmission generates no gross advertising revenue from the rebroadcast of these advertisements, and therefore there is nothing to tax. If this were the case, however (as it probably should be), it potentially opens the door to creative arrangements by other advertisers and advertising networks intentionally engaged in digital advertising, designed to avoid liability.
 47 U.S.C. § 151.
 Sandra Benson Brantley, “Re: Senate Bill 2 – Digital Advertising Gross Revenues – Taxation,” Office of Counsel to the General Assembly, Attorney General of Maryland, Feb. 7, 2020.
 Performance Mktg. Ass’n, Inc. v. Hamer, 2013 IL 114496 (Ill. 2013).
 Barclays Bank PLC v. Franchise Tax Bd. of Cal., 512 U.S. 298 (1994).
 Oklahoma Tax Comm’n v. Jefferson Lines, 514 U.S. 175 (1995).
 Westinghouse Elec. Corp. v. Tully, 466 US 388 (1984).
 See Hughes v. Oklahoma, 441 U.S. 322 (1979) and Bacchus Imports Ltd. v. Dias, 468 US 363 (1984).
 Ruth Mason, “Maryland’s Proposed Digital Tax May Be Unconstitutional,” Medium.com, Jan. 30, 2020, https://medium.com/@ProfRuthMason/marylands-proposed-digital-tax-may-be-unconstitutional-9be58831315b.
 Goldberg v. Sweet, 488 U.S. 252 (1989).
 Complete Auto Transit, Inc. v. Brady, 430 U.S. 274 (1977).
 Daniel Bunn, “The Davos Digital (Tax) Détente?” Tax Foundation, Jan. 23, 2020, https://taxfoundation.org/davos-digital-tax-detente/.
 In a 2018 statement, U.S. Secretary of the Treasury Steven Mnuchin said, “The U.S. firmly opposes proposals by any country to single out digital companies. Some of these companies are among the greatest contributors to U.S. job creation and economic growth. Imposing new and redundant tax burdens would inhibit growth and ultimately harm workers and consumers. I fully support international cooperation to address broader tax challenges arising from the modern economy and to put the international tax system on a more sustainable footing.” This continues to represent the U.S. approach. Department of the Treasury, “Secretary Mnuchin Statement on OECD’s Digital Economy Taxation Report,” Secretary Statements & Remarks, Mar. 16, 2018.
 Japan Lines Ltd. v. County of Los Angeles, 441 U.S. 434 (1979).
 Ruth Mason, “Maryland’s Proposed Digital Tax May Be Unconstitutional.”
 Jared Walczak and Janelle Cammenga, “State Sales Taxes in the Post-Wayfair Era,” Tax Foundation, Dec. 12, 2019, https://taxfoundation.org/state-remote-sales-tax-collection-wayfair/.
 Sandra Benson Brantley, “Re: Senate Bill 2 – Digital Advertising Gross Revenues – Taxation.”
 Brian Witte, “Tax on Internet Ads Gets Traction in Maryland,” Associated Press, Mar. 14, 2020, https://baltimore.cbslocal.com/2020/03/14/tax-on-internet-ads-gets-traction-in-maryland/.
 Sandra Benson Brantley, “Re: Senate Bill 2 – Digital Advertising Gross Revenues – Taxation.”