Digital Advertising Taxes After Romer: Maryland's Real-World Test

Summary

Romer's proposal — a progressive tax on targeted-digital-advertising revenue, designed to push platforms from ads toward subscriptions rather than primarily to raise money — did not stay a thought experiment. In 2021 Maryland enacted the Digital Advertising Gross Revenues Tax, the first US state tax explicitly aimed at digital advertising, after Romer personally testified for the bill. This page does not re-derive Romer's mechanism (see the linked page for the full design and its "if big is bad, tax big" logic); it covers what happened when a version of the idea became law: the statute's actual design (which departs from Romer's in several respects), the revenue it has raised against its own projections, the litigation that has consumed most of its life so far — culminating in a unanimous Fourth Circuit ruling (August 2025) striking down the tax's "pass-through ban" on First Amendment grounds — and the still-open question of whether platform advertising revenue is the kind of thing a rent tax should be reaching at all.

From Op-Ed to Statute

Romer's May 2019 New York Times op-ed proposing a progressive ad-revenue tax reached Maryland's legislature directly. The state Senate President introduced a digital-advertising-tax bill in January 2020, and the Senate Budget and Taxation Committee invited Romer to testify in its favor on 20 February 2020 — his testimony is preserved in the Maryland General Assembly's own committee-testimony archive.[1] Romer told the committee he wanted targeted digital advertising to stop, not merely to be taxed, and that he "would be happy" if a tax built to his specification raised no revenue at all, because that would mean the targeted-ad business model had disappeared.[1][2] He separately praised Maryland's specific effort as an attempt to "claw back for the citizens of Maryland revenue which has been sucked out of their state."[2]

The 2020 bill passed with veto-proof majorities but was vetoed by Governor Larry Hogan; the legislature overrode the veto in February 2021, and the tax took effect 14 March 2021, applying retroactively to tax year 2021.[3][4]

The Statute — Where It Follows Romer, and Where It Doesn't

Maryland's Digital Advertising Gross Revenues Tax (Md. Tax-General Article, Title 7.5) is a graduated tax on gross revenue derived from digital advertising services in Maryland ("banner advertising, search engine advertising, interstitial advertising, and other comparable advertising services"), with audio-only and non-programmatic advertising, and advertising on broadcast or news media entities' own platforms, excluded by later amendment.[3][5]

Rate schedule. The rate that applies to a company's Maryland ad revenue (the "assessable base") is set by the company's global annual gross revenue, not by how much of that revenue is Maryland-sourced:[3][4]

A person must file if their Maryland digital-advertising revenue is at least $1 million annually.[4] The Tax Foundation's analysis flags a structural quirk this table implies: the rate applies to a company's entire Maryland assessable base once it crosses a global-revenue threshold, not just to revenue above that threshold — a "cliff" rather than a marginal-bracket structure, which can make a firm just over a threshold worse off than a slightly smaller competitor and creates an incentive to understate global revenue or restructure below a cliff.[6]

Two departures from Romer's own design are worth naming plainly. First, Maryland's tax reaches "digital advertising services" generally, not specifically targeted advertising the way Romer's proposal was scoped — Romer's tax was meant to exempt non-targeted formats (e.g., a plain banner ad not keyed to a user's data) as part of the subscription-escape design; Maryland's statute left this distinction largely to the Comptroller's regulations rather than writing Romer's targeting-based exemption into the statute itself.[7] Second, Maryland's graduated brackets are keyed to a firm's absolute size, which is directionally Romer's "tax big" logic, but the statute contains no equivalent of Romer's explicit subscription escape hatch or the incentive to split up rather than grow — it is a size-graduated revenue tax on a broader ad-services base, closer in mechanics to a DST than to Romer's own narrower, more surgical design.[3][6] The wiki should not describe Maryland's tax as "Romer's tax enacted" — it is a state legislature's adaptation of Romer's big-is-bad logic, filtered through ordinary state tax-drafting, and the two instruments' incidence and legal exposure need not be identical.

The Litigation — the Dominant Story So Far

No US jurisdiction has operated a tax like this for long enough, or with clean enough revenue data, to generate a settled incidence record. What Maryland's tax has generated instead is five years of litigation on two parallel tracks.

State track: procedure, not merits, decided so far. Comcast, Verizon and other industry plaintiffs sued in Maryland state court in April 2021. In October 2022 an Anne Arundel County circuit judge invalidated the tax, agreeing that it violated the federal Internet Tax Freedom Act (which bars states from singling out electronic commerce for discriminatory taxation) and discriminated against interstate commerce by keying its rate to companies' global revenue.[8] The Maryland Supreme Court reversed in 2023 (per curiam order 9 May 2023, opinion 12 July 2023) — but only on a procedural ground: Comcast and Verizon had not exhausted the state's administrative tax-appeal remedies before going to court. The court's opinion did not reach the constitutional or Internet Tax Freedom Act arguments at all.[9][10] The tax survives in force today because of that procedural ruling, not because any court has yet upheld it on the merits — a distinction several of the professional tax-law summaries covered here are careful to preserve, and one the wiki should preserve too. A separate merits case is still pending in the Maryland Tax Court.[11][12]

Federal track: a First Amendment win against the state, on a narrow but real provision. A parallel federal suit — the U.S. Chamber of Commerce, NetChoice, the Computer & Communications Industry Association and the Internet Association v. Franchot (the Maryland Comptroller) — challenged the tax itself under the Internet Tax Freedom Act, the Commerce Clause and Due Process, and separately challenged a distinct statutory provision: Maryland's "pass-through prohibition," which barred a company subject to the tax from directly passing its cost to Maryland customers through a separate fee, surcharge, or line item.[12][13] In March 2022 the district court held the federal Tax Injunction Act barred it from hearing the challenges to the tax itself (federal courts generally cannot enjoin state tax collection where a state offers its own remedy) but allowed the narrower pass-through claim to proceed, since blocking a speech restriction is not the same as blocking tax collection.[12][13] The district court dismissed even that claim as moot in December 2022; the Chamber appealed.[12]

On 15 August 2025 a unanimous Fourth Circuit panel reversed, holding Maryland's pass-through ban unconstitutional under the First Amendment. The court characterized the provision as one designed to shield state lawmakers from "criticism and political accountability" for the tax's real-world costs rather than to serve any legitimate regulatory purpose — a company could raise its prices to cover the tax, but Maryland's law required, in the reporting characterization several outlets used, that it "do so in silence": no line item, no surcharge, no itemized explanation naming the tax.[14][15][16] The case was remanded to the district court, which entered final judgment on 15 October 2025 permanently enjoining Maryland from enforcing the pass-through ban in any form.[16][17] The core tax itself was not struck down — it remains in force, collectible, and separately contested in the pending Maryland Tax Court merits case.[11][17]

The pass-through ruling is itself an incidence-adjacent finding worth naming. A state legislature does not normally bother banning visible pass-through of a tax whose incidence it expects to fall entirely on the nominally taxed firms; the ban's own existence is circumstantial evidence that Maryland's drafters anticipated — and wanted to suppress public attribution of — cost pass-through to consumers. That is not a measured incidence estimate, and the wiki does not treat it as one; it is a fact about the statute's design that the record supports and that a reader assessing incidence should know.

Revenue: Well Under Projection, and Not Officially Itemized

Maryland's own fiscal estimate for the bill projected up to $250 million per year, largely earmarked for the "Blueprint for Maryland's Future" education funding program.[17][18] Press reporting sourced to the Comptroller's office puts actual collections at roughly $93 million in FY2022 and $82.5 million in FY2023, with a cumulative total of roughly $419 million collected since 2022 by the time of an October 2025 report — averaging closer to $100-120 million per year than the projected $250 million.[7][18] This figure needs an honest caveat: it is drawn from secondary press reporting attributed to the Comptroller's office, not from a Comptroller-published table this page could independently fetch and verify — one account explicitly notes exact year-by-year revenue has not been made public in full, partly because of the litigation.[18] The direction (well under the $250 million projection) is corroborated by every source consulted; the exact annual figures should be re-verified against a primary Comptroller report before being treated as precise.

Is Advertising Revenue Rent? Arguing It Both Ways

This is the question the whole instrument stands or falls on, and the wiki's rent gradient requires carrying both sides honestly rather than assuming platforms' ad revenue behaves like land rent.

The case that it is (at least partly) rent. Prat and Valletti's "Attention Oligopoly" (American Economic Journal: Microeconomics, 2022) models dominant digital platforms as attention brokers — intermediaries with proprietary information about users' preferences who sell targeted ad space to retailers competing for that attention.[19] Their result: when attention brokers become more concentrated, ad prices rise, fewer ads reach potential market entrants, and consumer welfare in the underlying product markets falls — a mechanism in which a platform's scale advantage, not just the quality of its matching technology, is what lets it extract higher ad prices.[19] The paper also warns that competition-policy assessments relying on aggregate platform usage "can be highly biased," implying regulators may be underestimating how much of platform ad revenue reflects bottleneck power rather than service quality.[19] This dovetails with the Furman Review's finding that Google and Facebook dominated an £11.55 billion (2017) UK digital-advertising market, and its call for a dedicated market study into whether the digital-advertising value chain is actually competitive — the same "durable-position" evidence the wiki treats as the empirical precondition for calling a return a rent rather than a transient lead.[20]

The case that a meaningful share is not rent. Targeted advertising funds real services users value at a zero sticker price — search, maps, social connectivity — and part of the revenue platforms earn reflects genuine matching quality: better ad targeting is a real technology that raises advertiser return on spend, not merely an extraction mechanism. The Furman Review itself, despite documenting the concentration above, is explicit that "there is nothing inherently wrong about being a large company or a monopoly" and that platform dominance often "reflect efficiencies and benefits for consumers" — the quasi-rent reading the wiki refuses to assume away.[20] Romer's own framing, discussed on the instrument-comparison page, reinforces this: he motivates his tax as a behavioral and democratic correction — he wants targeted advertising discouraged because of its effect on political discourse and attention markets, not because he has identified and measured a pure rent component to be captured. A tax calibrated to make firms abandon the ad model is, by construction, not calibrated to leave a genuine-service return untouched — it is closer to a Pigouvian correction than to a Georgist rent tax, a point this page inherits directly from the proposal page.[2]

What this means for the "is it rent" question, honestly stated. Prat & Valletti supply real evidence that concentration in attention markets raises ad prices and harms downstream competition — that is evidence for a rent-like component tied specifically to market power, not to the underlying matching/targeting technology. It does not, and does not claim to, decompose platform ad revenue into a rent share and a quasi-rent share the way the wiki can point to for land. No source consulted in this research pass provides that decomposition for advertising revenue specifically. The honest reading is the gradient's own: some of platform ad revenue very likely reflects bottleneck power over attention (the Attention Oligopoly mechanism, and the same durable-position evidence the Furman Review documents), and some reflects a genuine, valuable matching service — and no one has cleanly separated the two.

Honest Limits

• No ex-post incidence study of Maryland's tax itself was found. Unlike the UK DST, which has a real empirical pass-through estimate (Muddasani & Langenmayr on Amazon), this research pass did not locate a comparable peer-reviewed or working-paper measurement of who actually bears Maryland's ad tax. The pass-through-ban litigation is suggestive of the legislature's own expectation, not a measured result, and the wiki does not present it as one.
• Revenue figures rest on press reporting, not a fetched primary Comptroller table. The $93m/$82.5m/~$419m-cumulative figures are consistently reported across multiple outlets citing the Comptroller's office, but this page could not independently verify them against an official, itemized Comptroller revenue report; they should be re-checked before being cited as precise.
• The core constitutional merits case remains unresolved. As of this session's research (mid-2026), the tax survives only because of the Maryland Supreme Court's 2023 procedural ruling and a Tax Injunction Act bar on the federal claims; the Internet Tax Freedom Act and Commerce Clause arguments that a state court once found persuasive on the merits (October 2022) have never been affirmed or rejected by an appellate court. This is a live, moving litigation and any "the tax survives" framing should be dated.
• Maryland's statute is not a clean test of Romer's own design. Because the enacted tax departs from Romer's proposal (broader ad-services base, a cliff-rate structure, no explicit subscription escape hatch), its legal and revenue troubles are evidence about a state ad-revenue tax, not a direct verdict on Romer's specific progressive-schedule mechanism.
• The is-it-rent question is not resolved here and should not be treated as resolved. Prat & Valletti's concentration mechanism and the Furman Review's durability findings are real evidence for a rent-like component; neither source, nor any other found in this pass, quantifies what share of platform ad revenue is rent versus quasi-rent.

See Also

• Romer's Progressive Tax on Digital Advertising — the proposal's own design, incidence logic, and the Acemoglu–Johnson variant, not repeated here
• Taxing Tech Rents — Instrument Comparison — where Romer's ad tax is graded C as a rent instrument against ACE/DBCFT, DSTs, data dividends, and antitrust/DMA dissolution
• Digital Services Taxes and Their Incidence — the sibling revenue-tax instrument and its one rigorous ex-post pass-through estimate (Amazon/UK)
• Digital Services Taxes as Actually Implemented — the UK/France/India/Canada revenue and litigation record this page's Maryland case study parallels
• Platform and Data Rents — the underlying rent-or-quasi-rent diagnosis this page's central question extends
• Furman Review — Unlocking Digital Competition — the official UK diagnosis of digital-advertising-market concentration
• Pigouvian Taxation — the correction-not-capture framing Romer's tax shares with Acemoglu & Johnson
• Geoism — the rent gradient this page's honest-limits section applies

Sources

1. Paul Romer, testimony to the Maryland Senate Budget and Taxation Committee, 20 February 2020, Maryland General Assembly committee-testimony archive. mgaleg.maryland.gov PDF — used for the fact and date of Romer's testimony (A-claim; primary legislative record; the PDF's text could not be machine-extracted this session, so its content is corroborated via [2] rather than quoted directly from the PDF itself).
2. Paul Romer, quoted in Kate Klonick / ProMarket, "If You Think Moderation is Censorship, You've Got a Competition Problem," 15 January 2021. ProMarket — used for the "would be happy if raised no revenue" framing, the progressive marginal-rate design description, and the "claw back for the citizens of Maryland" quote (D-claim; interview/journalistic source quoting Romer directly; fetched this session).
3. Maryland Tax-General Article, Title 7.5 (Digital Advertising Gross Revenues Tax), as summarized in Sales Tax Institute, "Maryland Enacts New Tax on Digital Advertising." Sales Tax Institute — used for the rate schedule, effective date (14 March 2021), $1 million filing threshold, and the assessable-base definition (F/A-claims; professional tax-advisory secondary source; the statute itself was not machine-readable via the sources tried this session — see note below).
4. Maryland Comptroller, "Tax Guidance — Digital Advertising Gross Revenues Tax," official Taxpayer Services knowledge base. services.marylandcomptroller.gov — cited as the page's source_url (official primary guidance); corroborates the rate schedule and filing threshold reported via [3] (F-claim; official state guidance).
5. Description of "digital advertising services" scope (banner, search-engine, interstitial advertising; broadcast/news-media and audio-only/non-programmatic exclusions), as reported in web search synthesis of Maryland Comptroller Technical Bulletin No. 59 and related practitioner commentary — used for the statutory definition of the taxed activity (F-claim; the Technical Bulletin PDF itself could not be machine-extracted this session; content corroborated across multiple independent secondary summaries, not directly quoted).
6. Tax Foundation, "Worse Than Advertised: The Legal and Economic Pitfalls of Maryland's Digital Advertising Tax." Tax Foundation — used for the cliff-rate-structure critique, the Internet Tax Freedom Act and dormant Commerce Clause legal analysis, and the tax-pyramiding critique (D-claim; policy-institute analysis explicitly opposed to the tax, presented as that institute's own argued position, not adopted as settled fact; fetched this session).
7. Search-synthesized reporting on the gap between Romer's "targeted advertising" scope and Maryland's broader enacted definition, and on the ~$419 million cumulative revenue figure attributed to the Comptroller's office — used for the design-departure point and the cumulative revenue figure (B/D-claims; multiple secondary press/practitioner sources, not independently confirmed against a primary Comptroller dataset; flagged in Honest Limits).
8. Grant Thornton, "Maryland tax on digital-ad services still in effect," and RSM US, "Maryland digital advertising tax: Where are we now?" (2023) — used for the October 2022 Anne Arundel County circuit court ruling invalidating the tax on Internet Tax Freedom Act and dormant-Commerce-Clause grounds. RSM (B/A-claims; professional tax-advisory secondary sources; fetched via search synthesis this session).
9. Comptroller of Maryland v. Comcast of California/Maryland, LLC, Maryland Supreme Court (per curiam order 9 May 2023, opinion 12 July 2023). Maryland Courts PDF — cited for the case name and outcome; the PDF's text could not be machine-extracted this session, so the exhaustion-of-remedies holding and the court's non-reaching of the constitutional questions are corroborated via [10] rather than quoted directly from the opinion (A-claim; primary appellate opinion, content via secondary corroboration).
10. Aprio, "Maryland Supreme Court Ruling Keeps Digital Advertising Tax Alive . . . For Now," and Conduit Street (Maryland Association of Counties), coverage of the 2023 ruling — used for the procedural (exhaustion-of-remedies) basis of the reversal and confirmation that the constitutional and Internet Tax Freedom Act claims were not addressed (A-claim; corroborated across independent secondary legal-commentary sources; fetched via search synthesis this session).
11. Conduit Street (Maryland Association of Counties), "Federal Court Issues Final Ruling Blocking 'Pass-Through' Ban in Maryland's Digital Ad Tax," 21 October 2025. Conduit Street — used for the 15 October 2025 final-judgment date, the permanent injunction against the pass-through ban, confirmation that the core tax remains enforceable, the pending Maryland Tax Court merits case, and the $250 million original revenue projection (A/B-claims; specialized state-government-adjacent policy tracker; fetched and read this session).
12. U.S. Chamber of Commerce Litigation Center, "Chamber of Commerce v. Franchot," case summary. U.S. Chamber — used for the February/April 2021 filing and amended-complaint timeline, the March 2022 Tax Injunction Act ruling (barring the claims against the tax itself but allowing the pass-through First Amendment claim to proceed), and the December 2022 mootness dismissal that was later appealed (A-claim; litigant's own case-tracking page, corroborated by [13]; fetched this session).
13. Web search synthesis of Chamber of Commerce v. Franchot, 595 F. Supp. 3d 423 (D. Md. 2022) case commentary (vLex, Davis+Gilbert) — used to corroborate the Tax Injunction Act ruling and the specific claims raised (ITFA, Commerce Clause, Due Process, First Amendment pass-through) (A-claim; corroborating secondary legal sources; fetched via search synthesis this session).
14. Chamber of Commerce of the United States of America v. Comptroller of Maryland, No. 24-1727 (4th Cir., 15 August 2025). Fourth Circuit PDF — cited for the case name, number, and date; the PDF's text could not be machine-extracted this session, so the holding and reasoning are corroborated via [15][16] rather than quoted directly from the opinion (A-claim; primary appellate opinion, content via secondary corroboration).
15. BDO, "Fourth Circuit Finds Maryland's Digital Ad Tax Pass-Through Restriction Unconstitutional." BDO — used for the unanimous-panel holding, the "criticism and political accountability" characterization, and the confirmation that other methods of explaining a price increase remain permitted (A-claim; major accounting/tax advisory firm's legal-update summary; fetched this session).
16. Maryland Matters, "Appeals court rules provision of digital ad tax violates First Amendment protections," 15 August 2025; and NBC Washington / other outlets reporting Circuit Judge Julius Richardson's characterization that firms "must do so in silence." — used for the plain-language characterization of the pass-through ban and the "in silence" framing (A-claim; journalistic sources reporting the same ruling as [14][15]; corroborated via search synthesis, direct article fetch returned HTTP 403 this session).
17. Robert H. Smith School of Business, University of Maryland, "Digital Tax Debacle." rhsmith.umd.edu — used for the $250 million annual projection versus roughly $90 million/year actual collections, the "must do so in silence" quote attributed to Circuit Judge Julius Richardson, and the university's overall assessment that the tax fell well short of its revenue goal (B/D-claims; university business-school news summary, not a primary fiscal document; fetched and read this session).
18. Search-synthesized press reporting (multiple outlets citing the Maryland Comptroller's office) on FY2022 (~$93 million) and FY2023 (~$82.5 million) digital-ad-tax revenue and the ~$419 million cumulative total reported by October 2025 — used for the specific annual figures, flagged in Honest Limits as unverified against a primary Comptroller table (D-claim; press reporting, not independently confirmed this session).
19. Andrea Prat & Tommaso Valletti, "Attention Oligopoly," American Economic Journal: Microeconomics 14(3), August 2022 (working paper version 2019-2020). AEA · working paper — used for the attention-broker model, the finding that concentration among attention brokers raises ad prices and reduces consumer welfare, and the caution against merger assessments based on aggregate usage alone (B/C-claims; peer-reviewed journal article; abstract and secondary summary fetched this session, full text not independently retrieved).
20. Digital Competition Expert Panel (Furman, chair), Unlocking Digital Competition, HM Treasury, March 2019 — used for the £11.55 billion (2017) UK digital-advertising market figure, the digital-advertising market-study recommendation, and the "nothing inherently wrong about being a large company or a monopoly" / efficiencies quote. Full citation and this page's own reading of the report are carried at furman-review-digital-competition; not re-fetched independently this session (A/D-claims via that page's own sourcing).