The current code taxes labor reliably while allowing capital, wealth, perks, and inherited appreciation to escape or defer tax. The Accord taxes progressively by broadening the base, applying higher rates at the top, and pairing broad consumption taxation with monthly rebates and luxury surcharges.
Today's federal corporate tax allows a firm selling $10 billion of product into the US market to book the underlying IP in Ireland or the Cayman Islands, attribute the bulk of the profit to that low-tax jurisdiction via intercompany licensing, and report minimal US taxable income — even though the customers, the sales force, and the demand are all American.
This is not aggressive tax planning at the margin. It is the dominant operating model of the largest multinationals, and it is the largest single driver of the gap between corporate book income and corporate taxable income that the book minimum addresses from the other side. The book minimum closes the deduction-stacking route. Sales-factor apportionment closes the headquarters-shopping route.
Pillar Two of the OECD/G20 Inclusive Framework establishes a 15% global minimum on multinationals — the floor under headquarters arbitrage. Sales-factor apportionment is the complementary US-side instrument: it determines what share of multinational profit is properly American before any rate is applied.
Worldwide consolidated income of a multinational corporate group is apportioned to the US by a single factor: US sales as a share of worldwide sales. The apportioned share is then taxed at the applicable corporate rate (with the book minimum as backstop).
A firm with $50B worldwide income and 40% of its sales in the US has $20B in US-apportioned income, regardless of where the IP, the headquarters, the holding company, or the cost-sharing arrangement is domiciled. Intercompany transfer prices become irrelevant to the apportionment math — they are a within-group accounting artifact that no longer drives the US tax base.
The instrument is "single-factor sales" in state-tax terminology: it abandons the historical three-factor formula (sales × payroll × property) in favor of customers-only. State experience confirms that single-factor sales is administrable, audit-resistant, and constitutionally durable; California, New York, and roughly 30 other states have adopted single-factor sales for corporate apportionment over the last two decades.
The federal version applies to the consolidated worldwide group, not to the US subsidiary in isolation. This is the architectural shift: the US tax base is a share of the group's total income, computed by the share of the group's sales delivered to US customers.
Multinational corporate groups with substantial US sales whose effective US tax rate today falls below the apportioned-and-rated benchmark. The bite falls hardest on the technology, pharmaceutical, and platform sectors that have built their tax structures around offshore IP holding companies.
US-only firms with no foreign operations are unaffected — the apportionment factor is 100% US sales / 100% US sales = 1, so the entire income is already in the US base under any framework. The architecture targets the gap between US sales and US-reported profit, not domestic operations.
Domestic-only firms — the apportionment math reduces to the existing federal corporate tax. Small and mid-cap firms below the multinational threshold. Firms with foreign operations whose foreign-sales share genuinely matches their foreign-income share — sales-factor apportionment leaves their US base unchanged because the formula already reflects the underlying economics.
The protection is structural: sales-factor apportionment penalizes the gap between economic substance (where customers are) and reported profit (where IP is parked). Firms whose substance and reporting align face no incremental tax.
Pending canonical scoring (federal sales-factor apportionment has no current statutory basis; estimates pending JCT/CBO methodology).
State-level evidence is suggestive of scale. California's single-factor adoption in 2011 increased California's corporate base measurably; the federal aggregate effect would be substantially larger because the federal rate is higher and the offshore-IP gap is concentrated in federal-base computations.
The instrument's primary function is base protection rather than rate increase: it ensures that the headline corporate rate and the book minimum apply to the share of multinational income that is economically American. Without apportionment, the rate and the minimum apply to whatever share of income the group's tax planners declare to be US-taxable — which is the failure today.
See tax ladder · fiscal scoring
- Offshore IP holding companies
- Closed by apportionment. The IP can sit anywhere; the apportionment factor is sales destination, not IP location.
- Cost-sharing arrangements with offshore affiliates
- Closed by consolidated-group basis. Within-group cost shares are eliminated in consolidation; the group's worldwide income is the base, not the US subsidiary's standalone income.
- Intercompany royalty stripping
- Closed by transfer-price irrelevance. Royalties paid from US to foreign affiliate are within-group transactions that wash out in consolidation.
- Tax-haven holding-company chains
- Closed by single-factor sales. Holding-company location does not affect the apportionment factor; only end-customer location does.
- Inversions
- Substantially neutralized. After an inversion, the US sales share and the US-apportioned income are unchanged — the architecture taxes the same share of the same group income regardless of where the parent is incorporated.
The architectural innovation is shifting the US tax base from a transfer-pricing question (what is the arm's-length royalty for the IP held in Ireland?) to an apportionment question (what share of the group's customers are American?). The latter is observable from third-party data; the former is a pricing argument with hundreds of contestable inputs.
- Corporate book minimum
- Backstop on the apportioned base. The 15% book minimum applies to the US-apportioned share of consolidated book income — closing the deduction-stacking route on the apportioned base.
- OECD Pillar Two (15% global minimum)
- Pillar Two is the headquarters-side floor; sales-factor apportionment is the destination-side complement. They are not substitutes — Pillar Two prevents headquarters from racing to zero; apportionment prevents the US base from being relocated by intercompany pricing.
- Alliance Incentive
- Reciprocal market-access pricing for jurisdictions adopting either Pillar Two or comparable apportionment. The Accord prefers symmetric architectures across allied markets.
- Comprehensive withholding
- Same architectural principle applied to compensation: substance over form. Compensation is taxed where work is performed; corporate profit is taxed where customers are served.
Sales-factor apportionment is the headquarters-side complement to the corporate book minimum. The book minimum closes the deduction-stacking route within US-reported income; sales-factor apportionment determines what counts as US-reported income in the first place. Together they form the architectural answer to multinational corporate base erosion.
Single-factor sales apportionment is a state-level instrument with constitutional authority anchored in Commerce Clause cases. Extending it to federal corporate tax is novel and may invite challenge under the Foreign Commerce Clause, treaty-override doctrine, or WTO/OECD coordination commitments. Trading partners may retaliate.
The constitutional path is well-mapped. The federal government has plenary authority over its own corporate tax base; the state-level Commerce Clause constraints (which require apportionment to reflect activity in the taxing state) are weaker at the federal level because the federal taxing power is plenary. Treaty-override is a Congressional prerogative used routinely; the relevant treaties contain savings clauses preserving the underlying right to tax.
The OECD coordination point is the substantive one. Sales-factor apportionment as the primary federal apportionment instrument should be coordinated with the OECD Pillar One framework, which is itself a partial sales-factor apportionment for the largest multinationals. The Accord's position: Pillar One is the floor; the Accord's apportionment is the framework's full implementation domestically. Allied jurisdictions adopting comparable architectures receive Alliance Incentive market-access reciprocity.
Retaliation risk is real but bounded. The largest target firms are themselves dependent on US market access; their retaliation lever is limited. The architecture's negotiating posture is symmetry: any jurisdiction adopting comparable apportionment receives reciprocal treatment.
Honesty about gaps. The Accord's credibility comes partly from explicit acknowledgment of what is not yet specified. The items below are flagged for v10.2 specification or for outside expert review.
- Multinational threshold: the consolidated-revenue floor above which sales-factor apportionment applies. State practice ranges $25–50M; the federal threshold should be substantially higher to focus on multinational groups (likely $500M–$1B consolidated).
- Sales-sourcing rules for digital services and platform revenue: where is the 'customer' located when a US user accesses an ad-supported service hosted offshore? State-level guidance varies; federal rule pending.
- Pillar One coordination: the OECD framework already implements partial sales-factor apportionment for the largest ~100 multinationals. The Accord's relationship to Pillar One — full implementation, complement, or substitute — needs explicit specification.