Who Cuts Innovation—and Why
The U.S. R&D capitalization rule applied to a common federal tax system, but it did not create a common economic burden. This chapter explains why the Stanford study finds the clearest R&D reductions among domestic-only, research-intensive, and financially constrained firms. The distinction is useful because it turns a seemingly mixed average result into a test of mechanism: firms cut more when more of their research is exposed or when they have less ability to bridge a near-term cash-tax demand.
The relevant contrast is not between firms that care about innovation and firms that do not. A company can regard R&D as strategically essential and still face a financing constraint. Nor is “U.S. firm” a sufficient description of exposure. A U.S.-headquartered multinational may conduct some research abroad, while a domestic-only company has fewer ways for foreign operations or foreign tax treatment to alter the immediate scope of the U.S. change.
Exposure is a property of the R&D program, not just the company
The Stanford researchers define domestic-only firms using pre-policy disclosures of foreign income and foreign tax expense. For the firms in this comparison, the distinction is a practical measure of how much R&D the U.S. capitalization rule reaches. If a company operates entirely in the United States, the tax timing change applies to all of the R&D spending considered in the study. If a multinational conducts research abroad, the local tax treatment in the country where that research occurs is not changed by the U.S. rule.
The paper also notes a separate 15-year amortization treatment for foreign R&D in certain calculations involving foreign income for U.S. tax purposes, including GILTI, Subpart F, and foreign tax credits. That rule affects how foreign R&D enters particular U.S. tax computations; it does not mean that foreign operations simply escape tax consequences, and it does not change the foreign country’s own R&D tax treatment. The important point is narrower: a multinational’s R&D portfolio can contain components exposed to different timing rules.
This difference produces a testable prediction. If the delayed deduction is causing the response, domestic-only firms should reduce R&D more than multinational firms, relative to their respective comparison groups, because the tax shock reaches a larger share of their research spending. The study estimates a 2.7%-of-assets decline for affected domestic-only firms. Based on an average asset level of $488 million, the estimate is about $13.2 million per firm in the first year. Across 176 affected domestic-only firms in the sample, it amounts to approximately $2.3 billion.
The percentage is scaled by assets because firms differ greatly in size. It should not be read as 2.7% of revenue or as 2.7% of the tax bill. The paper also translates it into a decline of about 10.6% relative to the domestic-only group’s average R&D investment of 25.5% of assets. Stating both denominators matters.