Why this paper is being considered
The paper, written with Congressional Budget Office staff and Heidi Williams, develops two ways to estimate how federal R&D spending affects the economy: a capital-stock approach and a components approach based on the number and education of researchers. It bears directly on budget scoring of science funding. An evaluation would focus on the return and depreciation assumptions, the treatment of spillovers, and how sensitive budget effects are to the choice of approach.
AI-generated criterion ratings and reasoning
Scores come from AI prioritization; the accompanying explanations include AI-assisted source checks. These are provisional judgments, separate from human ratings and commissioned evaluations.
Decision relevance
8.0/10
The paper informs decisions about federal R&D appropriations, dynamic scoring, and the budgetary treatment of science and innovation policy. Its estimates could affect Congressional decisions around NSF, NIH, DOE, NASA, CHIPS-and-Science-style spending, and R&D-related reconciliation provisions, with broader implications for productivity growth, health innovation, climate/energy technology, and long-run fiscal policy. The global welfare link is indirect rather than GiveWell-style intervention choice, but US federally funded R&D can generate large knowledge spillovers and public goods beyond the United States.
- Paper claim to check An additional federal dollar of nondefense R&D is assessed to increase nonfederal R&D by about 25 cents over the relevant 10-year window. Source abstract
Value of added scrutiny
8.5/10
The method is used institutionally. The CBO has published related working-paper and report material on estimating the economic effects of federal R&D investment, and an independent economics blog has discussed the payoff question. Such use is by the authors' own agency, not an outside adopter.
Timing
9.5/10
Very high timing value: this is a July 2026 working paper in the CBO working paper series and NBER working paper stream, based on a May 2026 NBER conference presentation, and explicitly circulated to stimulate discussion and critical comment. Feedback could still influence CBO's analytic framework for future Congressional cost estimates and macroeconomic assessments of R&D policy.
- Source record Publication-stage and date evidence should be checked in the linked paper record. NBER
Methodological potential
7.0/10
The main challenge is that this is an analytical framework rather than a clean empirical identification paper. Evaluation would need to assess parameter choices, literature mapping, dynamic macro assumptions, treatment of lags, complementarity between federal and private R&D, depreciation of R&D capital, researcher-training channels, and the heavy reliance on a small number of recent estimates such as Fieldhouse and Mertens. Some components may be difficult to replicate or adjudicate because the relevant evidence base is thin and because CBO's model choices are partly institutional and forecasting-oriented rather than purely academic.
- Paper claim to check An additional federal dollar of nondefense R&D is assessed to increase nonfederal R&D by about 25 cents over the relevant 10-year window. Source abstract
Prominence
9.0/10
The scoring model estimated prominence from the paper's venue, authors, institutional setting, and visibility. The model did not supply a criterion-specific explanation. Current public-attention status: Used in the authors' agency; some independent commentary.
- Institutional report use CBO published an analysis of how federal R&D investment affects the economy and budget. CBO staff are authors of the paper, so this is use by the authors' institution. Congressional Budget Office
Likely influence
8.5/10
The scoring model estimated how far the findings could shape later research or decisions, without supplying a criterion-specific explanation. Current public-attention status: Used in the authors' agency; some independent commentary. See public-attention evidence below.
- Institutional report use CBO published an analysis of how federal R&D investment affects the economy and budget. CBO staff are authors of the paper, so this is use by the authors' institution. Congressional Budget Office
What the paper says
Source abstract · Full source abstract recovered from the NBER paper page; HTML entities and whitespace normalized.
In recent years, there has been Congressional interest in changing various policies related to research and development (R&D). Recent legislation has included provisions that would modify federal funding for R&D investments and tax provisions affecting the after-tax price of R&D. The Congressional Budget Office (CBO) has developed analytical frameworks for estimating how changes in R&D investments affect the economy and the federal budget. This paper describes the agency’s current analytical framework for modeling the economic effects of changes in federal funding for R&D, focusing on two distinct approaches that CBO has developed: an R&D capital stock approach and an R&D components approach.
Claims to check
- An additional federal dollar of nondefense R&D is assessed to increase nonfederal R&D by about 25 cents over the relevant 10-year window.
- A policy increasing federal nondefense R&D appropriations by $30 billion per year from 2027 to 2036 is estimated to raise potential GDP under both the R&D capital stock and R&D components frameworks, with larger long-run effects under the capital stock approach.
- Over a 30-year horizon, each federal dollar spent on R&D is estimated to increase present-value GDP by roughly $3.57 if deficit-financed and $3.82 if deficit-neutral, while the total deficit effect depends strongly on financing and interest-rate dynamics.
Methodological or theoretical issues flagged for evaluation
The main challenge is that this is an analytical framework rather than a clean empirical identification paper. Evaluation would need to assess parameter choices, literature mapping, dynamic macro assumptions, treatment of lags, complementarity between federal and private R&D, depreciation of R&D capital, researcher-training channels, and the heavy reliance on a small number of recent estimates such as Fieldhouse and Mertens. Some components may be difficult to replicate or adjudicate because the relevant evidence base is thin and because CBO's model choices are partly institutional and forecasting-oriented rather than purely academic.
Opus 5.5 re-evaluation (experimental)
Read with care. This is an experimental re-evaluation by a different model (Claude Opus 5.5, 2026-10-01). Across a calibration sample Opus scored about 15 points lower than GPT-5.5, so 15 points are added to compare it with GPT-5.5 scores. The calibration is a small sample (n=44) and is not human-validated, and the two models disagree about the order of papers within the top group. The AI score at the top of this page is the usual evaluation-priority score shifted by the difference between the adjusted Opus score and the earlier holistic score; the synthesis uses that shifted value, and the unshifted score is shown beside it. On the dashboard you can switch the Opus scores off. Read the methods note.
Opus rationale (AI-generated)
This NBER working paper by CBO staff documents how the Congressional Budget Office models the macroeconomic and budgetary effects of changes in federal R&D funding, using an R&D capital-stock approach and an R&D components approach. It is squarely in scope (quantitative public economics and innovation policy), and it is unusually decision-proximate: these frameworks feed directly into how Congress sees the costs and returns of R&D appropriations and related legislation. An Unjournal evaluation could add real value by having innovation economists scrutinize the key parameters: assumed returns to federal versus private R&D, crowd-in/crowd-out, depreciation and diffusion lags, the treatment of defense R&D, and how uncertainty is (or isn't) carried through to point estimates. Agency methodology papers rarely get that kind of public, independent critique, and CBO has a track record of updating its methods in response to expert input. Concerns: this is primarily a methods description rather than new causal evidence, so evaluators would assess the calibration choices rather than an identification strategy. The welfare frame is US output and the budget, with no attention to global spillovers or distribution. And the global-welfare stakes are more diffuse than for our core LMIC or catastrophic-risk work. Worth monitoring or considering, especially if we can recruit evaluators with expertise in returns to public R&D (for example, people who have worked on NIH/NSF returns estimates or growth accounting).
Dashboard details and provenance
Full AI dashboard scoring rationale
This seems like a strong Unjournal candidate because it is a prominent, policy-facing NBER/CBO working paper on a live federal budget question: how Congress should score and design large changes in nondefense R&D funding. The work would matter directly for Congressional committees, CBO, OMB, OSTP, NSF, NIH, DOE, and policy funders deciding whether federal R&D spending should be treated as a high-return public investment, and how much fiscal feedback to expect. It is not a standard causal-effect paper; it is a modeling and evidence-synthesis framework with substantial assumptions, but that is exactly where independent evaluation could add value because CBO explicitly presents it as preliminary developmental work meant to elicit external review, and the stakes are large relative to the amount of independent public scrutiny so far.
AI decision-relevance rationale
The paper informs decisions about federal R&D appropriations, dynamic scoring, and the budgetary treatment of science and innovation policy. Its estimates could affect Congressional decisions around NSF, NIH, DOE, NASA, CHIPS-and-Science-style spending, and R&D-related reconciliation provisions, with broader implications for productivity growth, health innovation, climate/energy technology, and long-run fiscal policy. The global welfare link is indirect rather than GiveWell-style intervention choice, but US federally funded R&D can generate large knowledge spillovers and public goods beyond the United States.
AI timing assessment
Very high timing value: this is a July 2026 working paper in the CBO working paper series and NBER working paper stream, based on a May 2026 NBER conference presentation, and explicitly circulated to stimulate discussion and critical comment. Feedback could still influence CBO's analytic framework for future Congressional cost estimates and macroeconomic assessments of R&D policy.
Intake, review, and crux connections
Community crux
New Focus Area: Abundance and Growth
· 72% match
Directly estimates economic effects of federally funded R&D, central to whether innovation grants can clear funding bars.
Community crux
Tax Cuts and Innovation
· 62% match
Evidence on federal R&D returns informs public versus privately induced R&D allocation comparisons.
Community crux
New Focus Area: Abundance and Growth
· 50% match
Quantifies US innovation-policy effects relevant to whether growth and innovation funding is competitive with EA alternatives.
Public attention and use
The method is used institutionally. The CBO has published related working-paper and report material on estimating the economic effects of federal R&D investment, and an independent economics blog has discussed the payoff question. Such use is by the authors' own agency, not an outside adopter.
Institutional report use direct institutional use
CBO published an analysis of how federal R&D investment affects the economy and budget. CBO staff are authors of the paper, so this is use by the authors' institution.
Source: Congressional Budget Office · Relationship: coauthor institution
Working paper direct project context
CBO issued the paper as a working paper.
Source: Congressional Budget Office · Relationship: coauthor institution
Independent blog visible circulation
An economics blog discusses the payoff of federal R&D in light of this work. We did not check whether it engages the methods in depth.
Source: Conversable Economist · Relationship: independent media
What the search did not establish
- No exact-title EA Forum or LessWrong discussion surfaced in the targeted search.
- No outside methodological critique of the two approaches surfaced.
- The most decision-relevant next signal is whether Congress or other scorekeepers adopt the approaches or the research community contests the assumed returns.
Targeted search checked 2026-10-01. Search scope: Targeted exact-title searches across the open web, institutional and author pages, news and public social-media results, EA Forum/LessWrong, and policy/white-paper contexts. Evidence records distinguish commissioning or report use from independent discussion, media attention, indexing, and post-publication policy use. A search miss is reported as uncertainty, not proof of absence. Entries with check_depth "quick" rest on roughly one to two searches and should be read as especially uncertain; entries checked on 2026-10-01 were added for the shortlist of papers without human ratings.
Human feedback so far
No human ratings have been submitted for this paper yet, so there is no human aggregate or synthesis score. If you know the work, rate or discuss it on the dashboard card; the team reviews feedback before it affects prioritization.