The following policy brief is a summary of the commentary titled “Credible Climate Planning Under Uncertainty: Beyond Incredible Certitude” by Charles Manski and has been prepared with the author’s kind permission.
Climate policy debate in the United States has become deeply dysfunctional. The exchange between mainstream and contrarian scientific perspectives has hardened into something resembling a religious war, ideologically charged, politically entangled, and resistant to productive communication. Both camps assert their conclusions with a certainty that the underlying science cannot honestly support, producing what the paper calls “incredible certitude”: firm predictions resting on unsubstantiated assumptions.
This discord takes four recognizable forms: a) Conventional certitude consists of predictions accepted by consensus without decisive evidential foundation; b) Dueling certitudes arise when opposing camps make contradictory predictions without ever engaging the substance of the disagreement; c) Conflating science and advocacy occurs when researchers design analysis to confirm a predetermined conclusion; d) Wishful extrapolation applies untenable assumptions to generate forecasts that are particularly damaging in long-run climate projections involving tipping points with no historical analogues. All four are visible in the current American climate discourse, and the paper argues they reflect a failure of analytical discipline, not an inevitable feature of a hard scientific problem.
Two Reports, Two Failures
1. The 2025 NASEM Report
The National Academies of Sciences, Engineering, and Medicine released a self-commissioned fast-track report in September 2025 responding to the EPA’s proposed rescission of its 2009 Endangerment Finding which is the legal foundation for federal climate regulations. Drawing on mainstream climate science, its sixteen authors concluded that “EPA’s 2009 finding that the human-caused emissions of greenhouse gases threaten human health and welfare was accurate, has stood the test of time, and is now reinforced by even stronger evidence”.
The report’s findings are scientifically defensible as far as they go, but structurally incomplete as a basis for policy. The committee was asked only a binary question—does climate change endanger public health or not?—and its conclusions are correspondingly qualitative throughout. The report never quantifies the magnitude of damages, and it contains no discussion of abatement costs whatsoever. Without weighing costs against damages, it cannot provide coherent guidance on how aggressive regulations should be, how fast to act, or how to trade off present burdens against future harms. References to uncertainty are pervasive but almost always binary and non-quantitative. The report also ignored the contrarian literature entirely, citing none of the research produced by the authors of the competing DOE report and engaging with it only in passing. In a discourse already defined by mutual incomprehension, that silence deepens the impasse.
2. The 2025 DOE Report
The Department of Energy’s July 2025 report, authored by five scientists with established contrarian reputations, reached starkly different conclusions. Acknowledging that warming is occurring, it disputed its severity and the appropriateness of aggressive mitigation, arguing that CO₂-induced warming is likely less economically damaging than mainstream research suggests and that “excessively aggressive mitigation policies could prove more detrimental than beneficial”.
Unlike the NASEM report, the DOE report explicitly discusses abatement costs and the conditions under which aggressive policy would or would not be economically justified. It devotes a full chapter to uncertainties in climate attribution and engages quantitatively with IPCC-sourced uncertainty ranges for equilibrium climate sensitivity (ECS), the key parameter governing how much warming results from a doubling of atmospheric CO₂. These are genuine analytic virtues. Yet the report ultimately undermines its own balanced framing: the same concluding chapter that calls for nuanced, evidence-based policy opens with the categorical claim that “U.S. policy actions are expected to have undetectably small direct impacts on the global climate,” a sweeping anti-regulatory assertion offered without the uncertainty accounting the rest of the report advocates. The report accuses mainstream science of conflating science with advocacy, while exhibiting precisely that tendency itself.
A Constructive Path: Welfare Economics
A constructive proposal is to apply welfare economics as a disciplining framework that can raise the quality of policy analysis on both sides. Familiar to non-specialists through the language of benefit-cost analysis, the approach formalizes society’s objectives in a social welfare function and evaluates alternative policies by their welfare consequences. Applied to climate, this requires three quantified components: the physical relationship between GHG emissions and temperature increases; a damage function translating warming into economic and social harms; and an abatement cost function expressing the cost of reducing emissions from a baseline trajectory.
This structure, embodied in Integrated Assessment (IA) models such as Nordhaus’s DICE model, makes the fundamental tradeoff transparent. Moderate mitigation is cheaper but allows greater warming and greater potential damage; aggressive mitigation reduces potential damage but at higher cost. The NASEM report discusses only damages; the DOE report minimizes damages while emphasizing costs. Both commit the same error from opposite directions, and neither provides the quantitative cost-benefit framework that credible policy evaluation requires.
Decision-Making Under Deep Uncertainty
The most technically substantive contribution concerns uncertainty. Multiple credible climate models yield different forecasts for the same emissions scenario. The standard practice of multi-model ensemble (MME) analysis, i.e. averaging across model outputs, lacks a compelling theoretical foundation, and the IPCC has acknowledged that “no universal, robust method for weighting a multi-model forecast ensemble is available”. In the paper “Addressing Partial Identification in Climate Modeling And Policy Analysis,” by Manski, Sanstad, and DeCanio (2021) the authors propose replacing MME averaging with a deep-uncertainty framework that treats the range of model forecasts as a problem of partial identification, applying the minimax-regret (MMR) decision criterion: choose the policy that minimizes worst-case regret across all plausible scenarios, without assigning subjective probabilities to models or retreating to the extreme conservatism of maximin rules.
This approach extends naturally to economic uncertainties, including the intergenerational discount rate—perhaps the most consequential and contested parameter in climate economics. A low discount rate assigns substantial weight to future generations’ welfare and favors rapid, aggressive emissions reduction; a high rate tilts the calculus toward more modest near-term action. This disagreement is not resolvable by empirical research alone; it reflects genuine normative disagreement about intergenerational equity. The MMR framework handles it by treating the discount rate as an unknown quantity spanning the full range used in the literature, rather than imposing a single contested value.
The Missing Distributional Dimension
Current IA models measure social welfare through present discounted gross world product, a convention that ignores how climate costs and abatement burdens are distributed across the population. Climate change disproportionately harms poorer and more vulnerable communities in ways that aggregate GDP figures entirely conceal. A welfare economics approach that takes equity seriously would specify social welfare functions explicitly sensitive to distributional outcomes—a dimension absent from standard IA models and doubly absent from the qualitative reports examined here.
The Broader Institutional Stakes
The collapse of constructive climate discourse has not remained confined to academic journals. In December 2025, the Federal Judicial Center released a new edition of its Reference Manual on Scientific Evidence containing a chapter on climate science that summarized mainstream research without engaging contrarian interpretations. On January 29, 2026, the Republican Attorneys General of twenty-seven states demanded its immediate withdrawal, alleging conflicts of interest among its authors. The FJC complied on February 6. The episode illustrates how deeply the politicization of climate science has penetrated institutions, including the federal judiciary, that depend on credible scientific guidance for their functioning.
Recommendations
Public-facing climate reports should adopt the welfare economic framework, presenting quantitative estimates of both climate damages and abatement costs simultaneously. Reports that address only one side of the ledger are analytically incomplete regardless of scientific perspective. Uncertainty should be communicated quantitatively, not as a binary qualifier since presenting ranges of outcomes across credible model specifications gives policymakers an honest picture of what science can and cannot say. The minimax-regret framework offers a principled approach to decision-making that is robust across the full range of scientifically credible scenarios and does not require the assignment of subjective probabilities to contested unknowns. Both mainstream and contrarian researchers should engage seriously with what the other camp gets right, rather than ignoring or caricaturing it. And welfare analysis should be extended to incorporate distributional equity, making explicit who bears the costs of both warming and abatement, the most politically and ethically consequential dimension of the problem, and the one most consistently hidden by aggregate metrics.
The ultimate goal is not to dissolve genuine scientific disagreement, which may persist, but to replace incredible certitude with credible analysis: transparent about its assumptions, honest about its uncertainties, and structured to illuminate rather than obscure the tradeoffs that climate policy unavoidably demands.