by M. Ramesh and Michael Howlett
Policy analysis is often expected to deliver clear, evidence-based recommendations. Yet many contemporary policy problems—from climate change to pandemics—are characterised by uncertainty, complexity, and disagreement about both facts and values. In their recent article, Inferential Reasoning in Policy Analysis: Knowledge Use under Uncertainty and Complexity, M. Ramesh and Michael Howlett examine how policy analysts can generate useful advice under these conditions and argue for a greater role for inferential and abductive reasoning in policy analysis.
The limits of conventional policy analysis
Much traditional policy analysis assumes relatively stable conditions in which analysts can gather reliable evidence, clearly define policy problems, and evaluate options using established tools such as cost–benefit analysis or impact assessment. Ramesh and Howlett suggest that these assumptions often break down in contemporary policy environments.
Many policy challenges combine empirical uncertainty with political disagreement and strong value conflicts. Under such conditions, technical analytical tools may struggle to produce timely or actionable guidance. At the same time, approaches that rely primarily on participatory deliberation can face different challenges, including difficulties integrating empirical evidence into decision-making.
Integrating multiple forms of policy knowledge
To address these limitations, the authors propose a multidimensional framework that integrates several forms of knowledge used in policy reasoning.
Drawing on classical philosophical distinctions, they identify five complementary knowledge types involved in policy analysis: episteme (empirical or factual knowledge), techne (technical expertise and knowledge of policy instruments), phronesis (practical judgment about political and social contexts), sophia (theoretical knowledge about policy processes and systems), and nous (inferential reasoning that integrates these forms of knowledge).
The authors argue that effective policy analysis depends not on privileging any single knowledge type, but on combining them in ways that allow analysts to diagnose policy problems and propose workable responses.
Abductive reasoning and policy learning
The framework centres on abductive reasoning, a form of inference that seeks the most plausible explanation for incomplete observations. Rather than waiting for definitive evidence, analysts develop provisional explanations and policy responses based on available knowledge, and then update their understanding as new information emerges.
In this sense, policy interventions can be treated as working hypotheses. Implementation generates feedback that allows policymakers to refine both their understanding of the problem and the effectiveness of potential solutions through iterative learning.
Making policy reasoning explicit
Ramesh and Howlett suggest that experienced practitioners often use this kind of diagnostic reasoning intuitively. However, because it remains largely implicit, it is rarely taught systematically in policy education or formalised within analytical frameworks.
By articulating inferential reasoning more explicitly, the authors aim to provide a clearer foundation for how analysts might work productively under real-world constraints of limited information, time pressure, and political contestation. While the effectiveness of such approaches remains an empirical question, the framework draws attention to forms of reasoning that policy analysts may already use in practice but that remain underdeveloped in both policy training and analytical frameworks.
You can read the original research in Policy & Politics at
Ramesh, M., & Howlett, M. (2026). Inferential reasoning in policy analysis: knowledge use under uncertainty and complexity. Policy & Politics (published online ahead of print 2026) from https://doi.org/10.1332/03055736Y2026D000000088
If you enjoyed this blog post, you may also be interested in reading:
Ylöstalo, H. (2020). The role of scientific knowledge in dealing with complex policy problems under conditions of uncertainty. Policy & Politics, 48(2), 259-276 from https://doi.org/10.1332/030557319X15707904457648
Policy & Politics Journal Blog 