Demonstrating Research on Climate and International Relations
Show how your work connects climate-related evidence to an international-relations question while keeping causal claims qualified.
A credible demonstration of climate–international relations research rests less on the volume of climate evidence you can cite than on a visible chain of reasoning: an international-relations question that an institution actually faces, the climate-related evidence that bears on it, and an inference whose causal strength is stated rather than assumed. Reviewers of this kind of expertise tend to test both ends of that chain — whether the political question is a real decision with identifiable actors, and whether the physical or environmental evidence is used within the limits its own literature sets.
This article sets out that chain as a working tool, the question–source–inference map, and applies it to a clearly fictional research profile. It then covers the vocabulary of qualified causal claims and the practical checks (contribution, permissions, data) that come before describing such work in a network application. It is guidance for researchers and practitioners weighing participation in an expert network. It is not a vacancy notice, a request for a particular project, a promise of admission or assignment, or an income claim.
Where climate–IR work usually loses coherence
The two bodies of research behind this field carry different evidentiary conventions. Hydrological reconstructions, attribution studies and scenario projections are explicit about uncertainty ranges, baselines and model structure. International relations research is explicit about institutions, bargaining positions, mandate and the counterfactual against which a political change is judged. A reader from either side will probe the half they do not work in.
Presentations that fail tend to fail in recognisable ways:
- The question arrives late. A dataset is described for several paragraphs before the institutional decision it might inform appears, if it appears at all.
- Evidence is used at the wrong scale. A global mean temperature series is offered as context for a dispute about dry-season allocation in one basin, without a step that connects the two scales.
- The mechanism is asserted rather than specified. "Climate stress drives instability" names no pathway (no price channel, no displacement route, no mandate trigger) and therefore cannot be examined.
- The claim outruns the design. A case study of one negotiation is presented as evidence about a class of negotiations.
The map below addresses the first, second and fourth of these directly. The third is a drafting discipline: if you cannot write the mechanism as a sequence of steps, the inference is not yet ready to present.
The question–source–inference map
The map is a four-column worksheet. Each row carries one claim you intend to make about the link between climate-related evidence and an international-relations question.
The first column states the IR question in institutional terms: which body, facing which decision or constraint, over what period. A question framed as "the geopolitics of climate" cannot be answered; a question framed as whether an allocation rule remains operable under a defined flow regime can be.
The second column records the climate-related evidence: the dataset, reconstruction, projection or measurement, with its provenance, its spatial and temporal resolution, and the scenario or baseline it assumes.
The third column states the inference the evidence supports — not the finding you hope for, but the strongest statement the second column will bear.
The fourth column records what the inference does not establish. This is the column that distinguishes a specialist from an advocate, and in practice it is the one reviewers read first.
A fictional worked example
The profile below is invented for illustration. It does not describe any applicant, engagement or client of Policy Expert Network, and its outcomes should not be read as findings about any real basin, commission or government.
Fictional researcher "Dr L." works on regional water cooperation, with training in comparative politics and resource economics. Dr L. wants to present a body of work linking changing precipitation variability to the bargaining dynamics of a shared-river basin commission.
| IR question | Climate-related evidence | Inference the evidence supports | What the inference does not establish |
|---|---|---|---|
| Whether the commission's dry-season allocation rule retains an operating margin under more variable flows | Peer-reviewed hydrological reconstruction for the basin, plus two scenario projections to mid-century under different emissions pathways | Under the stated scenarios, the frequency of years falling below the rule's current threshold rises, and the rule's operating margin narrows | That member states will reopen the rule, or that any particular state will defect. Rules are renegotiated for reasons the hydrology does not capture |
| Whether downstream ministries have begun to frame allocation as a security matter | A coded corpus of ministry statements and commission records for a defined period, triangulated with a small number of background interviews | A framing shift is observable within the documented venues and period, with stated limits on how representative the corpus is | That the framing caused a change in negotiating position, or that it extends to venues and ministries outside the corpus |
| Whether regional adaptation finance reaches the most exposed districts | Project-level finance records matched to district exposure indicators | A gap between recorded allocation and measured exposure can be described for the years and categories covered | That the gap reflects intent. Reporting categories and coverage gaps limit what the match can show |
Two features of the example matter more than its content. First, the first column would still make sense to an official who had never read a climate paper; the second would still make sense to a hydrologist. Second, every row has an honest fourth column. A presentation built this way invites scrutiny of the inference rather than of the presenter's judgement.
Qualified causal claims: a working vocabulary
Causal language in this field is not a matter of hedging for politeness. Different claim types require different evidence, and mixing them is the most common source of overstatement.
| Claim type | What has to be established | Language that fits |
|---|---|---|
| Association | Co-variation in defined data, with the comparison and period stated | "is associated with", "co-varies with over the period covered" |
| Mechanism | A specified pathway, with evidence at more than one link where possible | "is consistent with the pathway in which…", "the proposed route runs through…" |
| Attribution | That a named driver changed the magnitude or probability of an outcome, against a stated counterfactual | "contributed to", "raised the likelihood of… relative to the stated counterfactual" |
| Conditional projection | A scenario, a model and its assumptions, stated together | "under the [named] scenario, the frequency of… rises by…" |
Attribution and projection are not interchangeable, and neither substitutes for a mechanism. A counterfactual left implicit will be supplied by the reader, usually in a form less favourable to your argument than the one you would have written.
Contribution, permissions and data
Before describing work in any application, separate your own contribution from that of co-authors, the wider literature and the institutions you worked within. "Contributed hydrological inputs to a basin assessment" and "led the assessment" are different claims, and only one of them is true.
Institutional permissions are individual matters to verify in your own circumstances. Data-sharing agreements, research ethics approvals, funder communication rules, co-author consent and the preferences of a government or organisational partner can each constrain what may be described publicly, and the constraints differ by jurisdiction and by institution. A researcher employed by or affiliated with an institution should confirm the position with that institution rather than infer it from a colleague's practice.
Citing an assessment, a dataset or a public framework is not a credential. Using a national climate assessment or an international scientific report means you relied on it; it does not mean you contributed to it, and it does not imply the issuing body endorses your work or your conclusions.
The receiving side of this exchange has its own guidance. The UK Government's academic engagement toolkit for policy makers is written for officials working with researchers, and reading it is a reasonable way to see the questions a policy audience brings to an academic presentation: which decision the work informs, over what horizon, and what it cannot settle.
What a network application can and cannot carry
The current application route asks for a name, a contact email, a CV, acknowledgement of the terms and consent to processing. That is the whole form. There is no separate profile builder, no field for regional coverage or availability, and no applicant dashboard where status can be checked; the form forwards the CV to the applications mailbox without email verification. The Applicant Privacy Notice sets out how that material is handled.
Because the CV carries the evidence, it is worth arranging so the map is visible in it. A short research-focus line naming the IR question, the region and the period; a methods-and-data line naming the principal sources and any language or access advantages; and selected outputs with your role stated. Unpublished client material and unnecessary sensitive personal data do not belong in it.
Public pages describe the network's own scope and confirmed material: the About page states what the organisation is and how its published case material was approved. Reading it before applying is a more reliable basis for expectations than inferring a service catalogue from individual articles.
Two questions that tend to follow
What if my climate evidence is qualitative rather than quantitative? The map still applies. Interview evidence with negotiators, archival records of commission sessions or documentary analysis of mandate language can occupy the second column, provided the sampling, access and period are stated and the fourth column records what such material cannot establish — typically representativeness and the absence of a counterfactual.
Can work in progress be presented? Yes, if it is labelled as such and the partial state is specified: which data are complete, which analysis is provisional, and what would change the conclusion. Presenting a provisional inference as settled is a larger problem than presenting it as provisional.
A reasonable next step
Researchers who want to see how participation is framed can read the participation pages, and those who decide to proceed can use the current application route. Neither step constitutes an offer of work or a guarantee of eligibility; suitability is assessed separately, and the terms of any specific project are set out in its own agreement rather than in general guidance.