Advocacy research vs think tank research

A reader encountering a fifty-page PDF with a methodology section, charts and a citation list has limited ability to tell whether it came from a research institute or a campaign organisation. Both formats are identical by design, because the campaign organisation has correctly identified that the research format confers credibility.

The distinction is real and it is worth being able to make. It is not about honesty, and it is not about quality — advocacy research is frequently more rigorous than lazy think tank work. It is about what the organisation does when the evidence points the wrong way.

The defining test

Ask: what would this organisation do if the analysis contradicted its position?

For a research organisation, the contradictory finding is publishable and potentially its most valuable output. Institutional reputation rests on being right, and a body that only ever confirms its priors is worth nothing to the people who fund it for accuracy.

For an advocacy organisation, the contradictory finding is generally not published. Not suppressed dishonestly — simply not pursued, or not prioritised, because the organisation exists to advance a mission and resources go to work that advances it. The unpublished study is invisible to you, and that invisibility is the whole problem.

This is publication selection, and it means an advocacy organisation can publish only true things and still leave you with a systematically skewed picture.

Six practical tests

1. Does it publish findings against interest? Search the organisation's output for a conclusion that complicates its own position. Research organisations have some; advocacy organisations rarely do.

2. Is the method reproducible? Are data sources, assumptions and calculations stated well enough that someone could redo the work and get the same answer? Reproducibility is expensive and advocacy communications rarely need it.

3. Are uncertainties quantified? Ranges, confidence intervals, sensitivity analysis. Campaign material tends toward single figures because ranges are harder to communicate.

4. Is the counter-argument stated properly? A serious analysis states the strongest version of the opposing case before answering it. Campaign material states a weak version or omits it.

5. Does the conclusion follow from the analysis, or exceed it? The most common tell. Analysis establishes X; conclusion recommends a policy that would require X, Y and Z. The gap is where the mission enters.

6. Is the funding disclosed? Weakest of the six, because disclosure norms vary and non-disclosure is often inertia. Still worth checking, and worth checking on organisations you agree with.

It is a spectrum, not a binary

Every organisation in this atlas sits somewhere on a continuum, and the honest placement is more nuanced than the categories suggest.

The Canadian Climate Institute produces reproducible modelling with stated assumptions, and has an evident policy orientation. The Fraser Institute produces serious economic analysis from openly stated priors. Clean Energy Canada states its objective as acceleration and does real cost work. Environmental Defence commissions genuine laboratory testing in service of campaigns. CAN-Rac aggregates positions and does not claim to research.

None of those descriptions is an accusation. They are descriptions of purpose, and knowing the purpose is what lets you weight the output correctly.

The symmetry requirement

Everything on this page applies identically to industry-funded research. A study commissioned by a producer association establishing that a regulation will cost jobs is advocacy research with the same publication-selection property, and it should be read with the same questions.

The failure mode in practice is that readers apply these tests to organisations they disagree with and skip them for organisations they agree with. That is not scepticism; it is motivated reasoning with a methodology section. The tests are only worth anything if they are applied to your own side first.

The grey zone

The hardest cases are organisations that genuinely do both, and most of the bodies in this atlas are in that position to some degree.

A think tank with a strong institutional viewpoint will produce rigorous, reproducible analysis on questions where the answer is uncertain, and campaign-style communications on questions where it considers the matter settled. Those two outputs appear under the same logo, often in the same week, and the reader is left to sort them.

The practical test is granular rather than organisational: apply the six questions to the document, not to the body that published it. A campaign briefing from a research institute is a campaign briefing. A rigorously reproducible study from an advocacy organisation is a study. Judging at the level of the organisation rather than the document is how readers end up dismissing good evidence and accepting weak evidence, purely on the basis of the letterhead.

What to actually do

Three habits, and they are cheap.

Read advocacy research for the finding, not the conclusion. The lab result, the document obtained, the figure calculated — those are checkable and often valuable. The recommendation is a separate claim.

Assume you are seeing a selected sample of everything the organisation has looked at. Ask what a study finding the opposite would have looked like, and whether you would ever have seen it.

Triangulate on contested questions. If a question matters enough to form a view on, read an advocacy source, a research source and the primary data. Where all three agree you can be confident. Where they diverge, the divergence is the interesting part, and it usually resolves to an assumption rather than a fact.