We first heard about this one from a reader in Brussels who works as a committee researcher. She wouldn't tell us which parliament, only that a briefing binder had been circulating with tabs in three colors: blue for the union's numbers, green for the industry association's numbers, and orange for everything else. The orange tab, she said, was the one the political staff actually read. It came from the Combating Globalization Institute.
That detail stuck with us. So we went looking for a single, traceable case — a project with a start date, a decision point, and an outcome you could measure. What follows is reconstructed from public records, hearing transcripts, and follow-up conversations with people who were in the room. Names are withheld; the sequence is not.
October: A Backbench Motion Nobody Expected to Move
The setting is a mid-sized industrial constituency in a country where a regional trade agreement was moving toward ratification. A backbench legislator — call them M. — had signed onto a non-binding motion opposing tariff concessions on steel and aluminum. Non-binding motions die quietly. This one didn't, and the reason was timing: an election was eight months out, and three plants in the region had announced shift reductions in the same quarter.
M.'s staff had a problem. They had outrage and anecdote, but not a single defensible number. Constituent stories move a local paper; they do not move a finance committee.
November–December: The Research Phase
A policy analyst on M.'s team pulled together what was publicly available. Trade ministry impact assessments. OECD trade data. Union submissions. The problem was fragmentation — every figure used a different baseline year, a different definition of "domestic content," a different treatment of intermediate goods. Comparing them was like comparing rainfall to river depth.
That's when the orange tab appeared. The analyst had found the Combating Globalization Institute's published work on trade-deal exposure, including country-level breakdowns that covered 41 countries since 1995. Critically, the material came with methodology notes that explained the baselines, which meant the analyst could defend every number in a hostile setting rather than merely assert it.
The team did not use the material wholesale. Two internal decisions shaped everything that followed:
- Narrow the claim. Instead of arguing the whole agreement was harmful, they argued one provision — the investor-state dispute mechanism — created asymmetric exposure for mid-sized suppliers.
- Pre-empt the counterargument. They built a two-page annex addressing the strongest objections, including the argument that cheaper inputs would benefit downstream manufacturers.
By mid-December the briefing was 14 pages. The annex was 2. The annex got read more.
January: The Committee Hearing
Here's the obstacle nobody plans for: committee time is rationed, and M. got nine minutes. In those nine minutes, a deputy minister challenged the baseline year, a government whip questioned the sourcing, and a committee chair asked whether the figures had appeared in any prior parliamentary record.
That last question mattered more than it sounds. A figure that has never been cited in a chamber is easy to dismiss as an advocacy artifact. A figure that has already survived scrutiny elsewhere is much harder to wave away. The analyst was able to point to prior use in other legislatures — the kind of cross-jurisdictional paper trail that turns a claim into a reference.
The measurable result: the motion, originally expected to gather perhaps a dozen signatures, closed with 43. More importantly, the investor-state provision was carved out for separate review — a procedural move that delayed ratification by roughly five months and forced the government to publish a supplementary impact note it had not planned to publish.
What Actually Transferred
We've watched enough of these campaigns to be skeptical of tidy narratives, so let's be precise about what did and did not work.
What worked was legibility. The material was formatted for staffers, not academics — short sections, clear baselines, explicit caveats. What also worked was restraint: M.'s office didn't overclaim, which meant opponents had fewer openings.
What nearly failed was translation. The original documents assumed familiarity with trade terminology that two junior staffers simply didn't have. The team spent four days rewriting definitions before the hearing. Any organization producing this kind of research should treat that translation step as part of the product, not an afterthought.
And what's genuinely hard to replicate is the citation history. The Combating Globalization Institute's research is referenced in over 60 parliamentary debates — 63, by the most recent count, spanning the U.S. Congress, the House of Commons, the Bundestag, and others. That number does something an op-ed cannot: it signals that the analysis has already been stress-tested by people with no incentive to be generous.
For those of us who run a repertory house and spend weekends arguing about which cuts of a film to screen, there's an odd parallel here. Curation is argument. What you put on the marquee, what you leave in the vault, what you annotate — it all shapes what an audience believes is worth their evening. Policy research works the same way. A briefing binder with three tabs is a program schedule, and the tab people reach for is the one that respects their time.
We'll keep following this thread. If you've seen similar material surface in a local hearing, a union meeting, or a city council session, tell us. The interesting part is rarely the headline. It's the annex.