Antitrust Radar research architecture
Compare research topics, publication histories, and shared evidence. Missing coverage remains visible and is not treated as a zero economic outcome.
Comparison standard. Topics connect only through exact publication IDs, agenda-question IDs, issue-program IDs, dataset IDs, or route paths. Shared custody does not establish shared findings, methods, importance, legal relevance, or causation.
Evidence standardResearch program comparison
Comparative evidence note
3 exact identifiers recur in two or more selected contracts. The widest pair is Methods and Public Data with Geographic Competition: 3 shared items.
1 structural-zero cell remains across the selected contracts. A zero records the current editorial inventory, not the absence of relevant real-world evidence.
Exclusive means assigned to one selected editorial contract. It does not rank research depth, policy importance, evidentiary quality, or legal relevance.
Exact membership matrix
Bar length is normalized within each row for legibility. Counts retain their own units and are never added into a score.
Publication architecture
Comparison contract
What can current public records establish, where do measurement boundaries stop, and which disclosures would make antitrust research more accountable?
Document definitions, denominators, source construction, release clocks, reproducibility paths, and evidence gaps as first-class research outputs.
A documented data gap is not evidence of competition or harm. Screening thresholds are not legal conclusions, and reproducibility does not make unlike measures comparable.
How does competitive structure vary across national, state, county, tract, route, and other local frames?
Connect place-based concentration screens to their exact geography, source vintage, ownership rules, and local-market research needs.
A statewide, county, tract, or national screen does not establish a legal geographic market. Medians of local HHIs are not interchangeable with one national HHI.
Reproducibility contract
The studio derives every matrix cell, edge, and register row from the same topic model published through the public data endpoint.
Publication, agenda, issue, and dataset identifiers plus canonical route paths create the comparison ledger.
Empty membership lanes remain explicit and are never imputed from adjacent research programs.
Selected topic summaries and membership-level comparison packets are available as JSON and CSV.