Public comparison infrastructure
Antitrust Radar / comparative research desk
Comparison Observatory
Compare evidence within a consistent definition, period, and unit. Choose a subject to see the available comparisons and their limits.
- Research frames
- 7
- question-led routes
- Comparison workspaces
- 31
- shareable instruments
- Atlas profiles
- 5,478
- 19,193 declared edges
- Synthetic scores
- 0
- unlike evidence stays separate
Comparability ruleA visible difference becomes analytical evidence only inside a declared common frame.Audit the rule
Question-led comparison router
Start with the claim you need to test.
Each route carries its own unit, population, period, source, and exclusion rule into the analytical workspace.Current research frame Places2 comparison systems
State insurance concentration
Where is concentration highest under one geographic frame?
Compare states or counties within a single dataset and vintage, with ranks and benchmarks that preserve geography class.Recommended starting instrumentState comparison studio
Compare state individual-insurance HHI, leading-insurer share, rival depth, national ranks, and regional benchmarks from one release.
- Current scope
- 51
- state profiles
- Strict frame
- Same insurance dataset, vintage, unit, and geography class
Comparability passport
What travels with this comparison
- Unit
- State insurance or county branch-deposit screens
- Alignment
- One dataset, vintage, metric, and geography class
- Output
- Maps, ranks, regional benchmarks, and local profiles
- Never inferred
- State and county evidence systems are never merged
Available instruments
Places and local markets
- 01State insurance concentrationState comparison studioScope51state profilesComparison contractSame insurance dataset, vintage, unit, and geography classA state screen is not a local product-market definition
- 02FDIC branch-deposit screensCounty banking comparison studioScope3,104county deposit screensComparison contractSame FDIC vintage, county unit, and deposit-share denominatorCounty deposit screens are not combined with state insurance
Comparison systems ledger
All public comparison engines
| Research frame | Instrument | Current scope | Evidence boundary | Launch |
|---|---|---|---|---|
| Markets | MarketsDefinition-aware market screens | 12public market profiles | Profile-native concentration screens and within-profile historiesNo average across unlike products or synthetic market-power rank | |
| Markets continued | IndicatorsExact paired analytical plane | 30bounded indicators | Same entity, same period, declared indicator relationshipDescriptive association is not a causal or legal conclusion | |
| Firms | ParticipantsWithin-market participant shares | 55participant profiles | One market definition and one reported yearNo comparison of shares drawn from different denominators | |
| Firms continued | OperatorsOfficial attribution footprints | 20operator dossiers | Declared attribution records, kept separate by marketNo inferred ownership, control, or aggregated ecosystem share | |
| Places | StatesState insurance concentration | 51state profiles | Same insurance dataset, vintage, unit, and geography classA state screen is not a local product-market definition | |
| Places continued | County bankingFDIC branch-deposit screens | 3,104county deposit screens | Same FDIC vintage, county unit, and deposit-share denominatorCounty deposit screens are not combined with state insurance | |
| Economy | IndustriesDetailed Census industry structure | 899six-digit industries | Same NAICS level, Census measure, and strict matched-year rowsIndustry concentration is not a legal antitrust market | |
| Economy continued | SectorsBroad-sector structure and flows | 17structure sectors | Exact measures remain separate across their native vintagesNo sector HHI and no blended structure-flow score | |
| Economy continued | Business dynamicsOfficial Census flow histories | 19 x 46sectors x annual vintages | Same Census BDS metric, year, and sector levelBusiness flows are not concentration or causal estimates | |
| Economy continued | Consumer pricesOfficial monthly BLS price histories | 9 x 115series x monthly slots | Same BLS area, monthly frequency, and adjustment statusNo concentration, market-power, margin, pass-through, welfare, or causal inference | |
| Economy continued | Labor dynamicsOfficial monthly BLS worker flows | 12 x 114industry scopes x months | Same JOLTS rate, national industry scope, monthly frequency, and adjustment statusNo employer concentration, monopsony, wage, local-market, or causal inference | |
| Economy continued | Employment + earningsOfficial monthly BLS payroll context | 11 x 115industry scopes x months | Same CES measure, national private-industry scope, month, adjustment, and sourceNominal averages and payroll counts do not measure monopsony, bargaining power, job quality, or causation | |
| Economy continued | Productivity + costsOfficial quarterly BLS production context | 6 x 38major sectors x quarters | Same BLS index basis, major-sector scope, quarter, adjustment, and source releaseNo market concentration, markup, profit, employer-power, innovation, or causal inference | |
| Economy continued | Labor shareLong-run BLS production accounts | 3 x 318aggregate scopes x quarters | Same BLS 2017=100 reference, aggregate production scope, quarter, adjustment, and source releaseNo profit share, markup, concentration, employer power, bargaining power, legal market, or causal inference | |
| Economy continued | Yearbook yearsExact vintage-to-vintage evidence | 49observed yearbook vintages | Exact domain, entity, metric, unit, and period-basis pairsUnmatched rows stay visible and never become zero change | |
| Policy | Policy systemsRegulation and authority activity | 65economy evidence systems | PMR and enforcement remain distinct, source-labeled tracksNo policy effectiveness score or causal ranking | |
| Policy continued | RecommendationsPublished think-tank policy architecture | 37published recommendations | Exact classifications, report custody, and repeated evidence IDsNo similarity, feasibility, adoption, effectiveness, or impact score | |
| Policy continued | Merger review yearsFederal HSR reporting pipeline | 10official HSR fiscal years | One annual source, federal fiscal-year basis, and exact published fieldsNo merger-rate estimate, agency-performance score, or harm inference | |
| Policy continued | JurisdictionsSame-source authority histories | 59jurisdictions | Same OECD metric, annual period, and jurisdiction classAdministrative counts do not measure deterrence or welfare | |
| Policy continued | Enforcement industriesIndustry-level authority activity | 20enforcement categories | Same OECD activity metric, category system, and periodDecision counts do not establish underlying violation rates | |
| Policy continued | ProceedingsDeclared matter chronologies | 9declared proceedings | Explicit matter membership, exact dates, stages, and repeated identifiersNo posture, agency-quality, legal-effect, or causal ranking | |
| Policy continued | Official recordsCase-stage and evidence-window audit | 14official records | Declared record stages, dates, entities, and evidence windowsChronological proximity is not a causal event-study result | |
| Policy continued | Issue programsPublic-interest research architecture | 6issue programs | Ten exact evidence layers and shared source custodyPrograms are not ranked and unlike market measures stay apart | |
| Research | Research questionsSignal-derived empirical contracts | 46public research questions | Native metric lanes and exact topic, source, method, entity, metric, and request membershipsNo common measurement axis, research-priority score, or causal conclusion | |
| Research continued | Research reportsLong-form publication contracts | 11long-form research reports | Exact source IDs, market links, themes, dates, series, and sectionsNo research quality, importance, rigor, impact, or policy-effect score | |
| Research continued | Insight dispatchesPublished change-field evidence | 7published insight dispatches | Exact dispatch memberships, signal IDs, evidence layers, and releasesShared topics or sources do not imply equal findings, quality, or impact | |
| Research continued | Publication seriesInstitutional publication continuity | 6permanent research series | One institutional release clock with exact series and theme membershipsPublication volume, breadth, and overlap are not quality or impact scores | |
| Research continued | Research topicsZero-preserving subject histories | 8research topic profiles | One publication register with exact topic and series membershipsRepeated memberships do not become topic similarity or research priority scores | |
| Research continued | Public sourcesEvidence coverage and custody | 45public source dossiers | Exact observed years, native coverage channels, publishers, and custodyNo source quality, authority, completeness, or evidentiary-strength score | |
| Atlas | Relationship pathsBilateral declared-edge audit | 5,478canonical Atlas nodes | Exact node IDs and declared edges in the selected layerMulti-hop proximity never becomes a direct endpoint claim | |
| Atlas continued | Research setsMulti-profile evidence composition | 2-6 of 5,478canonical profiles per set | Pairwise direct, bridged, or disconnected statusA bridge route does not imply ownership, liability, or causation |
Cross-domain research