Datasets That Can Change A Score

Unemployment, labor force, payroll jobs, wages, CPI, PPI, and import/export prices.

Cadence

Monthly for most headline series.

Query targets

LNS14000000 unemployment rate; CES0000000001 payroll employment; CUUR0000SA0 CPI-U

Ledger use

Recession and labor stress; Inflation shock; Citizen harm baselines

Caveat

Store vintage date and revision date. Do not mix seasonally adjusted and raw series.

Federal Reserve, Treasury, labor, credit, rates, yield curves, GDP, money, and vintage macro series.

Cadence

Depends on source series; ALFRED keeps vintages.

Query targets

FEDFUNDS; T10Y2Y; BAA10Y; SP500; ALFRED vintages

Ledger use

Inflation and recession context; Credit stress; Policy-rate shocks; Historical revision checks

Caveat

Use ALFRED when the claim depends on what officials knew at the time. Free price-index rows are not total-return market history.

Population, ACS, poverty, migration, education, housing, income, and local geography.

Cadence

Annual ACS, decennial census, and program-specific releases.

Query targets

ACS 5-year poverty; Population by county; Educational attainment; Housing cost burden

Ledger use

Citizen exposure denominators; Regional comparisons; School and housing context

Caveat

Small geographies have margins of error. Show denominator and geography.

Debt, deficit, outlays, receipts, auctions, savings bonds, and Treasury accounting tables.

Cadence

Daily, monthly, quarterly, or annual by endpoint.

Query targets

Debt to the Penny; Monthly Treasury Statement; Treasury auctions

Ledger use

Fiscal stress; Debt-limit conflict; Emergency spending context

Caveat

Keep units explicit. Some endpoints return dollar strings, not numbers.

Federal awards, obligations, recipients, agencies, programs, locations, and spending categories.

Cadence

Frequent updates from agency award reporting.

Query targets

Fiscal-year obligations; Agency awards; Recipient records; Spending category

Ledger use

Federal spending context; Grant and contractor exposure; Fiscal-stress denominators

Caveat

Obligations are not final outlays. Reconcile with Treasury outlays before scoring fiscal burden.

Budget projections, economic projections, historical budget data, spending, revenues, debt, and deficit baselines.

Cadence

Recurring updates tied to CBO reports and budget baselines.

Query targets

Historical budget data; Budget and economic outlook tables; Long-term budget projections

Ledger use

Fiscal stress; Debt and deficit baselines; Policy-cost comparisons; Historical burden comparisons

Caveat

CBO projections are conditional baselines, not predictions. Store report date, table, units, and scenario assumptions.

Daily VIX open, high, low, close history from 1990 onward.

Cadence

Daily market days.

Query targets

VIX close; VIX drawup windows; Volatility regime breaks

Ledger use

Market fear proxy; Crisis timing; Stress comparison across eras since 1990

Caveat

VIX is not a full market-history source. Pair with licensed or exchange-grade return data.

Company filings, submissions, XBRL facts, insider transactions, and fund records.

Cadence

Near real time for filings; facts update after filing acceptance.

Query targets

Company submissions; Company facts; Forms 3/4/5; 10-K risk factors

Ledger use

Corporate exposure; Financial-sector stress; Conflict-of-interest checks

Caveat

Use a clear User-Agent. EDGAR is disclosure data, not a market-price feed.

Presidential documents, agency rules, notices, proposed rules, document metadata, dates, citations, and full-text links.

Cadence

Daily on federal publication days.

Query targets

Executive orders; Presidential memoranda; Agency rules; Document citation

Ledger use

Official action text; Regulatory timing; Executive-power cards; Source replacement for press summaries

Caveat

API metadata can change after correction. Store document number, publication date, and retrieval date.

Public laws, congressional reports, Federal Register, Code of Federal Regulations, and official PDFs.

Cadence

Collection specific.

Query targets

Public laws; Congressional reports; Federal Register documents

Ledger use

Primary law text; Regulatory change; Official investigation reports

Caveat

Use package and granule IDs so citations can be checked later.

NAEP reading, math, long-term trend, subgroup, grade, jurisdiction, and assessment tables.

Cadence

Assessment cycle dependent.

Query targets

Age-9 long-term trend; Grade 4 reading; Grade 8 math; State subgroup gaps

Ledger use

Learning-loss claims; School-system damage; Distributional harm

Caveat

Assessment design matters. Store population, subject, grade, jurisdiction, and scale notes. Scores alone do not prove school-closure causality.

Global news event records, source metadata, themes, locations, and television archive signals.

Cadence

Frequent updates.

Query targets

Event volume; Theme counts; Source geography; TV mentions

Ledger use

Press-intensity checks; Event discovery; Narrative-volume controls

Caveat

Press volume is not damage. Treat it as a discovery and salience signal.

Public web captures, texts, audio, video, software, metadata, and collection search.

Cadence

Continuous archive growth.

Query targets

Wayback captures; Public-domain texts; TV archive metadata

Ledger use

Source preservation; Historical public records; Quote verification

Caveat

Archive capture proves availability, not truth. Respect rights metadata.

IRS Form 990 filings, nonprofit metadata, compensation, revenue, grants, and filing history.

Cadence

IRS release dependent.

Query targets

Form 990 revenue; Grants; Officer compensation; Filing history

Ledger use

NGO funding trails; Advocacy incentives; Institutional trust checks

Caveat

IRS filings lag. Grant purpose text can be vague.

Federal campaign finance, committees, candidates, contributions, and independent expenditures.

Cadence

Frequent files plus cycle-level bulk releases.

Query targets

Individual contributions; Committee summaries; Independent expenditures

Ledger use

Donor capture; Campaign-finance escalation; Outside-spending concentration

Caveat

Names need normalization. Refunds, transfers, and joint committees require careful handling.

Campaign contributors, candidates, committees, ideology estimates, geography, transactions, and election-cycle records.

Cadence

Dataset release dependent.

Query targets

Contributor ideology; Candidate ideology; Geographic donor networks; Cycle-level transactions

Ledger use

Donor ideology context; Campaign-finance escalation; Polarization and money links; State and federal influence networks

Caveat

Ideology estimates are modeled from contribution behavior. Store version, release date, and estimation caveats before scoring motive or capture.

County, precinct, state, and national election returns across several offices and years.

Cadence

Dataset releases after election cycles.

Query targets

County presidential returns; Precinct returns; Senate and House returns

Ledger use

Election realignment; Turnout shocks; Geographic polarization

Caveat

Precinct boundaries change. Use documented joins and avoid false precision.

Public CSV datasets used in FiveThirtyEight projects, including polling, elections, and political behavior datasets.

Cadence

Project dependent; some archived datasets are static.

Query targets

Polling averages inputs; Election forecasts inputs; Political behavior datasets

Ledger use

Polling trust checks; Election-context comparisons; Public replication examples

Caveat

Treat archived project data as a secondary analytical layer. Prefer original official returns when scoring elections.

Congressional roll-call votes, member ideology scores, parties, chambers, and Congress metadata.

Cadence

Updated with congressional voting records.

Query targets

DW-NOMINATE; Roll-call votes; Member party history

Ledger use

Polarization; Party sorting; Legislative coalition change

Caveat

Ideology scores summarize roll calls. They do not prove motive.

Expert and public surveys on democratic performance, institutional norms, and threats to American democracy.

Cadence

Survey-wave dependent.

Query targets

Expert ratings; Public ratings; Norm violations; Institutional performance

Ledger use

Guardrail perception checks; Norm-erosion comparisons; Uncertainty around elite and public assessments

Caveat

Survey ratings measure perception and expert judgment. Pair with events, documents, and outcomes.

Cleaned democracy indicators, V-Dem-derived series, Polity, Lexical Index, and related historical country datasets.

Cadence

Dataset dependent.

Query targets

Democracy indices; Regime categories; Historical country-year comparisons

Ledger use

World-comparison baselines; Historical context; Cross-index checks

Caveat

OWID is a distribution and visualization layer. Cite original datasets when a score depends on them.

GDP, personal income, industry accounts, regional accounts, and international accounts.

Cadence

Monthly, quarterly, and annual depending on table.

Query targets

NIPA real GDP; Regional personal income; Industry value added

Ledger use

National output shocks; Regional decline; Industry exposure

Caveat

Use table names and line codes in stored metadata. BEA revisions can be material.

Bills, amendments, members, committees, summaries, text, actions, nominations, and treaties.

Cadence

Daily or near daily during active sessions.

Query targets

Bill actions; Bill text; Sponsors; Committee actions

Ledger use

Legislative production; Emergency statutes; Institutional repair attempts

Caveat

Bill status does not measure policy effect. Pair with implementation and outcome data.

Opinions, dockets, oral arguments, judges, citations, and RECAP records where available.

Cadence

Continuous updates from court and RECAP sources.

Query targets

Opinions; Dockets; RECAP filings; Citation graph

Ledger use

Primary legal support; Court-check timing; Lawfare and civil-liberties claims

Caveat

REST access requires credentials. Coverage varies by court and docket, and paid PACER gaps can remain.

NIBRS, hate crime, law-enforcement officers killed and assaulted, and national crime estimates.

Cadence

Annual and periodic releases.

Query targets

National estimates; Agency participation; Hate crime tables; NIBRS offenses

Ledger use

Crime-policy claims; Public-order comparisons; Police-capacity debates

Caveat

Participation changed during NIBRS transition. Label coverage gaps.

Long-run U.S. stock market, dividends, earnings, interest rates, CPI, and CAPE valuation series.

Cadence

Periodic updates.

Query targets

S&P composite price; Dividends; Earnings; CAPE; Long rates

Ledger use

Long-run market context; Crisis valuation comparisons; Citizen wealth exposure; Historical market regime checks

Caveat

Check license and definitions before monetized exports. Use total-return and constituent-aware data when the claim requires investable returns.

Federal campaign-finance summaries, donor and industry profiles, lobbying, outside spending, and state/local money through FollowTheMoney.

Cadence

Cycle and filing dependent; API and bulk access depend on account terms.

Query targets

Industry giving; Outside spending summaries; Lobbying clients; State and local donor networks

Ledger use

Donor capture; Influence-network context; Lobbying pressure; State and municipal money trails

Caveat

Use FEC bulk data as the primary federal record. Treat OpenSecrets as a cleaned analytical layer unless license terms allow raw redistribution.

State and local public records, including campaign contributions, campaign expenditures, licenses, inspections, salaries, and other government datasets.

Cadence

Dataset dependent; many collections are historical snapshots from state or local publishers.

Query targets

State campaign contributions; State campaign expenditures; Original columns; Normalized search fields

Ledger use

State and local donor networks; Municipal corruption checks; Public-record discovery; Entity matching outside federal files

Caveat

Treat TAP as a normalization and discovery layer. Keep original publisher, date range, original columns, and TAP processing notes before scoring.

Survey microdata and toplines for political values, trust, media, religion, identity, and public opinion.

Cadence

Survey dependent.

Query targets

Political typology; Institutional trust; News consumption; Partisan attitudes

Ledger use

Social-trust context; Media-trust claims; Polarization baselines

Caveat

Use survey weights and question wording. Public opinion explains climate; it does not prove institutional damage by itself.

State voting-law changes, bill classifications, restrictive and expansive voting measures, and yearly roundups.

Cadence

Periodic roundups during and after state legislative sessions.

Query targets

State; Bill type; Restrictive or expansive classification; Effective date

Ledger use

Voting-access claims; State-by-state democracy comparisons; Election-administration stress

Caveat

Advocacy institution with transparent classifications. Pair with bill text and state records for final scoring.

Candidate and officeholder risk indicators tied to election denial, threats to election administration, and anti-democratic behavior.

Cadence

Election-cycle dependent.

Query targets

Candidate; Office; State; Risk classification; Evidence links

Ledger use

Election-denial network mapping; State and local office risk; Democracy-threat triage

Caveat

Use as a lead source. Verify each entry against primary statements, court records, election records, or official actions.

Reports and datasets on election denial, state office risks, litigation, and democracy-threat actors.

Cadence

Report dependent.

Query targets

Candidate lists; Office type; State; Election-denial evidence; Legal posture

Ledger use

State election-risk cards; Election-denial backfill; Actor network context

Caveat

Use with primary records and clear methodology labels because the source is advocacy-aligned.

Protests, demonstrations, political violence, actors, locations, dates, sources, and event notes.

Cadence

Provider release dependent.

Query targets

Event type; Actor; Location; Fatalities; Source; Date

Ledger use

Public-order claims; Political-violence comparisons; Protest and riot denominators

Caveat

Event coding and media coverage can bias counts. Store source counts and coding definitions.

Political violence, threats and harassment, demonstrations, mitigation efforts, and local civic-space records.

Cadence

Dataset dependent.

Query targets

Threat event; Actor; Target; Location; Event type; Source

Ledger use

Political-violence denominators; Threats against officials; Local public-order stress; Event cross-checks with ACLED

Caveat

Use with dataset-specific codebooks and BDI/CivicPulse methodology notes. Event coverage depends on reporting, coding rules, and source availability.

Authoritarian-risk events, democratic guardrail pressure, institutional category, direction of movement, and event links.

Cadence

Updated as new actions are reviewed.

Query targets

Event; Guardrail category; Threat direction; Institution; Evidence link

Ledger use

Democracy-threat triage; Executive-power context; Directional guardrail movement

Caveat

Use as a discovery and categorization layer. Store primary documents before changing a score.

Legal challenges, case posture, filings, policy area, parties, and court actions for major public-law disputes.

Cadence

Updated as litigation develops.

Query targets

Case; Court; Issue; Posture; Filings; Decision date

Ledger use

Legal-status fields; Court-check timing; Executive-power and civil-liberties claims

Caveat

Use trackers to find cases. Store primary court records before changing a score.

Executive-order lawsuits, issue areas, plaintiffs, court posture, and outcomes.

Cadence

Updated as litigation develops.

Query targets

Order; Case; Issue; Court; Status; Outcome

Ledger use

Executive-order litigation triage; Legal posture comparison; Source discovery

Caveat

Use as a tracker and chronology aid. Primary dockets and opinions remain the scoring record.

Open datasets, text corpora, metadata, and processing tools across many providers.

Cadence

Dataset specific.

Query targets

Public text corpora; Dataset cards; License fields; Revision hashes

Ledger use

Document classification; Corpus search; Claim clustering

Caveat

Review each dataset card and license. Do not mix synthetic text with primary evidence.

Survivorship-aware equity returns, delistings, index constituents, credit, rates, and corporate actions.

Cadence

Vendor specific.

Query targets

Total return indices; Delisting returns; Constituent history; Corporate actions

Ledger use

Long-run market drawdowns; Sector stress; Wealth and pension exposure

Caveat

Do not make all-history market claims from free symbol downloads alone.