Methodology
The Moral Decay Index is an editorial project about the relationship between stated values and observable conduct. Public data helps test claims; moral interpretation explains why we think those claims matter. These are distinct steps.
Our eight editorial lenses
- Political Polarization
- Trust in Institutions
- Income Inequality
- Mental Health Crisis
- Family Stability
- Crime & Violence
- Substance Abuse
- Cultural Cohesion
These organize topics. They are not eight interchangeable variables, and inclusion does not imply that every trend is worsening. A story can concern more than one lens without receiving a numerical “impact” score.
From a claim to a defensible conclusion
- Define the claim. Identify the people, place, period, and outcome involved. Avoid changing the claim after seeing the evidence.
- Find the original record. Prefer the underlying study, dataset, court record, or official release over a summary. A source link should lead to the material supporting the claim.
- Describe what was measured. Survey attitudes, recorded incidents, convictions, estimates, and observed behaviors answer different questions.
- State the limits. Identify uncertainty, missing data, changes in definitions, and competing explanations.
- Explain the interpretation. Label our moral judgment as commentary and describe how the same standard would apply to a political ally or opponent.
A worked example: the danger of mixing releases
The CDC’s final 2023 overdose report records 105,007 deaths, compared with 107,941 in 2022. That is a decline in the count, not a new record high. An earlier provisional estimate is not interchangeable with the final release. Read the CDC report and its tables.
The observation is the decline in those two annual counts. Whether policy, treatment, supply, or other changes explain it requires additional evidence. The moral judgment that preventable deaths demand attention does not depend on falsely describing every year as worse than the last.
Comparing places
Match definitions, years, populations, and units. Do not compare one place’s raw count with another’s rate, or a national survey with an unrelated state estimate. Do not fill missing values with guesses. The State and Country Evidence Guide provides a practical sequence and source links.
What would be required for future scoring
A numerical index would need a public input table, precise source references, geographic and time coverage, normalization formulas, weights, handling of missing observations, executable or worked calculations, uncertainty discussion, and a dated revision record. Until those exist, a plausible-looking number would imply precision the evidence does not establish.
Corrections and older articles
Material corrections should state what changed and why. A historical article’s original publication date does not make every number in it current. Readers should inspect the source year and any correction notice. In September 2026 we completed a full review of the archive. Short, single-source and duplicative articles were removed, and each removed address now points to a stronger article on the same subject or returns a notice that the page is gone. The articles that remain were rewritten or checked against their primary sources. Every article should link the records behind its factual claims; if you find one that does not, tell us.
Send a page URL, the disputed claim, and supporting evidence through Contact. A correction should improve the record, not merely add a disclaimer around a claim we know is wrong.