01DGrade D
69.99phantom-program
A verified phantom program. Any reconciliation whose match type is phantom.
How grades work
Three weighted components, five hard caps and five bands. Code applies them to every jurisdiction with ledger coverage.
50 %Reporting integrityDoes the certified report tie to the ledger.
40 %Spending conformityDid the money reach an approved use.
10 %TransparencyDid the records arrive when asked for.
CapThe lowest cap that firedA cap puts a ceiling on the composite.
Composite = 0.5 RI + 0.4 SC + 0.1 T, then the lowest ceiling any cap applies.
Who gets a letter
Eligibility
A letter grade requires complete ledger coverage of the opioid accounts over the graded window.
Partial coverage yields a provisional grade, marked as such, with the coverage window shown.
No ledger yields no letter at all. The jurisdiction carries a status badge instead. A refusal is displayed as prominently as an F, because a refusal is the one thing a jurisdiction fully controls.
A report is due September 1 after its fiscal year closes. A year the State lists as No Report Received has no report on file. It is never read as a report certifying 0.00. A filed report that leaves the amount expended blank or N/A states no figure. It reads Not stated and is left out of reporting integrity. Filed report years feed reporting integrity. So does a year whose report is past due and not on file when the books show spending in it, and that year scores 0. Spending in a year whose report is not yet due counts only in spending conformity. That year reads as Pending on the jurisdiction page, with what has been paid so far. When a report lands it enters reconciliation and the grade updates.
The status badgesRecords refusedRecords pendingNot yet requestedNo spending reported
The formula
Three components, zero to one hundred each
Each component is scored on its own, then weighted into the composite.
Reporting integrity
50 %From reconciliations across all filed report years, and from every past due year with no report on file. Misreported dollars over the larger of certified or actual, plus a count of failure types present.
Per filed year: misreported = |certified_expended - ledger_expended| ratio = 100 * (1 - misreported / max(certified, ledger)) Both zero gives 100. Books that net below zero count as 0 spent. The ratio is held from 0 to 100. A report that leaves the amount expended blank or N/A is left out. A past due year with no report and spending in the books gives 0. A year not yet due is left out. Subtract 10 for each failure type present across all years. Average the per year ratios, apply the deductions once, floor 0.
Spending conformity
40 %Non conforming dollars, plus half of unverified dollars, over total spend using the Exhibit E gate classification. Unverified dollars can lower the score but never raise it. A supplantation confirmed row counts as non conforming whatever category it names.
100 * (1 - (non_conforming + 0.5 * unverified) / (conforming + non_conforming + unverified)) over ledger spend in every year, reported or not. A zero denominator gives 100. Unverified counts at half weight. The score is the lower of this and 100 * (1 - non_conforming / (conforming + non_conforming)), so unverified never raises it.
Transparency
10 %Records request behaviour, and whether the governing body ever named a use in a resolution.
produced-complete 100 produced-partial 60 deficient 40 pending 70 not-requested 70 silent 0 refused 0 litigation 0
The approved rubric lists four further failure types under reporting integrity. A certified encumbrance with no book commitment, unspent on hand misstated, a self contradiction inside one report, and a report resubmitted after a State flag. Version 3 does not score them, because they need appropriation and State flag data the pipeline does not yet load. They enter the formula when those loaders exist.
Ceilings
Hard caps
A cap is applied after the composite is computed. The lowest ceiling wins, and the finding that triggered it prints beside the grade.
02DGrade D
69.99zero-certification
A filed report year certifying zero expended over real spend in the books.
03DGrade D
69.99transfer-out-uncertified
An interfund transfer out of the opioid accounts to general, police or capital lines with no certified program.
04FGrade F
55zero-conforming-spend
Settlement money went to uses that failed the Exhibit E gate and not one dollar reached a conforming use. Clean arithmetic in the report does not lift it.
05CGrade C
79.99records-refused
Records refused, or litigation required to obtain them. A refusal is displayed as prominently as an F.
The bands
From a composite to a letter
Rubric version 5 grade bands.
01AGrade A
90 and aboveGrade A
The report ties to the ledger and the money reached an approved use.
02BGrade B
80 to 89.99Grade B
Small differences, none of them structural.
03CGrade C
70 to 79.99Grade C
Material differences, or records that did not fully arrive.
04DGrade D
60 to 69.99Grade D
A capped failure type is present.
05FGrade F
below 60Grade F
Certified spending that does not exist in the books, or spending with no conforming dollar in it.
Parking money is a separate displayed metric, the unspent share of what was received, and it is not penalised. An honest reporter with conforming spending lands at A regardless of how much was spent.
Worked examples
Grades as the code computed them
Published jurisdictions, read from their own panels.
Washington TownshipThe weighted figure already sits below every ceiling that fired, so no cap changed it.
Composite score11.15FGrade F
ComponentWeightScore
Reporting integrity50 %10.3 / 100
Spending conformity40 %0 / 100
Transparency10 %60 / 100
Weighted, before caps11.15
Caps that fired
phantom-programCertified spending that does not exist in the booksmax 69.99
zero-conforming-spendNon-conforming spending with no conforming dollar in the booksmax 55
Grade after capsF
Mantua TownshipThe weighted figure already sits below every ceiling that fired, so no cap changed it.
Composite score42.5FGrade F
ComponentWeightScore
Reporting integrity50 %65 / 100
Spending conformity40 %0 / 100
Transparency10 %100 / 100
Weighted, before caps42.5
Caps that fired
zero-certificationA report year certified at zero expended while the books show paymentsmax 69.99
transfer-out-uncertifiedMoney transferred out of the opioid account with no certified programmax 69.99
zero-conforming-spendNon-conforming spending with no conforming dollar in the booksmax 55
Grade after capsF
Deptford TownshipNo cap fired, so the weighted figure is the composite.
Composite score93.2AGrade A
ComponentWeightScore
Reporting integrity50 %100 / 100
Spending conformity40 %82.99 / 100
Transparency10 %100 / 100
Weighted, before caps93.2
Grade after capsA
Rubric versions
Versioning
Scores store a rubric version. A grade is always read at the version that produced it.
When the rubric changes, every jurisdiction is regraded at the new version. A correction from a jurisdiction that changes a verified row regrades that jurisdiction on its own and links the submission in the corrections panel on its page.
Version 2 adds one cap. A jurisdiction that spent settlement money with no conforming dollar in the books caps at 55, an F. The components, the weights, the bands and the four earlier caps are unchanged from version 1.
Version 3 counts unreported years. A year whose report is past due and not on file, with spending in the books, enters reporting integrity at 0. Spending conformity already ran over every year, reported or not.
Version 3 also pairs a transfer split across several opioid lines. A transfer out whose transfers in share its date and reference number and sum to it to the cent is a move inside the fund. Money that never left the fund is not spending. The weights, the bands and the five caps are unchanged from version 2.
Version 4 counts unverified spending at half weight in spending conformity. Before version 4 unverified dollars were left out of the score, so a jurisdiction whose spending the records could not place scored as if all of it conformed. Unverified dollars can lower the score but never raise it.
Version 5 narrows the version 2 cap. It applies only when some spending failed the Exhibit E gate and none conformed. Spending the records cannot place is unverified, not failed. It counts at half weight in spending conformity and no longer triggers the cap on its own.