How to read a department report card

The report cards on this site put a letter grade on a police department, and a letter grade invites a simple reading: good department, bad department. That reading is wrong, and the cards are built to resist it. Here is what a grade actually is.

A grade is a position among peers

Every graded department is compared only with agencies of the same type and roughly the same size. A small borough police department is measured against other small municipal departments, never against a county jail or the state police. Within that peer group, the site ranks each department on what the records can measure and reports which fifth of the group it lands in. Grade A is the lowest fifth on reported volume and severity. Grade F is the highest fifth.

So an F does not mean "failing." It means this department sits in the top fifth of its peers on the measured quantities. That is a real signal, and it is also a relative one. Move the department to a different peer group and the letter could change without a single record changing.

How the grades are distributed

Because a grade is a fifth of a peer group, the letters come out nearly even by construction. Of the 548 agencies on the site, 540 carry a grade: 108 are graded A, 107 B, 110 C, 107 D, and 108 F.^[2] The remaining 8 are ungraded, 7 because they have too few peers to rank against and 1 because it reported too little data to score.^[2]

That flat distribution is the first thing to know about the cards. A department with a D is not in a minority of bad departments; a fifth of every peer group gets a D. The letter locates a department on a curve. It does not say the curve itself is good or bad.

What "agencies like it" means

The peer groups are the whole method, so they are worth stating. Municipal police departments are split into four bands by the number of full-time sworn officers they reported to the FBI: 1 to 14 officers, 15 to 29, 30 to 59, and 60 or more. The other agency types each form a single group.^[2]

Peer group	Agencies in the group
Municipal police, 1 to 14 officers	100
Municipal police, 15 to 29 officers	171
Municipal police, 30 to 59 officers	110
Municipal police, 60 or more officers	93
County sheriff's departments	21
County prosecutor's offices	21
Corrections agencies	15
State agencies	10
Other agencies	6

Source: the report-card results this site computes and publishes.^[2]

The bands are why a suburban department of 20 officers is never on the same curve as Newark, and why a county jail is graded only against the other county jails. Where the FBI count is missing, as it is for county corrections and sheriff's agencies, the peer group falls back to the Attorney General's own officer count, and the card says so.

What goes into it

Three things, each shown on the card as its own sub-score. Internal affairs volume, measured per officer so a large department is not penalized for being large, at a weight of 40 percent. Discipline severity, which weights a termination more heavily than a short suspension, also at 40 percent. And the mix of allegations, weighted so that excessive force counts for more than a demeanor complaint, at 20 percent.^[1] The weights are an editorial choice, and the site publishes them rather than hiding them.

Reporting gaps do not quietly change the grade. They lower a separate confidence indicator, shown next to the grade with its reasons written out, so a grade built on thin data is labeled as such.

Why a grade can be low confidence

Confidence is reported on every card as high, medium, or low. Across the 540 graded agencies, 142 are high confidence, 371 medium, and 27 low.^[2] Three reasons account for nearly all of the downgrades. The most common, on 399 cards, is that the agency is small, so a per-officer rate swings widely on a handful of cases.^[2] The second, on 108 cards, is that the Attorney General listed the agency as reporting no internal affairs investigations in at least one year, which means part of the volume score rests on an absence.^[2] The third, on 19 cards, is that the agency reported both investigations and officer counts in only one or two years, too few to average.^[2]

Confidence and grade are independent. No F on the site is low confidence: the 108 F grades split 36 high and 72 medium.^[2] The low-confidence cards cluster at A and B, 13 and 5 of the 27, which is what you would expect when a small agency with a quiet year or a missing year lands at the low end of its peers on numbers that could not carry much weight.^[2] That is the reason the confidence label sits beside the letter rather than under it.

A worked example

Take Aberdeen Township Police Department in Monmouth County. Its card reads grade B, medium confidence, in the peer group of 110 municipal departments with 30 to 59 officers.^[2] Its combined position is the 21st percentile of that group, which places it in the second-lowest fifth and earns the B.

The three sub-scores show where that position comes from. On internal affairs volume, Aberdeen averaged 112.9 investigations per 100 reported officers across the years it reported both figures, which is the 18th percentile of its peers: low.^[2] On discipline severity it scored 44.1 severity-weighted records per 100 officers over 2020 to 2025, the 59th percentile: a little above the middle.^[2] On allegation severity its mean tier was 1.96 on a scale of 1 to 4, the 35th percentile.^[2] Weight those at 40, 40, and 20 percent and the combined percentile lands at 21. The confidence note says why it is medium rather than high: a department this size has few enough cases that one year could move the volume score a long way.

Nothing in that card says Aberdeen is a good department or a bad one. It says that among 110 departments of its size, it reported fewer investigations per officer than most and discipline of roughly middling severity, on figures thin enough to deserve a caveat. That is the whole claim, and every number in it is on the department's own page.

The one caveat that matters most

Higher internal affairs volume is ambiguous. It can mean more misconduct. It can also mean a department that investigates complaints seriously and reports them fully, which is the opposite of a problem. The grade weights volume anyway, because volume is what the record offers, but that ambiguity is the reason a grade is one lens and not a verdict.

Everything a grade rests on is one click away on the department's own page: the raw totals, the individual discipline records, the agency's own synopses. The grade is the site's analysis. The records are the facts. When the two seem to disagree, trust the records, and read the methodology for exactly how the grade was built.