The Civil Service Commission's salary lookup has a dropdown most people never touch. It lists fiscal years going back to 1985. Behind it sits every pay schedule the state has published for four decades, and we have now captured all of it: 36 schedules, 863 separate dates on which somebody's rates changed. One year's table tells you what your range pays. All 43 of them tell you whether it kept up.

The number that flatters

Start with the schedule most state employees are on. A-P-R-S covers the four CWA units at once, from clerical staff through senior supervisors. Step 1 of range 9, its lowest published rung, paid $10,888 when the rates took effect in December 1984. The same rung pays $39,358 today.

That is 3.61 times the 1984 figure, and stated on its own it sounds like a large gain. It is not, because a 1984 dollar and a 2026 dollar are different things. Running both figures through the Bureau of Labor Statistics consumer price index leaves a real increase of about 17 percent across 42 years. The schedule moved 43 times to get there.

So we give both numbers throughout. The nominal figure is what the Commission printed; the real one is what it bought.

Eight schedules that stopped

Of the 36 schedules with published rates, eight have not been updated in at least two years. Five of them were last set before the 2008 recession.

ScheduleCoversRates last setChanges everValue retained
EState Police managerial and exempt19991252%
XExempt, appointed and policy staff20061863%
MManagerial20071964%
MBLaw enforcement management outside State Police2007964%
MDHigher education administration2007964%
YHSergeant investigators, homeland security2019279%
YEDetective investigators, not represented2022691%
VHLieutenant investigators, homeland security2023395%

"Value retained" is the share of its purchasing power a schedule's current rates hold against the year those rates were set, using CPI-U in 2025 dollars. A schedule updated this year sits at 100 percent.

ERG E is the oldest table in the compendium. Its rates were set on 3 July 1999 and have not changed since, which leaves them worth about 52 cents on the dollar against the year they were written. The managerial schedules are close behind at 64 percent.

Worth saying plainly: managerial and exempt salaries are usually set individually rather than by stepping up a published ladder. The frozen tables are the formal schedule of record, not a description of what a manager takes home. What the freeze tells you is that the record itself stopped being kept current.

Four pairs covering the same work

Some kinds of state work are covered by two schedules at once, one negotiated and one not. The archive lets you watch them separate.

WorkScheduleRepresented byChangesStep 1 now
Managerial staffM-RepresentedIBEW14$76,287
MNot represented19$48,646
Lieutenant state investigatorsVIPBA Local 383B16$127,549
VHNot represented3$106,804
Sergeant state investigatorsYIPBA Local 383A16$113,608
YHNot represented2$76,041
Detective state investigatorsYDPBA Local 38315$84,180
YENot represented6$55,439

Step 1 of the range each schedule published most consistently, which is not the same range number in every schedule. Follow any code to its full table on this site.

Two of these pairs have a precise starting point. On 12 July 2014 the homeland security investigator schedules VH and YH first appear in the record, at $81,212 and $70,642. Those were exactly the rates VI and YI had been paying until that same day, when both of those schedules moved up. Twelve years later the sergeant pair stands at $113,608 against $76,041.

These are the Commission's own rates for comparable ranks. We are not going to tell you what caused the gap, because the schedules do not say. You now have the same figures we do.

Who did keep up

Most schedules did beat inflation over the full record, though usually by less than the nominal figures suggest. The three that gained the most are all negotiated law enforcement tables that moved more than 40 times each: State Police non-commissioned officers at 1.47 times their 1984 value after inflation, the troopers table at 1.36, and non-trooper superior officers at 1.31.

At the other end, ERG YE has not changed at all since 2018. In nominal dollars its step 1 is identical in both years. In real terms it is at 0.86, which is a pay cut that never appeared in any document as a pay cut.

Two things one year of data hides

New Jersey used to reset pay twice in one fiscal year. FY1986 carries two effective dates for nearly every schedule, and that pattern repeats through the 1980s and 1990s. Reading only the current compendium, you would never know the state once adjusted salaries mid-year as a matter of routine.

Rates can also fall. The State Troopers table published step 1 of range 17 at $17,055.80 in June 1984 and reissued the whole table about 3 percent lower that December, increment included. Both figures are in the pages the Commission serves. Why it happened is not something the record explains, so we are noting it rather than accounting for it.

Data and methodology

Schedules and effective dates come from the Civil Service Commission's Salary Schedules by ERG lookup, captured for all 43 fiscal years it publishes, FY1985 through FY2027. The raw pages are kept with a checksum for each one so any figure here can be traced back to the document it came from. Coverage is 36 schedules and 863 distinct effective dates.

Real-terms figures use CPI-U for all urban consumers, U.S. city average, all items, series CUUR0000SA0, published by the U.S. Bureau of Labor Statistics. Annual averages are matched to the calendar year containing each effective date, and every real figure is expressed in 2025 dollars, the most recent complete annual average. Schedules effective after 2025 are compared at the 2025 index. No inflation figure here is estimated or interpolated.

Growth comparisons use the range each schedule published in the most years, named in the tables above, so the two ends of a comparison are the same rung rather than whichever rung happened to be lowest. ERG Z is left out of the frozen list because its ranges carry three different effective dates and one figure would misdescribe the table.

Found something that looks wrong? Tell us and we will check it against the source pages.