How Correctional Officers Get Paid: The Post Cannot Go ...

Seventeenth in a series on jobs whose pay system is stranger than the salary. Earlier entries covered firefighters, whose work week federal law redefines, nurses, whose wage is quietly set by a patient-safety rule, and police officers, the other uniform on the same public payroll. Correctional officers belong here for a reason that is nearly the opposite of the harbor pilots' privileged scarcity. Their pay engine is a scarcity that traps them, not one they profit from at leisure. Every figure below is cited to a state audit, a state comptroller's payroll data, a pension plan document, or a wage survey, and where a number could not be verified at a primary source I say so rather than estimating.

Most jobs let the work go undone when nobody shows up. If a store is short a cashier, the line gets longer. If an office is short an analyst, the report is late. The shift that nobody filled simply does not happen, and the payroll is smaller that day.

A prison cannot do that.

A correctional post has to be staffed every hour of every day, because the thing being controlled does not pause for a staffing shortage. So when a facility comes up short, the missing shift cannot go empty. Someone already on duty is held over and put onto a second shift back to back with the first. That is the whole mechanism, and everything else about how a correctional officer gets paid is downstream of it. In a chronically understaffed system, this forced overtime stops being an exception and becomes the actual pay engine. The overtime routinely rivals the base salary, and in the extreme cases it dwarfs it.

The wage the public sees is not the wage the system produces. Hold both numbers in view at once, because the gap between them is the entire story.

The survey number, and why it is not the paycheck

Start with what the federal wage survey reports, because it is the number that ends up in headlines and career guides.

In the May 2024 federal occupational wage data, correctional officers and jailers had a median wage of $57,970 a year and a mean of $62,760, across about 387,500 jobs nationally. The spread the survey publishes:

Percentile	Annual
10th	~$41,750
50th (median)	$57,970
75th	~$75,330
90th	~$93,000

One caveat on provenance up front, the same one that recurs across this series. The Bureau of Labor Statistics blocks automated access, and it blocked the proxy too, relaying its own "Access Denied" page instead of the data. So these figures are BLS's published numbers surfaced through search summaries rather than read at bls.gov directly, and I am labeling them secondary for that reason. The 25th-percentile figure was not among the values that surfaced, so I am not printing one. The median, mean, employment, and the tenth, seventy-fifth, and ninetieth percentiles are BLS's own numbers, just not pulled from the primary page.

Here is what the survey cannot show you. That $57,970 median describes base wage. It does not capture the held-over double shift, because a survey of typical wages is not built to measure a system running on mandatory overtime. In the states where the understaffing is worst, the officer's real annual pay is a base salary plus an overtime figure that can equal or exceed it. The median is not wrong. It is measuring the wage, and the wage is not what the job pays.

Connecticut, where the overtime beats the base

The clearest single case comes out of Connecticut, and it is worth walking through in detail because it is the whole thesis in one paystub.

In 2024, the Connecticut Department of Correction paid $110.8 million in correctional-officer overtime, up 30.5 percent from $84.9 million in 2015. That is the departmentwide figure. Now look at what it does to one person.

According to state comptroller figures reported by CT Insider, the state's highest-paid officer that year was Leroy Lee. His base salary was above $71,000. His overtime was above $187,000. His total compensation cleared $258,000. Read the two middle numbers again in order, because their sequence is the point: the overtime was more than two and a half times the base. To get there he worked 449 hours in January and 503 hours in March, with no days off in those stretches. A standard month at 40 hours a week is roughly 173 hours. He worked closer to three of those months inside one calendar month, twice.

He is the extreme, but the shape is systemic. The same reporting, drawn from records obtained under freedom-of-information law, documents officer Terry Castro at Bridgeport Correctional Center working 195 hours over 14 days in June. He crossed a full 40-hour week's worth of hours in his first three days, then kept going for nine consecutive days, several of them 16-hour shifts. That is not a person chasing overtime for a boat payment. That is a person being held on post because there was no one to relieve them.

The union says so plainly. AFSCME Council 4, which represents the officers, put it this way: officers are "ordered to work overtime under a threat of discipline or termination." This is the single most important word in the entire subject, and it is why "overtime" is a misleading label. On a normal job, overtime is something you accept or decline. Here it is mandatory. You do not volunteer for the second shift. You are told you are working it, and the alternative is losing the job.

And notice what did not cause it. The overtime rose while the prison population fell. Connecticut's officer headcount dropped from 3,936 in 2015 to 3,578 in 2024, about a 9 percent decline. Over the same years the prisoner count fell much faster, from 16,025 to 10,584, a 34 percent drop. Fewer prisoners, and still a record overtime bill. The overtime is not being driven by a surge in the work. It is being driven by the collision between a post that has to be covered every hour and a workforce the state let shrink.

Nevada, where a state audit shows the machine's gears

Connecticut shows the paystub. Nevada shows the machinery, because the state's own internal auditors took the whole system apart in a 2024 report and published the gears.

The Nevada Division of Internal Audits found that the Nevada Department of Corrections paid $50,426,271 in correctional-officer overtime in fiscal year 2024, across 890,701 overtime hours. That was not a stable cost. Overtime had grown by 507,513 hours, up 132.4 percent, and by $35,305,780, up 233.5 percent, in just four years from fiscal 2020 to fiscal 2024. The audit projected the full-year figure would reach $55.3 million.

The audit then did something a paystub never will. It went into individual timesheets and measured how concentrated the overtime was on individual bodies. On the timesheets it tested, 14 of them, one half of one percent, showed more than 100 overtime hours in a single pay period. Another 169, about 6.1 percent, showed 60 to 99 overtime hours in one pay period. A pay period is two weeks. One hundred overtime hours on top of the regular schedule inside two weeks is the arithmetic of a person who is barely leaving the building.

Here is the part that shows the mechanism is not an accident but a rule. Nevada's own internal policy caps a shift at no more than 16 hours in any 24-hour period and forbids more than two consecutive double shifts. Those are safety limits. But the labor agreement overrides them at the exact moment they would matter. The audit reports that under the collective bargaining framework, the department is "prohibited from relieving a covered employee from duty for an overtime shift in progress, regardless of reason." So once you are on the held-over shift, the state cannot take you off it, and the audit found at least one employee who blew past the two-consecutive-doubles limit anyway. The cap says one thing. The contract says the officer cannot be relieved. The contract wins, because the post still has to be covered.

The auditors were explicit about the cause. Facilities are understaffed, "contributing to an increased reliance on overtime." The state's proposed fix is the tell: the department told the Legislature that approved pay raises plus cost-of-living adjustments would "boost morale, help with retention and recruitment, and decrease vacancies, which would reduce mandatory overtime." Raise the base to fill the posts, so you stop paying the punitive overtime to cover the empty ones. It appears to be working at the margin, with departmentwide vacancies down 34.4 percent from fiscal 2023.

The audit also priced the waste. Better overtime and pay policies could save the state up to $18.5 million a year. Covering the posts left vacant while cadets sit in the academy costs about $5.46 million per academy session, roughly $11.0 million a year for two sessions. And more than 38,000 overtime hours, worth about $2.1 million, in fiscal 2024 had no supporting documentation at all.

One number in the Nevada record should not be mistaken for a wage, and I want to flag it directly so nobody repeats it wrong. A $169,600 figure that circulates alongside this audit is a consulting contract for a staffing study, funded through 2023 legislation. It is not an officer's salary. Do not read it as one.

Oklahoma, and the staffing floor that guarantees the overtime

Oklahoma shows the third corner of the same triangle, the staffing shortfall stated as a bare fraction.

An analysis of state payroll data by the Tulsa World found the Oklahoma Department of Corrections paid $35.57 million in overtime in 2025, which was about 48 percent of all overtime paid to every state employee in Oklahoma. One agency, nearly half the state's entire overtime bill. That was down slightly from $37.48 million in 2024, but it was up 361 percent from $7.7 million in 2010. As in Connecticut, some employees routinely logged more than 100 overtime hours in a two-week pay period.

The reason sits in a 2022 report from Oklahoma's Legislative Office of Fiscal Transparency, which found the department's facilities operating with roughly 40 to 45 percent of the staff they needed. Sit with that fraction. If a post requires a certain number of officers to run safely, and you have fewer than half of them, the remaining officers do not each do a normal shift. They do a shift and then most of another one, again and again, because the coverage requirement does not scale down with the roster. The staffing shortfall is the overtime bill. They are the same fact, measured two ways.

This is the inversion that makes correctional officers strange inside this series. The harbor pilots collect their extraordinary income from a scarcity a state board deliberately enforces on their behalf, a cap of sixty people on a job that prints money. The correctional officer's pay also comes from scarcity, but it is a scarcity of officers, and it does not enrich them at leisure. It chains them to the post. The pilot profits from the empty chairs on the roster. The officer is the one who has to sit in them, one after another, until the state finds someone else.

The public payroll versus the private one

There is one more comparison worth making, because it shows what the same job is worth when a company rather than a state signs the check.

Not every correctional officer works for a government. Some work for private prison operators, and the pay is measurably lower. Reporting drawn from BLS ownership detail put public-prison correctional officers at a median around $44,710, more than $5,600 above their private-prison counterparts. At the two largest operators, employer-reported figures put CoreCivic officers around $23 to $30 an hour, a median near $26, and GEO Group around $19 to $27, a median near $23.

I have to be careful here, and I would rather flag the softness than paper over it. The public-versus-private gap and the two hourly ranges come from secondary reporting, the Pulitzer Center on the wage gap and employer-review aggregators on the hourly figures, not from a primary BLS ownership-detail table I read directly. The Pulitzer Center page itself returned an access error on the attempt. A commonly cited claim that public prisons run about five inmates per officer against roughly seven in private facilities traces back to 2005 federal data and is stale, so I am not treating it as current. What survives all of that caveating is the direction: the same work pays less when the payer is a corporation optimizing costs than when it is a state absorbing them, which is a useful reminder that "correctional officer" is not one wage but two.

The part that is actually generous: the retirement

If the overtime is the punishing side of this job's pay, the pension is the compensating one, and it is genuinely valuable in a way the private sector abandoned decades ago.

Correctional officers are usually classified for retirement the way police and firefighters are, under enhanced or hazardous-duty rules that let them retire earlier than an ordinary worker on a richer formula. The clearest primary example is New York. Under the state comptroller's State Correction Officers plan, an officer with 25 years of creditable service can retire at any age with a full benefit equal to one-half, 50 percent, of final average salary. With 20 or more years, the benefit accrues at 2 percent of final average salary per year of service, and the whole thing caps at 60 percent of final average salary. Retire at any age once you have the years. That phrase does not exist in a normal private retirement account.

The federal and other state versions rhyme, though I read them at a lower confidence. Federal Bureau of Prisons officers are covered by law-enforcement retirement provisions that let them retire at age 50 after 20 years of covered service, or at any age with 25 years, under the enhanced hazardous-duty rules. New Hampshire's corrections benefits describe new hires eligible to retire by age 52.5 with 25 years of service, and add correctional-officer hazardous-duty pay of an extra $30 a week, about $1,560 a year. The federal and New Hampshire figures come from search summaries of the government pages rather than direct primary reads, so I am labeling them secondary while the New York plan is verified at the comptroller's own document.

There is a dark arithmetic hiding in the combination. A pension is calculated on final average salary, and final average salary in a mandatory-overtime system can be inflated by exactly the held-over shifts that are grinding the officer down. The forced overtime that runs the base pay up also, in many plans, runs the pension base up. The system that cannot staff its posts ends up paying for that failure twice, once in the overtime and again, for decades, in the retirement it inflated. Whether any given plan counts overtime toward final average salary varies and I am not asserting it for a specific system, but the incentive is worth naming.

What a salaried reader should take from this

When a shift cannot go unfilled, "overtime" stops meaning what you think it means. The defining fact of correctional pay is not a dollar figure. It is that the post has to be covered every hour, so a short-staffed shift is not skipped, it is forced onto whoever is already there. In a job built this way, overtime is not a choice you make for extra money, it is a duty the employer assigns you under threat of discipline. The same 7(k) overtime structure that lets fire and police departments redefine the week shows up here as its harsher cousin. Before you envy a big overtime number, ask whether the worker volunteered for it or was ordered onto it with no days off.

A staffing shortfall and an overtime bill are the same number wearing two faces. Oklahoma running its facilities at 40 to 45 percent of needed staff and Oklahoma paying 48 percent of the entire state's overtime are not two problems, they are one. When you see an organization bragging that it holds headcount flat while output holds up, look for where the missing labor went. Often it did not vanish. It got reassigned as mandatory overtime onto the people who remain, and it shows up as burnout and turnover before it shows up in the budget.

The valuable part of this job is the part a statute forces someone to fund. The overtime is coerced and the base is ordinary, but the pension is real, early, and rich, because a state law obligates the government to provide it under hazardous-duty rules. That is the recurring lesson of this whole series, from the harbor pilots' statutory pension to the firefighters' cancer-presumption laws: the durable money comes from a structural obligation somebody else has to honor, not from anything you negotiate. Judge these jobs on the total lifetime package, not the paystub, because the best part of the deal never appears on it.

The same title can be two different wages. A public correctional officer out-earns a private one doing identical work, because a state absorbs the cost while a company minimizes it. Whenever you compare offers by job title, first ask who signs the check and what that payer is optimizing for. "Correctional officer" is not one pay figure. It is at least two, and the wage survey's single median hides the split.

Fact-check notes and sources

Overtime totals and the timesheet detail come from a state internal audit and state comptroller payroll data. The pension formula comes from a state plan document. The wage figures come from the federal wage survey. Where a figure is secondary or could not be verified at a primary source, it is flagged in the text and here.

National correctional-officer employment and wages (median $57,970; mean $62,760; about 387,500 jobs, May 2024; and the tenth, seventy-fifth, and ninetieth percentiles at roughly $41,750, $75,330, and $93,000) are BLS Occupational Outlook Handbook and Occupational Employment and Wage Statistics figures for SOC 33-3012, surfaced through the BLS correctional-officers page. These are treated as secondary: bls.gov bot-blocks automated fetch and the proxy relayed the site's own "Access Denied" page, so the numbers were not read at the primary page. The 25th-percentile figure did not surface and is not asserted. The Handbook's exact wording on correctional-officer schedules and overtime could not be pulled from the primary source and is not quoted.
Nevada overtime totals, growth, timesheet concentration, shift caps, the CBA relief prohibition, understaffing, vacancy change, potential savings, academy-coverage cost, and undocumented overtime hours are all from the Nevada Division of Internal Audits, Report 25-04, Department of Corrections Correctional Officer Overtime Management: $50,426,271 and 890,701 overtime hours in FY2024; overtime up 507,513 hours (132.4 percent) and $35,305,780 (233.5 percent) from FY2020 to FY2024 with a $55.3 million full-year projection; 14 timesheets (0.5 percent) at 100-plus overtime hours in one pay period and 169 (6.1 percent) at 60 to 99; the 16-hours-in-24 and two-consecutive-doubles policy caps; the CBA guidance that the department is "prohibited from relieving a covered employee from duty for an overtime shift in progress, regardless of reason" and the one employee who exceeded the two-double limit; understaffing "contributing to an increased reliance on overtime" and the department's statement that raises plus COLAs would "reduce mandatory overtime"; vacancies down 34.4 percent from FY2023; up to $18.5 million in potential annual savings; roughly $5.46 million per academy session (about $11.0 million a year for two); and the 38,000-plus undocumented overtime hours worth about $2.1 million. The $169,600 figure associated with this record is a consulting contract for a staffing study funded via 2023 legislation, not an officer salary, and is not presented as a wage.
Connecticut's $110.8 million 2024 overtime bill (up 30.5 percent from $84.9 million in 2015); officer Leroy Lee (base above $71,000, overtime above $187,000, total above $258,000; 449 hours in January and 503 in March with no days off); officer Terry Castro (195 hours over 14 days in June, nine consecutive days, multiple 16-hour shifts); the staffing decline from 3,936 officers in 2015 to 3,578 in 2024 against a prisoner drop from 16,025 to 10,584; and the AFSCME Council 4 characterization that officers are "ordered to work overtime under a threat of discipline or termination" are from CT Insider / The Register Citizen reporting by Bill Cummings, citing state comptroller and Department of Correction data and FOIA records, as republished by Corrections1. Treated as secondary reporting of primary payroll data.
Oklahoma's $35.57 million in 2025 overtime (about 48 percent of all state-employee overtime; down from $37.48 million in 2024; up 361 percent from $7.7 million in 2010; some employees over 100 overtime hours per two-week period) is from a Tulsa World analysis of state payroll data, and the 40-to-45-percent facility staffing shortfall (2022) is from the Oklahoma Legislative Office of Fiscal Transparency, both as republished by Corrections1. Treated as secondary.
The public-versus-private pay comparison (public correctional officers around a $44,710 median, more than $5,600 above private; CoreCivic roughly $23 to $30 an hour, median near $26; GEO Group roughly $19 to $27, median near $23) is from Pulitzer Center reporting and employer-review aggregators. Secondary and not fetched at a primary BLS ownership-detail table; the Pulitzer Center page returned HTTP 403 on the attempt. The often-cited "about five inmates per officer in public versus roughly seven in private" ratio traces to 2005 federal data, is outdated, and is not asserted as current.
The New York State Correction Officers pension formula (25 years of creditable service allows retirement at any age with a full benefit of one-half, 50 percent, of final average salary; 20-plus years accrues at 2 percent of final average salary per year; maximum 60 percent) is from the New York State Comptroller's State Correction Officers plan page and is primary-verified.
Federal Bureau of Prisons law-enforcement retirement (retire at age 50 after 20 years of covered service, or any age with 25 years, under FERS enhanced/hazardous-duty provisions) is from the Bureau of Prisons careers page, and New Hampshire's corrections benefits (retire by age 52.5 with 25 years of service; hazardous-duty pay of an extra $30 a week, about $1,560 a year) are from the New Hampshire DOC benefits page. Both are secondary: taken from search summaries of the government pages rather than direct primary reads. Whether overtime counts toward final average salary is not asserted for any specific plan.

This post is informational and journalistic, not career, legal, or financial advice. It describes a published state internal audit, state comptroller and payroll data, a state pension plan document, and a federal wage survey. Overtime totals, staffing levels, wage figures, and pension terms change, and several figures are as of 2020 through 2025 as noted, so verify current status before relying on any of them. Mentions of specific states, agencies, unions, and private operators are nominative fair use, and no affiliation is implied.

How Correctional Officers Get Paid: The Post Cannot Go Empty, So the Overtime Is Mandatory