Crime and policing in Croton-on-Hudson

What the record shows, what the department actually does all day, what it costs measured against fifteen comparable villages, and whether it has room to absorb more work than it currently gets.

Croton-on-Hudson employs 22 sworn police officers for about 8,300 residents. This report sets out what they cost, what they are called to do, and how much capacity the department has — using the state's own crime records, the department's own monthly reports, and the annual financial filings every New York village submits to the State Comptroller.

1. The record

Index crime — the state's standard set of seven serious offences — is defined and published by the NYS Division of Criminal Justice Services, not by the village. Croton has reported a full twelve months in every year the state records that field.

Reported index crimes and arrests, 1990–2024. Source: NYS DCJS, Index Crimes by County and Agency.

Reported crime is roughly a third of its 1990 level. It is also about double its 2015 low, and the recent rise is property crime rather than violence. Both things are true and the second is the one more often left out.

Index crimes per 1,000 residents, 2024, against the other river villages. Dashed lines are the county (11.2) and state (21.4) rates.

Violent crime is the part worth separating out, because the index total is dominated by property offences and the two behave differently. Over the last ten years Croton has averaged 2.5 violent offences a year — murder, rape, robbery and aggravated assault combined — in a village of about 8,300. That is roughly 30 per 100,000 residents, against a national rate of 363.8 per 100,000 in the FBI's most recent published year. Croton's violent crime rate is about one-twelfth the national one. In 2024 the village recorded a single violent offence.

Property crime is where the recent movement is, and the gap there is narrower: a ten-year average of about 494 per 100,000 against a national 1,934, so roughly a quarter of the national rate. The counts are small enough that year-to-year swings should not be read as trends — the difference between four violent offences and none is two incidents, and both have occurred in the last decade.

Against the county and the state, Croton's rate is low. Against its immediate neighbours it is not the lowest — at 7.1 per 1,000 it is the second highest of the seven river villages charted, behind Dobbs Ferry. A department that under-reported would not produce that number.

2. What the department actually does

Crime counts are a poor description of police work. Croton's department publishes a monthly report — calls, arrests, cases, accidents, tickets, fire and ambulance assistance — signed by the Chief, going back to 2018.

Calls for service by month, from the department's own reports. Gaps are months the department did not publish.

In 2025 the department logged 4,410 calls for service and recorded 58 index crimes in the same year — roughly 76 calls for every recorded serious crime. Serious crime is a small fraction of what the department is called to.

Arrests, cases investigated, motor-vehicle accidents and fire/ambulance assistance, by year. Only years with twelve published months are comparable.

Comparing the same five months each year — the only honest way to use a partial current year — calls are up about 18% on the 2018–2023 average. The fastest-growing category is not crime: it is fire and ambulance assistance.

3. What it costs

What Croton spends, and on what

Start with the village's own budget, because that is the document a resident can pull up. The figure usually called "the police budget" is the police-department line alone. It is not what the department costs: the pension, the health insurance, the civilian staff and the capital programme are all budgeted elsewhere in the same book.

Where Croton's own police money goes

The comparison above puts fifteen villages on one rule. This is the other view: Croton's own adopted budget for 2026–27, broken into what the money actually buys — including the capital programme, which sits in a separate fund and is easy to leave out of a "police budget" figure.

Croton-on-Hudson adopted budget 2026–27. The centre figure is the complete total: $8.4 million — $7.4 million of operating spending plus a $1,050,937 capital programme. The health, payroll-tax and compensation slice is an estimate: those benefits are pooled centrally for the whole village and no filing splits them by department, so it is apportioned on police's share of payroll. Every other slice is a figure printed in the budget.

The single largest thing the department costs is people, and the second largest is the pension and insurance that follow from employing them. Together payroll and benefits are about 85% of the complete figure. Equipment, supplies and the whole capital programme together are the remainder.

Why these figures had to be rebuilt

How that compares with fifteen neighbours

Every component of police spending, per resident, fiscal 2025, fifteen villages — capital included, unlike the operating-only figure used for year-to-year comparison. Capital is a five-year average because it is lumpy. Tarrytown and Dobbs Ferry show no capital segment because they run essentially all police capital through their operating budgets — it is already inside their other segments. Elmsford is absent: it has filed no annual financial report since fiscal 2017.

Croton totals $875 per resident against a fifteen-village median of $939.[1] The chart makes the shape of the difference visible: almost all of the spread between villages is payroll and the benefits that scale with it.

What has been normalised out, and it is marked on the chart. Two villages file spending under a police category that is not policing, and both are now removed rather than merely footnoted: Pelham −$15 per resident (code-enforcement officers), and Tuckahoe −$30 per resident (parking-lot maintenance labour). The dark red tick on those two bars marks where each village's figure sat as filed, so the size of the adjustment is visible rather than taken on trust.

[2]

View as table

Village	Payroll	Other op.	Benefits	Capital	Total
Pelham Manor	$831	$51	$521	$3	$1406
Ardsley	$731	$57	$501	$9	$1297
Irvington	$677	$53	$464	$15	$1209
Larchmont	$665	$35	$411	$45	$1155
Bronxville	$565	$91	$384	$80	$1120
Tarrytown	$559	$63	$398	$0	$1019
Tuckahoe	$520	$82	$367	$36	$1006
Pleasantville	$532	$31	$354	$22	$939
Pelham	$494	$38	$360	$26	$917
Sleepy Hollow	$513	$35	$335	$24	$907
Croton-on-Hudson	$495	$36	$306	$38	$875
Briarcliff Manor	$442	$27	$355	$17	$842
Rye Brook	$464	$25	$309	$26	$825
Dobbs Ferry	$408	$103	$271	$0	$783
Hastings-on-Hudson	$367	$28	$281	$39	$715

[3]

What the comparison actually shows

The spread is not really about what an officer costs. Across the fifteen villages, officers per 1,000 residents correlates with cost per resident at +0.96; cost per officer correlates at +0.06. An officer costs $294,000–$356,000 almost everywhere. Police density runs from 2.21 per 1,000 in Hastings-on-Hudson to 4.69 in Pelham Manor. Croton is at 2.52, third-thinnest of the fifteen. This is a measure of how heavily a village polices itself, not of how efficiently it does so.

4. Does the department have capacity?

There is a published professional standard for this, used in staffing studies nationally: the saturation index. It asks what share of available patrol time is consumed by answering calls. Below about 60%, patrol has slack — time for proactive work, and headroom to absorb surges. Above it, officers are doing little but moving from call to call.

Croton's shift floor, as the Village states it, is 3 officers on duty at all times — two officers and a sergeant. That is 26,280 officer-hours a year of patrol capacity at minimum staffing. Against 4,410 calls in 2025:

Saturation index = calls × officer-hours per call ÷ available patrol hours. "Headroom" is how many times current call volume could rise before patrol reaches 60%.

Croton would need calls averaging more than 3.58 officer-hours each — well above anything in the published literature, where typical figures run between half an hour and an hour and a half — before patrol crossed the 60% line at current volume.

What that means, on our reading

On any plausible assumption about how long a call takes, Croton's patrol force is not saturated. At one officer-hour per call the index is about 16.8%; at an hour and a half, about 25.2%. Both sit far below the 60% threshold. On the same arithmetic, call volume could roughly 2.4× before patrol reached that line at an hour and a half per call.

Two consequences follow, and they point the same way. The department has had, and still has, sufficient time per call — the constraint on it is not that officers are run off their feet answering the phone. And the growth in workload it has absorbed since 2018, roughly 18% on a matched basis, is well inside what the current establishment can carry: the village could have grown without strain at this staffing level.

What the overtime looks like

If a department were stretched by demand, its overtime should show up in the work it chooses to do. Croton's does not. Overtime is 13.6% of police payroll — $528,000 against $3,365,885 of base pay — and it is split like this:

Two-thirds of it — $350,000, or 66% — sits in a single general line that names no activity at all. The named activities the department runs by choice — police investigations, police community & youth, patrol boat, bicycle patrol — come to $128,000 between them, about 24% of overtime and under 3.3% of police payroll.

That shape is what a coverage floor looks like. When a shift must not fall below two officers and a sergeant, and there are twelve patrol officers to draw on, every illness, court date, vacation and training day has to be backfilled at premium rates whether or not anything is happening in the village. It is consistent with the saturation finding above: the department is not busy in the sense of being run off its feet by calls, but it is committed — the roster, not the workload, is what costs the money.

What actually binds a department this size is not call volume but the round-the-clock coverage floor. Two officers and a sergeant must be on duty at 3am whether or not anything happens, and covering every shift of every day with rotation and leave takes five squads. That floor is reached long before workload becomes the limiting factor — which is also why per-resident staffing league tables mislead for villages this small, and why the professional literature advises against using them for staffing decisions.

5. What we could not settle

Croton's own police borrowing is known; its neighbours' is not. Croton's adopted capital plan names a funding source for every project, so what the village intends to borrow for policing is a matter of public record: $973,998 of the 2026–27 police capital programme is debt-funded — the digital radio system ($666,250), police vehicles ($175,275), the staffing study ($59,450), IT upgrades ($46,373) and equipment ($26,650) — with body cameras ($76,939) funded by grant instead. The plan plus its four out-years carries $1,353,998 of debt-funded police capital in total, mostly a vehicle replacement cycle of $92,000–$98,000 a year.[5]

We could also obtain the equivalent for three neighbouring villages, by reading their own bond offering documents. Those state, project by project, what each village has authorised itself to borrow and for what:

Two villages cannot be placed at all: Elmsford has not filed an annual financial report with the state since fiscal 2017, and Sleepy Hollow's most recent police personnel return is 2022, which means that figure sits inside the staffing median we compare Croton against.[7]

Briarcliff Manor, the closest case

Briarcliff Manor is the village this argument has centred on. Its 2020 bond issue financed $124,083 of police capital against about $2,258,033 of fire and ambulance capital — two fire engines, an ambulance, a chief's car, turnout gear, a defibrillator. Roughly eighteen dollars for fire and ambulance for every one for police.[4]

The village most often cited as spending more than Croton is, on its own capital documents, overwhelmingly equipping a fire and ambulance service rather than a police department.

Police is a small share of what villages borrow for — between nothing and about 5% among the three we could check. Croton's 12.8% looks unlike them, and the reason is a single line: the $666,250 digital radio system, a generational replacement rather than a recurring cost. Take it out and police is $384,687, or 4.7% of the plan — level with Hastings' 4.8%, the highest of the three, and well above Ardsley's 1.3%. What villages mostly borrow for is roads, parks, buildings and fire apparatus; Dobbs Ferry's largest single authorisation is a $770,000 fire vehicle, and it has no police borrowing outstanding at all.[6]

6. A note on what is on the record

One theme runs through the checking. Almost everything in this report was reconstructable because Croton publishes the underlying documents, and several of the things we could not establish were blocked by the absence of an equivalent somewhere else.

The department publishes a monthly report — calls, arrests, cases, accidents, tickets — signed by the Chief, going back to 2018. Of the peer departments whose pages we could reach, only Ardsley publishes anything comparable. Croton also files its civilian police staff under a separate account that just six villages in the state use, which has the effect of making costs visible that are folded invisibly into other departments elsewhere. Its capital plan names a funding source for every project, which is why its police borrowing can be stated exactly and its neighbours' mostly cannot. Reported crime is filed in full every year the state records the field.

Against that, the document at the centre of this whole argument — the one-page comparison of sixteen departments — names no fiscal year, no source, and no definition of what "budget" includes. It is the least documented thing on the department's reports page, and it is the thing that has been quoted most.

There is a practical cost to that asymmetry. Two of the corrections in this report exist only because a village published enough detail to catch them: Briarcliff's budget line, which showed the comparison table had copied a figure that included pension and health insurance, and Croton's own budget lines, which showed our roster table had been understating overtime by $23,000. Where a village publishes less, we have fewer findings about it — not because there is less to find.

Sources: NYS Division of Criminal Justice Services (index crime by agency); Croton-on-Hudson Police Department monthly reports; NYS Office of the State Comptroller, account-level annual financial reports for villages, fiscal 2025; NYS DCJS Law Enforcement Personnel by Agency; US Census. Method and every correction we have made are recorded in full in our internal research file.