Is New York Safer Than Ever in 2026? Crime, Policing, and the Preparedness Gap

The bottom line: New York is safer by several major measures—but “safer” is not the same as “safe everywhere”

Through July 2026, New York City recorded its fewest murders, shooting incidents, and shooting victims for the first seven months of any year in the NYPD’s recorded history. Major crime fell 6.6%. Those are real gains. Yet felony assault was almost flat, reported rape rose under a broadened definition and with more delayed reports, transit crime rose slightly, and the Preliminary Fiscal 2026 Mayor’s Management Report—covering July through October 2025—reported a 9-minute-54-second end-to-end average for critical crimes in progress, measured from completion of dialing 911 to officer arrival. The responsible conclusion is confidence without complacency: support effective policing, demand honest policy evaluation, and build a layered personal-safety plan centered on awareness, avoidance, communication, first aid, secure storage, and lawful last-resort defense.

Begin with agreement

What “safer than ever” actually means

When officials say New York City is experiencing historic public-safety gains, they are not inventing the central evidence. The NYPD’s August 3, 2026 release reports record-low murders, shooting incidents, and shooting victims for the first seven months of the year. Any honest analysis should celebrate fewer dead and wounded New Yorkers. The NYPD launched CompStat in 1994; this article uses “recorded history” as the department’s term for its modern statistical series, not as a claim that every pre-1994 count is directly comparable without methodology caveats.

The problem begins when a precise claim becomes a universal slogan. “Record-low murders for this seven-month period in the NYPD’s recorded history” is measurable. “Every New Yorker is safer than ever” is not. A citywide total cannot tell a particular person the risk on a particular block, during a particular hour, or inside a particular relationship. It cannot erase nonfatal assault, harassment, disorder, unreported victimization, repeated exposure in a high-risk location, or the minutes before help arrives.

The right comparison is not “statistics or lived experience.” Both matter, but they answer different questions. Statistics estimate broad patterns. Lived experience identifies conditions, locations, and harms that a citywide average can obscure. Neither should be used to cancel the other.

The strongest current evidence

New York City crime in 2026: record lows and warning lights in the same report

The following figures compare January through July 2026 with the same seven months of 2025. They are preliminary NYPD counts, not final audited totals. Percentage changes below are the department’s published calculations.

Source and denominator note: NYPD press release dated August 3, 2026. Counts cover reported incidents in New York City during the first seven months of each year. They are not rates per resident or per trip, and current-year figures are preliminary.

The rape number requires care—not minimization

The NYPD reported that 28% of rapes reported in 2026 concerned conduct from prior years and 20% fell under the expanded definition created by the Rape is Rape Act, effective in September 2024. Those facts make simple year-to-year comparison difficult. They do not make the reports less serious. Better legal coverage and greater willingness to report can increase the measured count while improving recognition of victimization.

Public housing shows why one-month headlines can mislead

Public housing recorded its safest first seven months in the NYPD series for murders, shootings, shooting victims, and robberies. Yet public-housing major crime rose 3.5% in July. Both statements can be true because one describes seven months and the other one month. Honest public-safety writing always names the time window.

New York is more than New York City

Albany and the 28 GIVE departments: lower long-term violence, uneven local movement

New York State’s Gun Involved Violence Elimination initiative tracks shootings involving injury in 28 participating police departments outside New York City. The July 2026 monthly report, with data through June, shows strong regional declines: 197 shooting incidents involving injury compared with 244 during the same months of 2025, a 19% reduction; 223 victims compared with 279, a 20% reduction; and 26 people killed compared with 50, a 48% reduction.

Measurement boundary: GIVE counts shooting incidents involving injury outside New York City; NYPD reports its own New York City shooting-incident series. The two systems are kept separate here, and their rates should not be compared directly.

Albany therefore illustrates the correct level of nuance. The city’s first-half shooting incidents rose by one from 2025, victims were unchanged, and fatalities fell. At the same time, incidents and victims remained roughly half their recent five-year averages. A single percentage without the raw count and longer baseline would distort the picture in either direction.

The City of Albany separately reports that violent crime fell 11% in 2025 and property crime reached its lowest level since 2020, while also acknowledging that its overall crime rate remains above the national average. Long-term statewide context is encouraging: DCJS reports New York’s 2024 index-crime rate was 63% below its 1990 level. That long decline is real; it does not eliminate the need to prepare for low-probability, high-consequence events.

What police statistics do—and do not—capture

Police data count events known to police under defined categories. They are essential, but they are not a census of every harmful experience. The national Bureau of Justice Statistics victimization survey estimated 23.3 violent victimizations per 1,000 people age 12 or older in 2024, while the rate reported to police was 11.2 per 1,000. Those are national survey estimates—not New York counts—and that report does not publish an NYC-only victimization rate. They demonstrate why reported crime and experienced victimization are different measures without supplying a New York City estimate.

Other limits matter:

• Counts are not individual probabilities. A person’s exposure varies by location, routine, occupation, transportation, household, and relationship circumstances.
• Categories change. A broader statutory definition can increase a count without proving that underlying conduct suddenly became more common.
• Small numbers swing sharply. A change of one event can produce a dramatic percentage in a small jurisdiction.
• Crime data do not measure response gaps. An incident may be rare and still unfold faster than a caller can obtain help.
• Recorded crime is not perceived order. Disorder, threatening behavior, and repeated low-level exposure can affect public conduct even when they do not enter the major-crime index.

Viral graphic versus current evidence

Did a decade of laws “tie cops’ hands”? A claim-by-claim fact-check

The phrase is political shorthand, not a measurable variable. Different laws changed different stages of the system: police encounters, arrest processing, pretrial release, evidence disclosure, case readiness, and dismissal. Treating all of them as one policy—or every dismissal as a crime caused—is analytically weak.

The more defensible argument is narrower: some reforms created real workload and case-processing costs; some produced due-process and transparency benefits; some early claims relied on incomplete data; and policymakers should measure whether the combined system can protect rights, prepare cases, supervise higher-risk defendants, respond promptly, and sustain public trust.

Bail reform: rearrest is real, but the causal claim is harder

New York’s official 2019–2024 pretrial tables show that among 115,409 New York City Criminal Court releases in 2024, 25% had any rearrest within 180 days, 4% had a violent-felony rearrest, 11% a nonviolent-felony rearrest, and 9% a misdemeanor rearrest. Those figures should not be minimized.

They also should not be mislabeled. A rearrest is an allegation following a new arrest, not a conviction. The table does not tell us that bail reform caused the alleged conduct, that detention would have lawfully prevented it, or that populations released in different years had the same risk profiles. It also combines release outcomes with very different baseline risks.

A 2025 Data Collaborative for Justice quasi-experimental evaluation used matched comparison groups over 50 months. For cases made ineligible for bail in New York City, the reform group had lower overall rearrest (57% versus 66%), felony rearrest (33% versus 40%), and violent-felony rearrest (20% versus 25%). Effects in suburban and upstate regions were neutral. Reduced use of bail in eligible NYC cases was associated with modest reductions in overall and felony rearrest and no change in violent-felony rearrest.

This does not prove every aspect of bail reform works perfectly. The same research program has found that effects can vary in smaller, higher-risk subgroups. It does show why a viral before-and-after percentage cannot carry the weight of a causal conclusion.

Discovery reform: the strongest “system strain” evidence—but not the whole story

Discovery reform required prosecutors to disclose evidence earlier and more comprehensively. The due-process case is powerful: a person should not have to consider a plea without timely access to the evidence. Defense providers reported earlier, fuller information and better-informed case discussions.

The implementation problem is also well documented. Drawing on the state court dashboard, the CUNY Institute for State & Local Governance reported that speedy-trial dismissals of NYC misdemeanor arraignments rose from 9% in 2019 to 48% in 2024. The dashboard category is speedy-trial dismissal; it does not independently code every dismissal as caused by a particular missing discovery item. The researchers interpret the trend alongside documented discovery-compliance failures, but the 48% includes all speedy-trial dismissals. The comparable change was 1% to 5% in downstate suburbs and 2% to 4% upstate, with the upstate movement disappearing when two outlier counties were excluded. For the most serious indicted felonies—not felony cases still in Criminal Court—speedy-trial dismissals stayed below 1% in every year and region.

That pattern matters because it rejects two easy narratives. It is not credible to say discovery caused no serious operational problem in New York City. It is also not credible to imply that half of violent felony prosecutions were dismissed. The sharpest effect was concentrated in city misdemeanors, while prosecutors, police, defenders, courts, and technology systems faced different conditions across the state.

The legal baseline also changed. Amendments announced by the Governor took effect on August 7, 2025, altering compliance and dismissal procedures while retaining the early-disclosure framework. A graphic that ends with 2024 data is useful history, not a complete statement of 2026 law or results. The New York Courts discovery dashboard and annual reports should be the update point.

How Many Stops: capacity cost and accountability benefit can coexist

The How Many Stops Act requires NYPD reporting on Level 1, Level 2, and Level 3 investigative encounters. The NYPD now publishes quarterly encounter files, including first-quarter 2026 data.

The Thurgood Marshall Institute’s first-year research brief, analyzing NYPD files from July 2024 through June 2025, found 2.7 million reported encounters—more than 7,000 per day. Nearly 98% were Level 1 encounters, which do not require suspicion of criminal activity. The study reported major racial disparities and explained that prior public reporting of Level 3 stops showed only a small fraction of police-civilian contacts.

Police leadership has described substantial reporting and overtime demands. Those claims deserve an independent, apples-to-apples operational audit: time per report, duplication with existing records, officer and supervisor hours, technology cost, error rates, effect on patrol availability, and whether forms can be simplified without destroying the information’s value.

The wrong choice is between blind bureaucracy and invisible policing. The better objective is minimum necessary reporting that yields auditable public value. If a field never informs supervision, legality review, resource allocation, or public understanding, remove it. If the data expose disparities or unlawful patterns, preserve and improve it. Accountability should be designed as an operational system, not merely added as paperwork.

The strongest opposing case

What criminal-justice reform and gun-violence-prevention advocates get right

A durable preparedness argument must survive the strongest objections, not the weakest caricatures.

The gap nobody can legislate away

Police can be effective and still be minutes away

The latest Preliminary Fiscal 2026 Mayor’s Management Report measures end-to-end response from the completion of dialing 911 to officer arrival. During July through October 2025—the first four months of fiscal 2026—the average was 9:54 for critical crimes in progress, 13:14 for serious crimes, 15:54 for all crimes in progress, and 32:14 for noncritical crimes (NYPD discussion and table, pp. 15 and 17). These are preliminary fiscal-period averages, not guarantees for any call and not calendar-2026 year-to-date figures.

Measure definition: “End-to-end” includes call-processing time from completion of the 911 dial, dispatch, and travel through officer arrival. It is broader than the report’s dispatch-and-travel-only measure, which was 7:32 for critical crimes over the same period.

These averages are not an indictment of police. Officers cannot arrive before a call is made, and dispatchers cannot eliminate distance, traffic, simultaneous emergencies, scene uncertainty, or finite staffing. The federal Cybersecurity and Infrastructure Security Agency makes the same practical point in its active-shooter preparedness guidance: such incidents often end within 10 to 15 minutes, sometimes before law enforcement arrives.

That interval is the preparedness gap. Most people will never need force during it. Many will benefit from a locked door, an exit route, a charged phone, a clear address, a family code word, bleeding-control skills, or the discipline not to enter an avoidable conflict.

Original NY Safe framework

The Preparedness Ladder: eight layers before last-resort force

The ladder is ordered by how broadly useful each layer is. It is not a legal test and does not require a person to complete each rung during an unfolding emergency. Its purpose is to build safety before stress collapses time and choices.

1. Layer 1 · Information

   Know the actual pattern

   Use current local reports, building alerts, transit information, and verified warnings. Do not turn a viral video or a stranger’s appearance into a threat assessment.

2. Layer 2 · Avoidance

   Leave early, not dramatically

   Change route, create distance, decline arguments, move toward light and people, and trust specific behavior-based warning signs without profiling a category of person.

3. Layer 3 · Environment

   Use locks, lighting, barriers, and exits

   Reinforce doors, trim concealment, test alarms, identify two exits, keep vehicle doors locked, and choose a family rally point. Prevention works while nobody is afraid.

4. Layer 4 · Communication

   Call early and communicate clearly

   Know your location, describe conduct rather than conclusions, identify clothing and direction, say whether weapons are seen, and follow dispatcher instructions. Teach children how and when to call 911.

5. Layer 5 · Medical readiness

   Prepare to keep someone alive

   Learn CPR and bleeding control, maintain an accessible trauma kit, and know medical conditions in the household. The most valuable act after violence may be care, not confrontation.

6. Layer 6 · De-escalation

   Protect the mission from ego

   Use calm words, nonthreatening posture, distance, and clear withdrawal. Do not teach strangers lessons. Do not turn property, pride, or disrespect into a life-or-death event.

7. Layer 7 · Non-firearm options

   Match lawful tools to ability and context

   Consider alarms, flashlights, communication tools, and lawful defensive options after checking current state and local restrictions. Equipment without judgment can create false confidence.

8. Layer 8 · Last-resort defense

   Force only within law, necessity, and identification

   A firearm is not the plan and carrying is not mandatory. Any defensive force must be grounded in the facts, proportional to the threat, directed only at an identified imminent danger, and stopped when necessity ends. New York’s retreat, initial-aggressor, and reasonableness rules matter.

The anti-vigilante line

Why Bernie Goetz is a warning, not a model

The 1984 New York City subway shooting is often reduced to a cultural symbol. The more useful lesson comes from the controlling legal decision. In People v. Goetz, 68 N.Y.2d 96 (1986), the Court of Appeals held that justification contains both a subjective and objective component. The person must actually hold the claimed belief, and that belief must be reasonable in light of the circumstances. Fear by itself does not supply unlimited authority.

Current Penal Law § 35.15 generally permits physical force when and to the extent a person reasonably believes it necessary against the use or imminent use of unlawful physical force. It limits justification for provocation, initial aggression, and combat by agreement. Deadly physical force has additional requirements, including New York’s duty to retreat when the person knows complete personal safety is available, subject to statutory exceptions.

If a firearm is part of the plan, storage and suitability come first

A person who is unwilling to secure a firearm, train, manage anger, avoid alcohol or impairment while armed, or step away during a mental-health crisis is not prepared merely because that person owns equipment. The decision not to own or carry a gun can be responsible. So can temporarily not carrying when stress, medication, sleep loss, grief, relationship conflict, or impaired judgment changes the risk.

New York Penal Law § 265.45 requires qualifying safe storage where a person under 18 or a prohibited person resides. It also requires an unattended firearm in a vehicle to be unloaded, locked in an appropriate safe-storage depository, and out of sight; a glove box does not qualify. Local rules may be more restrictive.

The Johns Hopkins Center for Gun Violence Solutions—an organization that advocates stronger gun regulation—makes an important point gun owners should adopt rather than resist: home defense and safe storage are not mutually exclusive. A quality quick-access safe can restrict children, visitors, people in crisis, and thieves while preserving controlled access for an authorized adult. Johns Hopkins also reported in a 2018 national survey of gun owners that respondents whose storage decisions were influenced by home-defense concerns were 30% less likely to store all firearms safely—an association, not proof that the motivation caused the behavior.

If someone in the household is experiencing suicidal thoughts, an acute mental-health crisis, or substance impairment, reduce firearm access immediately and seek professional help. Lawful temporary off-site storage may be appropriate, but New York licensing and transfer rules can affect who may receive a firearm and how; verify the method before moving it.

Review NY Safe Inc.’s New York safe firearm storage guide and adapt the plan to the people who actually enter the home—not the ideal household imagined when the safe was purchased.

Turn concern into capability

A seven-day personal-safety reset

1. Day 1—Map: identify two exits from home and work, one family rally point, and the precise address or location information a dispatcher would need.
2. Day 2—Harden: test locks, exterior lighting, cameras, alarms, intercoms, and visibility around doors and vehicles.
3. Day 3—Communicate: choose a family code word, update emergency contacts, charge backup batteries, and teach age-appropriate 911 use.
4. Day 4—Learn care: enroll in CPR and bleeding-control training; inspect first-aid and trauma supplies.
5. Day 5—Rehearse restraint: practice simple exit language—“I don’t want a problem,” “I’m leaving,” “Call 911”—without insults, threats, or competing for the last word.
6. Day 6—Audit tools: check batteries, access, legality, storage, and whether every person who might encounter a defensive tool knows the rules.
7. Day 7—Stress-test the plan: ask what happens if the phone is dead, one exit is blocked, a child is in another room, a visitor accesses storage, or police arrive while facts remain unclear.

Preparedness should make daily life calmer. If a plan makes you scan every stranger as an enemy, seek confrontation, or imagine yourself as law enforcement, the plan is failing.

Frequently asked questions about New York crime and preparedness in 2026

Continue with the question you actually need answered

Further reading: law, response time, judgment, and storage

Methodology and primary source library

Method: NY Safe Inc. reviewed the latest available official crime, court, response-time, and statutory sources as of August 18, 2026. We preserved each source’s time period and denominator, treated current-year data as preliminary, separated arrests from convictions, and used causal language only where a research design attempted causal inference. We included contrary evidence from criminal-justice-reform and gun-violence-prevention sources. GIVE injury shootings and NYPD shooting incidents are treated as separate reporting systems and are not combined or directly rate-compared. No attempt was made to combine incompatible police counts and national victimization-survey estimates into one New York City risk rate.

• NYPD — first seven months of 2026 crime release
• NYPD — current citywide crime statistics
• NYPD — CompStat history and statistical context
• NY DCJS — July 2026 GIVE monthly shooting report
• NY DCJS — New York criminal-justice footprint
• NYC Mayor’s Office — Preliminary Fiscal 2026 Mayor’s Management Report
• Bureau of Justice Statistics — Criminal Victimization, 2024
• NY DCJS — Supplemental Pretrial Release Summary Tables, 2019–2024
• Data Collaborative for Justice — long-term bail-reform evaluation
• New York Courts — discovery dashboard and annual reports
• CUNY ISLG — discovery implementation and dismissal analysis
• New York Governor — 2025 discovery amendments taking effect
• NYPD — investigative-encounter data
• Thurgood Marshall Institute — first-year How Many Stops analysis
• New York Penal Law § 35.15 — justification
• People v. Goetz, 68 N.Y.2d 96 (1986)
• New York Penal Law § 265.45 — safe storage
• CISA — active-shooter preparedness
• Johns Hopkins — safe storage and home-defense compatibility
• Johns Hopkins — national gun-owner storage survey

Start with judgment before deciding whether a firearm belongs in your plan

NY Safe Inc.’s personal-protection webinar addresses New York law, realistic risk, avoidance, alternatives, and the responsibilities of firearm ownership without assuming that every attendee should carry. For the deeper decision standard, read New York Concealed Carry: 7 Critical Judgment Tests.

For journalists, researchers, and educators

Quote, verify, challenge, and link to the exact section

Brief quotations from this article and the NY Safe Preparedness Ladder may be reused with attribution to Peter Ticali, founder and lead instructor of NY Safe Inc., and a link to the published article. Please preserve the article’s central qualification: falling citywide crime and continuing individual preparedness needs can both be true.

Interview angles: how to discuss falling crime without dismissing victims; how response-time data should shape family planning; why responsible firearms training begins with avoidance; where reform critiques are supported and overstated; how secure storage fits both public-health and home-defense goals; and how journalists can distinguish rearrest, dismissal, acquittal, and conviction.

Attribution does not imply NY Safe Inc. endorses a publication or its coverage. For substantial republication, graphics licensing, data questions, or an interview, contact NY Safe Inc. through its verified website.

Experience, instruction, and public-safety education

About Peter Ticali: the experience behind this guidance

Peter Ticali is the founder and lead instructor of NY Safe Inc. His work combines firearms instruction, family-safety education, medical-response training, multi-jurisdiction licensing experience, and sustained study of New York firearms law and Article 35.

NRA Endowment Life Member · NRA & USCCA Certified Instructor · Licensed Firearms Instructor: NY, MD, DC, MA, UT · NY Pistol License Holder Since 1992.

Peter is an NRA Chief Range Safety Officer, an American Heart Association Basic Life Support Instructor, a graduate of the FBI Citizens Academy and Suffolk County Police Department Citizens Academy, and an instructor in safety programs that address crime prevention, children’s firearms safety, active-shooter response, first aid, and trauma response. Citizens-academy attendance does not confer police employment, certification, authority, or agency endorsement.

Credentials do not substitute for evidence. Their relevance is narrower: Peter’s training philosophy treats avoidance, de-escalation, secure storage, emergency communication, medical readiness, lawful decision-making, and restraint as one civilian safety system.

Review Peter’s complete credentials, licenses, community work, and media background.