Whose safety, and where.

I should begin with something I cannot claim as insight, only as a fact about myself. As a man, I walk through this district at night without thinking about it. The question of whether I will get home does not occur to me. That is not a virtue and it is not courage. It is a condition I was handed, and it has shaped what I assumed public safety policy was for.

It is not the general experience. Gallup’s most recent measure found forty percent of Americans afraid to walk alone at night within a mile of home — the highest reading in three decades. Broken out, that is fifty-three percent of women and twenty-six percent of men.¹ The gap is not new and it is not narrow: in 2010 the figures were fifty percent and twenty-two percent, and Gallup has found a comparable gap internationally.² Fear also tracks income and geography, running higher among households under forty thousand dollars and among city residents.³

Gallup has also observed that women are about as likely as men to be crime victims.² I do not read that as evidence that the fear is misplaced. I read it as evidence that fear itself is a cost — a tax on movement, paid nightly, by half the population and not by me.

So the first proposition of this piece is that public safety is not one condition. It is at least two, and they call for different instruments.

Safety in public space

What the State and the counties can do about a street is mostly physical, and the evidence is better than the debate usually suggests.

Light works, and we know it works. In 2016 New York City ran a randomized controlled trial — the first of its kind — placing temporary lighting in roughly half of eighty high-crime public housing developments chosen at random.⁴ Nighttime outdoor index crimes fell by about sixty percent at the treated sites, and by at least thirty-six percent after accounting for the possibility that crime simply moved a few blocks.⁵ The researchers found the mechanism was deterrence rather than arrest: crime went down and so did arrests.⁶ That last detail matters. This is a public safety intervention that does not expand the reach of the criminal justice system, and about two-thirds of residents surveyed felt favorably about it.⁷

Kalihi has streets where the lighting has not been adequate in my lifetime. This is not a hard problem. It is a budget line.

People work too. A street with neighbors on it — someone walking a dog, a store with a lit window, a bench where kūpuna sit in the afternoon — is a street with witnesses. This is the point at which public safety and land use turn out to be the same subject. A district zoned so that no one has any reason to be on the sidewalk between the hours of six and eight is a district that has legislated away its own supervision. I have written separately about what our zoning did to neighborhood commerce; the safety consequence belongs here. Eyes on the street are not a slogan. They are a design outcome, and we designed them out.

And the ordinary infrastructure. Sidewalks wide enough to use, crossings where people actually cross, sightlines not blocked by parked vehicles, bus stops that are lit and visible. None of this is exotic. Most of it is deferred maintenance with a safety justification attached.

What we ask of police, and what we should stop asking

I want a well-funded, well-trained, and responsive police department. I also think we spend a substantial share of sworn officer time on work that does not require a sworn officer, and that this is a disservice to officers and to the public at once.

The research on what police actually do is consistent across jurisdictions. A study of 4.3 million 911 calls across nine agencies found traffic-related calls were the single largest category at 16.8 percent, followed by disorder at 16.2 percent and suspicion at 12.8 percent; calls concerning violent incidents were 6.4 percent.⁸ A detailed analysis of Detroit found that only 44 percent of emergency calls for service related to crime at all, and that crime incidents consumed 51 percent of on-scene time — with traffic duties taking 16 percent, quality-of-life incidents 15 percent, and health calls 7 percent.⁹ A separate analysis across eight cities found that less than thirty percent of calls for service were related to any type of crime, and 1.4 percent to violent crime.¹⁰

Set beside that, consider what we know works. A randomized trial of foot patrol in hot spots found that even minimal patrol time — fifteen minutes a day — reduced serious violence across a large area, with no evidence of displacement into surrounding blocks and a cumulative benefit from patrolling on consecutive days.¹¹

The conclusion I draw is not that we need fewer officers. It is that we are spending a scarce and expensive resource on the wrong tasks. An officer standing in traffic, chalking tires, or taking a report on an abandoned vehicle is an officer not walking a hot spot, not investigating a burglary series, and not available when a domestic violence call comes in.

So: civilianize the work that does not require a badge. Parking enforcement and abandoned-vehicle response are the clearest candidates — they are enforcement functions, not policing functions, and every jurisdiction that has separated them has kept the enforcement and freed the officer. Non-injury crash reporting, traffic direction at planned events, and routine report-taking belong in the same conversation. This is the opposite of a reduction in police capacity. It is a transfer of officer hours from tasks that need a uniform to tasks that need an investigator.

Safety in the home is a different problem

Everything above concerns danger from strangers in shared space. Most serious violence is not that.

Women are most often harmed by intimate partners. Children are most often harmed by the adults responsible for their care: in the federal Child Maltreatment data, seventy-six percent of identified perpetrators are a parent of the victim,¹² and a victim-level analysis puts the figure at 89.1 percent maltreated by one or both parents.¹³ That statistic is not an indictment of parents. It is a statement about where children are, and therefore about where the risk is.

The implication is uncomfortable for how we have built the system. Our instruments — patrol, lighting, visibility, response time — are calibrated to a threat that arrives from outside. In the home, the threat has a key, knows the schedule, and is frequently the person whose account the system will hear first.

That changes what the State owes. It owes a process built around the person reporting harm rather than around the convenience of adjudication: advocates available before the first hearing rather than after, address confidentiality that actually functions, protective orders enforceable in practice, and the ability to keep or leave a lease without financial ruin. I set out the housing side of this in a separate piece; the legal side belongs here.

Children, and the adults at school

For a child, school is often the only setting outside the home where a trusted adult might notice. Federal data reflects this: professionals submit the majority of screened-in maltreatment reports, with education personnel among the largest reporting groups.¹⁴

That places a real obligation on how we train and staff our schools. Faculty and counselors need training that goes past mandated-reporter compliance into the harder skill of recognizing what a child is not saying. And in bullying and discrimination matters, a student needs someone whose role is to represent that student’s interest — not the same administrator weighing the school’s exposure. Those are different jobs, and asking one person to hold both is how a child learns that reporting accomplishes nothing.

What I am pursuing

• Basic infrastructure for safer streets. Most of what makes a street feel safe to walk at night is ordinary public works that has been deferred, and it should be budgeted as public safety spending rather than as maintenance:
  • Lighting, first. Audit street and pathway lighting across the district, prioritize by usage and by reported incidents, and fund it. The evidence base is unusually strong for how cheap the intervention is.
  • Sidewalks wide enough to use, continuous, and not surrendered to parked vehicles.
  • Crossings where people actually cross, with the signal timing a kupuna can make.
  • Sightlines kept clear at corners and along pathways.
  • Bus stops that are lit, visible, and sheltered — where people wait alone, at night, on a schedule someone else sets.
• Civilianize parking enforcement and abandoned-vehicle response, and examine non-injury crash reporting for the same treatment — returning sworn hours to investigation and patrol.
• Design for occupancy. Zone so that people have reason to be on the sidewalk at the hours when an empty street is a dangerous one. Small neighborhood commerce is a safety asset, and we legislated it away.
• Survivor-centered process: advocacy available before the first hearing, functioning address confidentiality, and lease protections for people leaving violent households.
• Student advocacy in schools that is structurally separate from the administrator assessing institutional risk, with training for faculty in recognizing abuse, bullying, and discrimination.

Not everything on that list requires money, and the parts that do are inexpensive relative to what we currently spend responding after the fact. But I would rather be honest about the frame than the price. For most of my life I have not had to think about this, and the policy I would have written from that position would have been worse.