what the survey shows · What predicts a school's ties to outside organizations

This is the same study with every parent measure taken out. Seven organizations rather than eight, because parents' groups is a tie to people inside the school rather than outside it. The finding holds, and the police result is larger than it is with parents in.

Introduction

Schools deal with organizations outside their own walls. Civic groups, local businesses and churches. Social services, mental-health providers, juvenile justice agencies and the police. Seven kinds in all. This study asks one question about them: what decides how many of them a school works with?

The answers come from the School Survey on Crime and Safety, which the US government sends to school principals. It ran six times between 2005-06 and 2021-22. We use 12,691 of the answers.

The survey also asks about parents: whether the school works with a parents' group, whether it has a formal process for consulting parents, how many parents turn up to school events. None of that is here. Parents are inside a school, not outside it, and this version of the study is about outside partners only. The companion site keeps them in, and the two differ in ways that are set out at the end.

Two things predict how many organizations a school works with.

The first is two things the school runs itself: security staff, and security equipment such as cameras, metal detectors and locked doors. Schools with more of these work with more outside organizations, by about +0.23 of an organization for every standard deviation, a measure of how spread out a set of numbers is.

The second is the area around the school, as the principal describes it. Principals are asked to describe the crime level as low, moderate or high. Schools whose principals say high work with more outside organizations, by about +0.09 of an organization per standard deviation, even after what the school runs is accounted for.

Four panels of dot-and-line charts. For each of the seven external organizations, and then for the count of them, the share of schools working with it rises from low- through moderate- to high-crime areas. Six of the seven rise and the police fall. The top row uses the principal's description of the school's own area, the bottom row the areas where students live, and the second is steeper. — Figure 1. How many of the seven external organizations a school works with, against how its principal describes crime in the area. The survey asks twice and both are shown: the top row uses the area around the school, the bottom row the areas where its students live. On the left, each of the seven on its own, as the share of schools working with it where the principal calls that area low, moderate or high crime. Six of the seven rise. The police, grey and dashed, is the estimate this site reports and does not build on, because on its own it reaches significance in one survey only; what the site builds on is the same tie at a fixed number of organizations, which is Table 4. On the right, the same as a count: all seven together, then the three community organizations and the four government ones. Every point is the average prediction from a survey-weighted regression holding school size, grade level, whether the school is in a city or the countryside, and which survey at the sample distribution, so the three points compare the same schools placed in low, moderate and high crime in turn. Asterisks come from a companion regression treating the crime report as one continuous scale, with standard errors clustered on the sampling stratum: ***p < 0.001, **p < 0.01, *p < 0.05.

Then the finding this site exists for. Where a principal reports more crime, the school works with more organizations, and the police are the one partner that does not come along. Take two schools working with the same number of organizations. The one whose principal calls the area high-crime is about 10 percentage points less likely to have the police among them.