曼哈顿政策研究所-犯罪热点：对2010年、2015年和2020年纽约市街道的研究（英）-2021.8-30页_1004kb

Summary of "Crime Hot Spots: A Study of New York City Streets in 2010, 2015, and 2020"

Core Content

This report examines the distribution of crime in New York City (NYC) across three years: 2010, 2015, and 2020. It highlights the significant concentration of crime on a small number of street segments, despite a general decline in crime rates across the city.

Main Findings

• Crime Concentration: A small proportion of streets in NYC account for a large share of the city's crime. Specifically:

  • 1% of streets produce 25% of all crime.
  • 5% of streets produce 50% of all crime.
  • These concentrations remain stable over time, even as overall crime levels decline.
  • 15–16% of streets produce 75% of crime, and 55–58% produce all crime.
  • 45% of streets have no crime in a given year.

• Overall Crime Trends:

  • In 2010, there were 509,731 total crime reports.
  • By 2020, this number had dropped to 413,412, a decline of 19%.
  • Violent crime dropped by 25% from 2010 to 2020.
  • Property crime decreased by 10% over the same period.

• Hot-Spot Stability:

  • The same streets were identified as hot spots in 2010 and 2020.
  • High-crime streets are concentrated in Manhattan, Brooklyn, and the Bronx.
  • There is significant street-by-street variability, with some streets having very high crime levels and others almost none.

• Violent Crime on Hot-Spot Streets:

  • 25% of violent crimes occur on 1% of streets.
  • In 2010, the average number of violent crimes on these hot spots was 17.7 (median 15).
  • In 2015, the average was 19.8 (median 16).
  • In 2020, the average was 16.5 (median 14).
  • Even in the second quartile (25–50% of crime), there are still 6.5–7.2 violent crimes per street on average.

• Key Variables Contributing to High Crime:

  • Population density and commercial activity are linked to higher crime levels.
  • Concentrated disadvantage (poverty, unemployment, public assistance use, etc.) and low collective efficacy (lack of neighborhood trust and willingness to intervene) are important predictors of high-crime areas.

Methodology

• The report uses NYPD crime reports with XY coordinates.
• Street segments (not intersections or address-level data) are the geographic units analyzed.
• The generalized Gini coefficient is used to measure crime concentration, with values ranging from 0.77 to 0.84 for violent and property crime, indicating high concentration levels.

Implications

• Policy Focus: The findings suggest that police and city agencies should focus on the top 1–5% of streets that produce the majority of crime.
• Misconceptions: There is a misconception that crime is evenly distributed across neighborhoods, but the data show otherwise.
• Intersection Exclusion: Intersections are excluded from the analysis, which may undercount the crime problem on high-crime streets by 13–19%.
• Policing Strategy: The report argues for targeted interventions on hot-spot streets, rather than broad neighborhood-level policing.

Conclusion

Despite an overall decline in crime, many NYC streets still experience high crime levels, requiring focused police and government action. The report underscores the importance of understanding localized crime patterns rather than relying on citywide statistics, which can mask significant spatial variability. The stability of hot-spot locations over time reinforces the idea that crime concentration is a consistent feature of urban environments, and that targeted policing remains a critical strategy for reducing crime effectively.