Underbelly

Public safety is an important aspect of the quality of life of a city’s citizens. Research suggests that cities with a high percentage of people who are satisfied with safety also tend be more satisfied with their overall quality of life.[1] At the same time, rising income inequality can affect the distribution of crime across victims, while theories have sharply divergent predictions about whether crime victimization increases more for the rich or the poor.[2] The problem to be addressed concerns inequities in the distribution of crime in City of Pittsburgh, Pennsylvania.

Inequity in the City of Pittsburgh is one of the key long-term stresses, where a 2017 study rated Pittsburgh’s equality score at 55 out of possible 100, signifying the prevalence of inequalities by race, income and gender.[3] Public safety was rated 44 out of 500 based on 5 indicators. Strong evidence of disparities in crime victimization included:

• Higher rates of homicides among black populations than white populations (58.6 and 4.6 per 100,000 residents, respectively), and

• Higher rates of domestic violence and property crimes (including theft, burglary and arson) among black populations.[4]

To explore the relationship between race, income and criminal victimization for Pittsburgh, I compare maps displaying serious crime rates with household income and racial demographics and identify census tracts with populations most at risk.

Is there a relationship between neighborhood income and racial composition? Using 2018 Census Data, I find that the proportion of Blacks or African Americans living in an Allegheny County census tract decreases as median household income increases. 

To understand how crime victimization at a neighborhood level varies by income distribution and racial composition, I analyzed crime incident data for homicides and property crimes between 2016 and 2018 as reported by the City of Pittsburgh Police.

The data replicates the finding that there are higher rates of homicides among black populations than white populations. Homicides tend to occur in neighborhoods with 30% or more Blacks or African Americans.

On the other hand, property crimes are prevalent in both low income and high income neighborhoods, supporting Levitt's (1999) theory that property crime victimizations have become increasingly concentrated among the poor.[5]

A Poverty and Crime Indicator was used to identify the top 20 tracts most vulnerable to poverty and serious crime. The indicator comprises of: 

• Wealth Quintile with scores between 1 to 5, where Quintile 1 is the poorest quintile and quintile 5 is the wealthiest based on median household income,

• Homicide Quintile with scores between 1 to 5, where Quintile 1 has the highest number of homicides and quintile 5 has the lowest, and

• Property Crime Quintile with scores between 1 to 5, where Quintile 1 has the highest number of property crimes and quintile 5 has the lowest.

The individual scores were summed with equal weight to get a composite score between 3 to 15. The top 20 with the lowest scores received scores between 3 to 5 and are indicated in the map below by a yellow outline.

Profile of the Top 20 Most Vulnerable Tracts (5% of total)

• 108 homicides (53% of total)

• 10,414 property crime incidents (20% of total)

• Total Population of 45,149 (4% of total)

• Black or African American population of 23,423 (16% of total)

• White population of 18,059 (2% of total)

• 9 of 20 tracts are majority Black or African American

• Median Household Income ranges from $15,185 to $47,986 

The exploratory analysis reveals that serious crime disproportionately affects low income as well as Black and African American populations. Identifying populations at risk can help inform where to target interventions. Further research into how these trends are changing overtime and reasons for why disproportionate crime victimization is occurring can inform what kinds of interventions can create the most impact to improve the equity of public safety and quality of life of the most vulnerable.

• Crime data: The City of Pittsburgh participates in the National Police Data Initiative. Open data from crime incidents between 2016 and 2018 was downloaded from the Western Pennsylvania Regional Data Center,  https://data.wprdc.org/dataset . The Police Incident blotter contains crime incident data reported by the City of Pittsburgh Police which has been validated and processed to meet Uniform Crime Reporting (UCR) standards. Data is reported at the block/ intersection level.
• Data on population demographics by race and income for 2018 was downloaded from the US Census bureau.
• Data for this project includes the census tract layer for Alleghany County, Pennsylvania which has the city of Pittsburgh within its boundaries. It was downloaded from the ESRI Tiger/ Line website,  https://www.census.gov/cgi-bin/geo/shapefiles/index.php .

[1] Węziak-Białowolska, Dorota (2016). “Quality of life in cities – Empirical evidence in comparative European perspective.” Cities, Volume 58, pp 87-96.  https://doi.org/10.1016/j.cities.2016.05.016 

[2] Levitt, Steven (1999). “The Changing Relationship between Income and Crime Victimization.” Economic Policy Review, September 1999.

[3] City of Pittsburgh (2017), Pittsburgh Equity Indicators: A Baseline Measurement for Enhancing Equity in Pittsburgh.

[4] Ibid.

[5] Levitt, Steven (1999).