Exploring the impact of neighborhood characteristics on crime rates in the United States: a case study of homicide using the multilevel modeling analysis

Crime is an unavoidable reality in the twenty-first century, especially in the United States. Existing literature has extensively examined the occurrence of crimes and their impact on the economy and well-being. In addition, other studies have investigated the causes of crime among different societal classes using interviews and traditional logistic regression models. However, the use of multilevel models to analyze the influence of neighborhood characteristics on crime rates, particularly, homicide in the United States remains underexplored. We used multilevel models to explore the impact of neighborhood characteristics on homicide across different cities in the United States, using the data from the 2010 National Neighbourhood Crime Study Wave 2 (NNCS2), United States, 1999–2013, which included 10,206 observations. The result from model 2 revealed that the Black race/ethnicity was positively correlated with homicide than the other races. Model 3 found that concentrated disadvantage and unemployment were statistically significant with homicide across different cities in the United States. The findings from our study support the conclusion that homicide is more prevalent among the Black race/ethnicity than the White and Asian races. The random intercept model performs better than the null model when additional predictors are introduced. We recommend that socioeconomic opportunities be made available to people, especially the Black race who are unemployed and living in disadvantaged neighborhoods across different cities in the United States to reduce homicide rates.