Inconsistencies in Associations between Crime and Walking: A Reflection of Poverty and Density

Abstract

Higher crime rates theoretically deter walking, yet empirical analyses show mixed results. It is hypothesized that more walking occurs in low-income, high-density municipalities that have higher crime rates. Gender, car ownership and relative wealth may also moderate associations between crime and walking. A statewide New Jersey survey (n = 673) of walking was linked to crime and census data. Women were more likely to walk for exercise, but less likely as crime rose. Carless households and wealthier respondents did more non-discretionary walking, but walked less in municipalities with higher crime rates. Poorer, high-density municipalities have higher crime rates and more walking.

Keywords:

• active transportation
• assault
• crime
• discretionary trips
• non-discretionary trips

Notes

¹Our overall research design included an oversample of 400 respondents from Jersey City. This data was not used in this analysis as we are measuring crime at the municipal level, and this would lead to a lack of variation on our crime data in the analysis.

²The AAPOR3 method “estimates what proportion of cases of unknown eligibility are actually eligible.” It is calculated as:

Where I = complete interviews (and screen-outs); P = partial interviews; R = refusals and break-offs; NC = non-contacts; O = other; e = the estimated eligibility of unknowns; UH = unknown households; and UO = unknown other and NE = not eligible (The American Association for Public Opinion Research 2009).

³Respondents were asked for the “reason” for their walk trip. This tells us what the “purpose” of their trip is. The latter is more conventional jargon within the transportation literature, but in our view would not be properly interpreted by survey respondents.

⁴This is a Stata module developed by David Roodman and available at http://ideas.repec.org/c/boc/bocode/s456882.html.

*p < .05.