What is going on in Duquesne? (Pittsburgh, Penn Hills: crime, houses, neighborhoods)

[PreservationPioneer]

It seems that murders are frequently being reported in Duquesne. Why is the crime so much worse there than in other distressed, low-income communities in the area? What is it about Duquesne that breeds so much violence and crime? Thanks!

[squarian]

Is it, though? I haven't looked into the crime stats - maybe someone will provide us with the data (Uptown?). Certainly Duq has been in the papers lately, and certainly it's a marginalized, impoverished crime-incubating area. But is it actually any worse than Homewood, East Hills, Wilkinsburg, etc, etc, etc?

[Love2Golf09]

it kind of surprises me too, since everytime I drive through duquesne it looks like a ghost town. It's very rare to even see a human outside. I think there is like one bar there that is always open, but I never, ever see anyone walking around, driving, or go in and out of homes. The whole place is basically abandoned. Perhaps a lot of the violence has to do with the large number of vacant homes serving as crack houses? I'd be surprised if the population of duquesne was higher than 9. If it is, then they must all sleep during the day.

[PeterRabbit]

Excluding city neighborhoods, Duquesne had the highest suburban murder rate in the county during 2011 - 66.7.

[Uptown kid]

Yep since the year of 2008 there have been 14-15 homicides in the City of Duquense. Only the larger Pittsburgh, McKeesport, Wilkinsburg, & Penn Hills have a larger homicide count... It seems like you can't pick a fight in Duquesne the by the time night falls you might have slugs in your chest!

[PreservationPioneer]

Originally Posted by PeterRabbit

Excluding city neighborhoods, Duquesne had the highest suburban murder rate in the county during 2011 - 66.7.

I just checked the crime stats site and Duquesne had 5 murders in 2011, which gives it a murder rate of 89.8 per 100,000 people! That is ridiculously high.

For comparison, the rate in Flint in 2010 was 48.5, and Detroit was at 34.5!

Penn Hills had 6 murders, but that is spread out over a population of 42,000, giving it a more reasonable/safe rate of 14.2. Wilkinsburg only had one murder in 2011, giving it a rate of only 6.3. That is assuming the stats on the site I linked to are correct. Could Wilkinsburg really be safer than Penn Hills?

Duquesne's population was 5565 at the 2010 census. It peaked at 21,000 in 1930.

[gg]

It is a bit scary that Duquesne has that many murders in the winter months. Usually the hoodie people are more active in the summer. Guess with a warmer winter, it is more active.

[ML North]

Originally Posted by alleghenyangel

I just checked the crime stats site and Duquesne had 5 murders in 2011, which gives it a murder rate of 89.8 per 100,000 people! That is ridiculously high.

For comparison, the rate in Flint in 2010 was 48.5, and Detroit was at 34.5!

Penn Hills had 6 murders, but that is spread out over a population of 42,000, giving it a more reasonable/safe rate of 14.2. Wilkinsburg only had one murder in 2011, giving it a rate of only 6.3. That is assuming the stats on the site I linked to are correct. Could Wilkinsburg really be safer than Penn Hills?

Duquesne's population was 5565 at the 2010 census. It peaked at 21,000 in 1930.

Safety -- such an elusive concept. So while I think your data seems perfectly reasonable, I think the comparison between large and small cities gets to be messy.

In areas with small populations, 1 or 2 random homicides may have a very large impact on the per 100k murder rate. Random homicides don't necessarily establish a trend that is indicative of a deeply-rooted violence problem. And since random homicides are generally a small percentage of total homicides, in large cities their effect is much less pronounced.

So the problem then is how to filter random homicides from environmentally-influenced homicides. Because if you're trying to determine how "safe" an area is, what you should care about is how the dynamics of that area propagate violence. A simple method to smooth out the effect of random homicides would be to use multiple years of data. Now doing that introduces other complications, but I think it may lend itself to more accurate comparisons against larger population centers.

[gg]

Originally Posted by ML North

So the problem then is how to filter random homicides from environmentally-influenced homicides. Because if you're trying to determine how "safe" an area is, what you should care about is how the dynamics of that area propagate violence. A simple method to smooth out the effect of random murders would be to use multiple years of data. Now doing that introduces other complications, but I think it may lend itself to more accurate comparisons against larger population centers.

I can agree to a point, but you certainly can get killed as a bystander or passerby when the druggies are having their normal shootouts.

Most homicides are drug related. Over 80% of the prison population is drug related crime. If you just reported murders that are not drug related, I suspect the number would be amazingly low.

Here is some interesting reading regarding prisons.

The prison industry in the United States: big business or a new form of slavery?

[ML North]

Originally Posted by h_curtis

I can agree to a point, but you certainly can get killed as a bystander or passerby when the druggies are having their normal shootouts.

Sure, but the chances of that happening are so much smaller than being the actual target.

Originally Posted by h_curtis

Most homicides are drug related.

Yep, and those crimes would fall into the category that I'm calling "environmentally-related homicides" which should directly figure into your assessment of an area's safety.

Originally Posted by h_curtis

If you just reported murders that are not drug related, I suspect the number would be amazingly low.

It would be, but it would also be difficult to filter out. For instance, if a strung out drug addict throws a brick through your window and runs away, prompting you to chase him into the street and shoot him, is your crime drug-related? At some point you have to establish limits as to what can be considered drug-related, and doing so would introduce some uncertainty to the assessment.