Apparently the Philadelphia homicide rate actually went down during the Street administration. Interestingly enough, it was at its second highest during the reign of Frank Rizzo, the former “law and order” mayor - who also had the most cops.
Adds something surprising to the debate over what to do about violent crime.




Read it again Susie……the blogger said the Murder rate was highest under Rendell. I bet population was far lower during Rendell than under Rizzo so Rendell’s murder was even worse per capita.
To try and spur improvements in society, you should try and be honest. That means you can’t cherry-pick data you like just because you disdain Rizzo and like Rendell. When you do that, it shows you lack intellectual integrity.
And to be fair , I would add the murder rate was higher under rendell but that does not mean he was a worse mayor than either Street or Rizzo or Goode.
First of all, I skimmed the piece and missed the last part about Rendell, or I would have included it. I have no great love for the man, so there goes your little theory.
Second, my point was that the perception of what we “should” do (get another mayor like Rizzo) is not borne out by the numbers. Many people are convinced the city was at its safest under Rizzo; it was not.
And finally, how do you figure a lower population makes for a higher per capita homicide rate? “Per capita” means per unit of population.
Murder rate was almost the same under Rizzo and Rendell but population was much lower under Rendell so his per capita murder rate was much higher than Rizzo’s.
For example:
Rizzo round numbers:
About 400 murders with 1.8 Million people = 22 per 10,000 people
Rendell round numbers:
About 400 murders with 1.6 Million people = 25 per 10,000 people or more than 10% higher per capita.
Sometimes selected raw data by itself can imply an incorrect conclusion.
Looking at the averaged deaths per year for each mayor in conjunction with an estimated average population for the city under each regime yields the following:
Rizzo 21.7 deaths per 10,000 population
Green 21.5 deaths per 10,000 population
Goode 23.4 deaths per 10,000 population
Rendell 25.5 deaths per 10,000 population
Street 24.2 deaths per 10,000 population
These results show a trend of an increasing mortality rate since 1972 in Philadelphia. However the year to year variation in the data is conveys a sense the trend may not be significant.
Since I didn’t agree with Mr. Rizzo’s politics or methods, I was hoping the data would prove their failure. But it doesn’t.
As an aside, some of my relatives were involved in the Rizzo regime. My uncle, Frank O’Donnell, was the city councilman who was reelected with Rizzo the day after he was buried. I heard the news on PBS in San Jose, CA while driving to work that Wednesday morning.
Did I ever mention I’m not good at math? Thanks.
My basic point still holds - people have this idea that the city was much safer under Rizzo and the variations still aren’t all that significant.
The real story is probably in the details. I’d like to see these homicides broken down by race and neighborhood.
The other big mistake is that the raw data isn’t the rate. Incidents divided by population divided by time is the rate. There are fewer murders in a smaller population. The murder RATE is nonetheless up. Lots of good news about Philly’s resurgence, but it’s still a dangerous place.
Susie said “Variations are not that significant”. I beg to differ Suzie.
These per capita stats indicate about 30-50 more people died per year under Rendell, Goode and Street than under Rizzo. Over their collective 24 years in office, that is about 1,200 more murders -try and tell their families the “variation is not significant”.
I guess not being good at math is a prerequisite for a member of the media. It is called innumeracy- suggest you read the book by the Temple prof.
Wendy, you make some good points. Granted Susie shouldn’t hide behind the “I’m just not good at math” argument, and the media generally tends to be lazier than Susie when it comes to numbers, but your comment about “not being good at math is a prerequisite for a member of the media” was unnecessary.
I would add, what were the murder rates in comparable-size cities in the US, in the state, and in the country over the same time frame? Those numbers would help me put the Philly numbers in context and give me a better idea of whether a particular mayor’s policies made a difference.
There is an error in my data in comment #4 above. Should have checked it better, but it was late. The death rates listed are per 100,000 population; not per 10,000 as noted.
That being said, here’s some additional analysis. The data show significant swings for each mayor in an Excel spreadsheet graph. Overall however, when a best fitted straight line is used, there is about a 25% increase in the death rate over 36 years.
In words here is the gist of what the graph shows.
Rizzo – The death rate was higher than the average for most years. There was a significant drop in the death rate for 1976 and 1977. The death rate was increasing for Rizzo’s final two years. [High – 23.94; Low – 18.19]
Green – Under Green, the death rate had a steady four year decline. [High – 25.89; Low – 18.82]
Goode – The first year under Goode had the lowest death rate for the five mayors. It increased for seven years to the highest rate for the five mayors; before backing off somewhat the last year. [High – 31.53; Low – 15.96]
Rendell – Under Rendell the death rate stayed around the rate attained in the last year of the Goode administration for six years and dropped significantly the last two years. [High – 28.05; Low – 19.15]
Street – Under Street the death rate held low for the first three years; then showed a steady increase. [High – 27.82; Low – 19.15]
As Susie noted these data are for the whole city. The data are not worth much without more specific information regarding the homicides. In addition to race and neighborhood as noted by Susie, was it a crime of passion or murder during a crime. Was it drug or gang related. The approach for reducing the number of homicides must address the type of homicide.
Describing the problem in simplistic black and white terms leads to a Homeland Security approach. Lots of money and personnel; a pandemic of alerts and self-serving pronouncements; ham-fisted, humiliating surveillance and interrogation methods; and very few actual accomplishments.