Did Forbes get it right?

Forbes-Logo.jpg

Last week, Forbes.com released their 15 most dangerous cities list. Although we at Onboard Informatics prefer to look at the positive side of things in our stories, it doesn't make their story any less valid (or interesting for that matter). What Forbes used as their criteria is the number of violent crimes (murder and non-negligent manslaughter, forcible rape, robbery and aggravated assault) committed per 100,000 people in the population of an MSA (Metropolitan Statistical Area) with total population over 500,000. Overall, you will see a number of cities on this list that intuitively should be there such as Detroit, Baltimore, and Philadelphia. [As a side note, I have been to all three of these cities and love them. All a lot of fun and the people are very nice. My personal feelings about the cities on this list have nothing to do with the data.]

photo by karpov the wrecked train

But as we read through this list, there were some that we thought, "Hey, what about so-and-so?" so we decided to rerun the numbers to see what our data would say about a list based on this same criteria and how Forbes got to their list. Here is what we found.

Forbes chose to base the geographies in question on MSAs vs. the cities themselves - which is fine. We came up with our own theories as to why. Our best was that people who live in these MSAs commute into these cities so these crimes affect the people in the whole MSA. Fair enough. This will work - if you clearly articulate that this is the level of geography that you are measuring. Forbes didn't do this. They took the crime incidences for the MSA but allude to the city proper in it's title 'America's Most Dangerous Cities'. This can be misleading.

You have to compare apples to apples here. Just because someone may get assaulted in Manhattan does not mean that they cannot be assaulted in Westfield, NJ (very low crime rates). The same applies on the flip side.  You may spend most of your time outside of this city where the risk is lower; you may want to look at things on the place level instead. All of this must be taken into account. For this reason, coupled with population numbers, Boston and Chicago are on our list along with some others that will not surprise you. [Again, I have been to both of these cities and they are a great time. I, personally, recommend you check them out.]

Another thing that we noticed was that Forbes used 2007's FBI crime statistics. Although the 2008 stats are available, they are preliminary (first 6 months of 2008).  They are more recent (however subject to update by the FBI) and a solid statistical sample for the purposes of such a simple story based in the now.  There are a bunch of other cities that would be on this list if we dropped the population threshold down a bit, which I would recommend. For now, I will share what this list would have looked like if we were inclined to do a story like this:

Onboard Informatics' List of America's Most Dangerous Metro Areas

OBI Rank

Forbes Rank

City

State

Population

Crime Incidence

Crimes/ 1ooK

1

1

Detroit

MI

860,971

8,443

981

2

2

Memphis

TN

669,264

6,259

935

3

10

Baltimore

MD

624,237

4,890

783

4

9

Nashville

TN

564,169

4,047

717

5

15

Philadelphia

PA

1,435,533

9,783

681

6

Milwaukee

WI

572,938

3,401

594

7

Houston

TX

2,169,544

12,385

571

8

Chicago

IL

2,824,434

15,902

563

9

Boston

MA

591,855

3,096

523

10

Jacksonville

FL

797,350

4,020

504

11

4

Las Vegas

NV

1,341,156

6,646

496

12

14

Charlotte-Mecklenburg

NC

733,291

3,537

482

13

Oklahoma City

OK

542,199

2,572

474

14

Dallas

TX

1,239,104

5,786

467

15

San Francisco

CA

733,799

3,388

462

As you can see, there is a pretty significant difference here. Our list starts to deviate at place #3 and it really doesn’t match up much throughout the rest. With all of this said, a story like this, in our opinion, takes significantly more information and analysis to appropriately vet this notion.

I don't write this post to punch holes in Forbes' story; rather I do it to demonstrate the importance of good data and how numbers can be very easily skewed.  The other half of this recipe (and some would say even more important) is the explanation of what all of this data means.  To really get the most out of a story like this the reader must be able to understand what you are saying here and what it's based upon.

There are factors of property vs. personal crime as well as how one's target audience defines "dangerous". All of this must be explained in a story like this so as to avoid confusion amongst the readers. The last thing anyone wants is for a place to be affected (property values or otherwise) by a highly publicized story - such as this - without taking all of the different variables into account. Also, it is important to do some individual, on-site reporting on all of these areas to really understand the why's and wherefore's behind what gives a place certain characteristics. For a good example of this please visit CNN's Best Places to Live and US News and World Report's Healthiest Places to Retire.

What's your opinion on all of this? I would love to hear it.

Interesting observation: With the exception of Las Vegas, for obvious reasons, every one of these places has at least one MLB, NFL, NBA, or NHL team.  Memphis, Jacksonville, and Oklahoma City only have one each.  From a distribution perspective, these might be good candidates to consider for expansion of these leagues. :-)

-Patrick Healy