Category Archives: Data

New government buzzwords

The New Republic ran delightful little piece last month about the buzzword “disruptive.” Author Judith Shevitz argues that the inventor of the term “disruptive innovation,” which originally applied to the tech industry, has let it spin out of control, so that it’s being applied to areas like education and public health. This trend, she says, is “a category error” because “[n]ot all civil services need to be hyper-efficient and bargain-basement and in a state of permanent revolution, especially when the private entities tasked with disrupting government operate largely outside public view.”

I think her point is well taken, but I particularly enjoyed the story art, which mocks vacuous buzzwords generally. This got me thinking about some of the buzzwords I’m guilty of using. Of course, first and foremost is the title of this blog, as “data-driven” is  ill-defined jargon. I decided to plot a few terms using Google’s ngram, which charts how often terms appear in Google Books over time (click the hypertext to see a clear version of the chart below).

buzzword ngram

 

There’s no real conclusion to draw here, but I thought the information was interesting regardless. Not suprisely, terms like data-driven, public-private, and civic engagement have increased dramatically in use in recent years. Intriguingly, “urban renewal” has been surpassed in use by “e-government.”

New interactive infographic rates risks in all 50 states

After the jump is a new infographic that was suggested as a follow up to my post highlighting that three Detroit neighborhoods had topped the list of most dangerous places to live. Just as depressingly, this map pulls data from various sources to rank all 50 states based on the terrible things that could happen to you there. You can check out the original webpage here. Notably for midwesterners, Illinois and Michigan are ranked 40th and 44th respectively for murders per 100,000 people (with rank 51 being the worst and going to D.C.), and Illinois has the highest percentage of adults reporting poor mental health (according to a Kaiser Family Foundation analysis of CDC data). Obnoxiously, as you’ll see if you interact with the map, the states are color-coded by total number of murders and traffic fatalities, even though the title for those sections says “per 100,000.” To clarify the accurate info click “see the data behind the rankings.” Continue reading

Renting versus owning: how homeowners come out ahead

I’m doing more digging into the 2011 American Housing Survey, as a follow up to this weekend’s post about the age of American housing.

The survey breaks down “owner occupied” versus “renter occupied” units, and the differences between them are compelling. In particular, the survey hints at public-health downsides to renting, though these results are likely tied more to poverty than to the fact a person is renting as the data does not account for socioeconomic status. Nonetheless, here are some interesting statistics.

1. Renters share bedrooms more often.

chart_1

As you can see, around 50% of households in either category have .51 to 1 person per bedroom. But for households with 1.51 or more people per bedroom, the renter category jumps to more than 20%, while only 6.7% of owner-occupied households are that crowded.

2. Renters are more likely to have unsafe drinking water.

Frankly, it’s amazing to me that nearly 10% of people in the United States are without safe drinking water, but it’s worse for renters than for homeowners.

3. Renters have more problems with pests, bad wiring, and holes in floors and walls.

chart_4

As you can see, homeowners have more mice problems, but that’s about it. Renters have more problems with rats and much more problems with open cracks or holes, exposed wiring, and cockroaches. I’ve experienced that last one while renting in the Deep South, and it’s terrible.

4. Renters are less healthy.

Maybe it’s because of all the other things mentioned, but renters report being in “very good” health less often than homeowners. Although both categories of people have the same number of people who reported “excellent” health, renters also report “poor” health more than homeowners.

chart_5

Let me know your thoughts on the comments below.

The age of housing in the United States

The Census Bureau recently released the results of the 2011 American Housing Survey. One noteworthy point for me was how old housing is in the U.S. The data basically shows a rough bell curve peaking between 1950 and 1979. The median year was 1974, 34 years ago. This number has crept up since 1985, when the median house age was only 23, according to the National Association of Home Builders.

age of houses in the united states

For more analysis on this issue, check out this oldhouseweb.com article from a few years ago. It shows which states had the highest concentration of old housing at that time. A surprising amount of it is in the Midwest, with Michigan, Illinois, Wisconsin, Indiana, and Ohio all in the top ten.

The varying costs of medical services by state

health symbolEarlier this summer, the Center for Medicare & Medicaid Services published interesting data on the charges submitted by hospitals in all 50 states for 30 different outpatient services. The Center published the amount ultimately paid for the services, and similar data for 100 different inpatient services.

As an example, I decided to create a visualization showing the varying average costs of one type of outpatient service, “Level II Cardiac Imaging,” by state (except for Maryland, which wasn’t included in the dataset). Keep in mind as you look at these charges that Medicare paid out a national average of $744.58 for the procedure, even though the national average submitted charge was more than $4,000. After clicking the image below, you can see exact amounts by hovering over a particular state:


 

As you can see, the amounts charged value greatly. In my view, these differences confirm at least two things. First, this data shows just how easy it might be to submit overcharges, given the wide discrepancies, and thus why the Obama administration saw a pressing need to crack down on Medicaid and Medicare fraud.

This data also unmasks America’s broken system for pricing medical procedures. Wired ran a great piece last year about this problem. The article noted that “a recent study of the costs of routine appendectomies performed throughout California” showed that, for nearly the same procedure, “the charges varied more than 100-fold—from $1,529 at the cheapest to $182,955 at the most expensive.” The article concluded that “Job One” was “transparency in treatment, cost, and institutions.” This data release is not perfect transparency, but it’s a good start.

Apparently, one reason for the price differences may be the migration of these services from doctors’ offices to hospital outpatient departments, which often charge more. Complicated, eh?

Detroit tops list of most dangerous neighborhoods

Detroit netted the first 3 spots on NeighborhoodScout’s recent list of the 25 most dangerous neighborhoods in America. The Midwest sadly dominates this list. Chicago is on there four times. Overall, the Midwest takes 14 of the top 25 (if you include St. Louis). In the top two neighborhoods in Detroit (W Chicago / Livernois Ave and Mack Ave / Helen St) you have a 1 in 7 chance of becoming a victim within a year. Here’s the link to the full list.

Legal aspects of big data

big data sheriff's star“Big Data”— the business-world buzzword for the collection and analysis of massive amounts of data—has caught on with local government officials in the past few years as many cities have developed extensive data portals providing citizens access to heaps of public information like data from 311 calls. And its not only local governments getting involved, in 2012 the White House announced the “Big Data Research and Development Initiative,” through which federal agencies would commit funding toward collecting and analyzing “huge volumes of digital data.”

So, what sparked this interest in “Big Data”? In short, innovations in computing, particularly the ability to allow users to remotely access large data sets stored on third-party servers, i.e., “cloud computing.” But as attorney John Pavolotsky wrote last November in Business Law Today, “[w]hile business publications have written widely about Big Data, legal commentators have written sparingly on the subject.”

Pavolotsky goes on to note three areas of legal concern he see with Big Data: privacy, data security, and intellectual-property rights. He does not dwell long on data security and IP rights for long, other than to note that API licenses should be reviewed carefully to determine the permissible scope of data distribution. He focuses instead on privacy, arguing that, because of the “inherent squishness” of the legal standard applied to collection of cellphone or GPS data under the Fourth Amendment—which protects guards people’s “reasonable expectation of privacy”—perhaps legislatures should limit the length of time data can be stored.

I’d like to add one other interesting question surrounding Big Data, though it leans more economic than legal: whether data collection should remain public or be privatized. This issue comes up with vacant-property registration, which I’ve written about before, as many local governments allow registration through MERS, a private company, rather than directly through local data systems. Government officials are then provided access to MERS.

Privatization of data collection and analysis provides many benefits, particularly in that it is cost-effective for local government to take advantage of an already developed platform. The primaru draw back, however, is the risk of industry capture, as with MERS and its association with the mortgage industry.

The best solution, when available, is for local governments to take advantage of open-source programs or nonprofit developers (as available through Code for America). Otherwise, there are companies like Socrata that, as far as I can tell, are not closely associated with any industry other than the cloud-computing and data-collection industry.

Interesting report on train ridership in the U.S.

0802130709I want to flag a recent report by Brooking’s Metropolitan Policy Program about rail ridership in the United States. Perhaps the best part is that they have created an interactive map using the data collected, so it’s fun to poke around and look at their findings. I encourage you to check out the full report, but it’s bottom line conclusions are:

  • Amtrak ridership grew by 55 percent since 1997, faster than other major travel modes, and now carries over 31 million riders annually, an all-time high.

  • The 100 largest metropolitan areas generate nearly 90 percent of Amtrak’s ridership, especially those in the Northeast and West.

  • Only ten metropolitan areas are responsible for almost two-thirds of Amtrak ridership.

  • The short distance routes consistently dominate Amtrak ridership share and captured nearly all of Amtrak’s recent growth.

  • Combined, Amtrak’s short-distance corridors generated a positive operating balance in 2011—while corridors over 400 miles returned a negative operating balance.

Efforts to make a smarter Honolulu

Honolulu is, somewhat surprisingly, the 10th largest municipality in the United States. IBM is working with the local government to update its technology—to “build a smarter Honolulu,” as they would say. Here is an interesting video on efforts to provide data to citizens, and the results from those efforts, namely, citizen-created apps. A bus-tracker app is mentioned, and similar apps have been developed in other cities (there are enough apps tracking the CTA in Chicago that it has created, as Jacqui Cheng put it, a “battle of the CTA bus trackers“). The tsunami app is, however, more unique.

Government debt in the 19th century versus today

I recently obtained a copy of Gaskell’s Compendium of Forms from 1889, and it has a fascinating chart of the top municipal debts per person at the time that provides an interesting perspective on the history of government debt:

municipal debt chart

I was surprised by how high these numbers were even back then. But how do these numbers compare to today?

According to Forbes, in 2010 New York City had a total debt of $64.8 billion or $7,760 per resident. That’s way up from 1889 when the debt per capita was under $100. The per-capita debt is up even if you account for inflation, which would make the 1889 debt rise to around $2,500. Detroit similarly has modern debt in the billions—$3 billion as of 2011—and with its smaller population, a debt per capita of more than $4,000 per person, up from $7 in 1889 ($200ish in today’s dollars).

According to the Sun Times, Chicago topped these levels in 2011 with a debt of $10,000 per residents ($27 billion total), up from $26 a century ago (or around $700 with inflation). This debt has led Chicago officials to consider all sorts of creative fundraising measures: fining mortgage lenders to register vacant properties, garnishing parking fines from tax refunds, selling ads on landmarks.

For an in-depth discussion of municipal debt and municipal bonds, check out this article over at citymayors.com.