IRE

NICAR’s copy of the Housing Mortgage Disclosure Act dataset for 2007 has been updated. This dataset, maintained by the Federal Financial Institutions Examination Council, provides information about property loans in the United States, including, for each loan application:

• the race, ethnicity and gender of the applicant
• how much money was requested in the loan
• the annual income of the applicant
• if the loan was considered “subprime” — defined in this dataset by being three points higher than the prime rate — how much higher its interest rate was
• The U.S. Census tract for the property location — highly useful for mapping

Reporters have used HMDA data for years to report housing trends with authority; in the current economic climate, that effort has become all the more important. Did lending institutions in your state grant fewer subprime loans than last year? By how much? This dataset can help answer those questions. Please contact the Database Library with any questions.

A great deal of the news coverage surrounding this week’s Democratic National Convention touched on whether the party could recover from a close primary election between Sens. Hillary Clinton and Barack Obama.

Articles focused on the desire from the Obama campaign to reach out to Clinton supporters, and whether a rift remained in the party.

But ABC News decided to dig a little deeper. They reasoned: campaign finance records are, of course, public — can we use them to take a look at Clinton supporters who may not be so enthused with Obama’s nomination?

ABC Reporter Marcus Baram knew that both Clinton’s donor list and the most recent electronic filings from the Obama and McCain campaigns could be cross-referenced to get an idea of how many donors were either sticking with the new Democratic candidate or shifting over to support McCain.

Although on the surface that may be straightforward, the data for processed FEC filings is kept in a slightly different format than the more raw electronic filings which are posted on the FEC’s site shortly after they are submitted by the campaigns.

Needing to save time, Baram contacted the NICAR Database Library for assistance. The result is an insightful look at how a reporter following the money showed, rather than simply told, how some of Clinton’s donors were switching sides.

- Jeremy Milarsky

By Julie Karceski
NICAR Data Analyst

One year later, and we’re back where we started.

The one-year anniversary of the Minnesota bridge collapse, Aug. 1, sparked a flurry of articles — more than 100 in the anniversary week — reflecting on the event.   Many revealed that bridges in certain states are in worse condition now than one year ago.

Newspapers across the country published stories with updates on the victims and investigations into what caused the collapse.  Many journalists also put a local angle on the anniversary by investigating bridges in their own communities and bringing public safety into question.

Data on bridges, including details such as age and inspection dates, were widely used in these stories.

Pam Sohn from the Chattanooga Times Free Press used a combination of data from the Tennessee Department of Transportation and NICAR for her story published Aug. 1. This was not the first time she had written about bridge safety. Two years ago, her paper investigated bridge safety in Tennessee with data from NICAR.

“About one year after our story, the collapse occurred,” she said. “We looked back and localized it using the data we had before.”

This year, Sohn wrote about the status of bridges in Chattanooga and the planning for repairs. Her story revealed that five out of Tennessee’s twenty – or one in four – highly-trafficked, “structurally-deficient” bridges are located in Chattanooga.

In preparation for increased attention on bridge safety, the Tennessee Department of Transportation released PDF documents on bridge inspections in the state.

“This time the Tennessee DOT did a little homework of their own- a preemptive strike,” Sohn said. “It showed where the worst bridges in the state were.”

Ariel Hart of the Atlanta Journal-Constitution also emphasized the importance of data in writing a story about bridge safety.

“You can’t just say it’s been a year since that bridge fell down,” Hart said. She published a story detailing how bridge ratings in Georgia have gotten worse in the past year, despite heightened attention on bridge safety.

Hart got the data for her story from the Georgia Department of Transportation. The Atlanta Journal-Constitution’s computer-assisted reporting specialist, Megan Clarke, calculated the bridge numbers for the state and gave Hart a few story angles to work from.

“Data made it a news story,” she said. “A big part of the picture has gotten worse. We have more structurally deficient bridges this year. That led us into the background about the budget crunch.”

Darryl Isherwood from The Morning Call in Allentown, Pa., scrolled through the 25,000 bridges listed in the Pennsylvania DOT Web site to provide context to his bridge anniversary story. He, like Hart, was struck by the fact that the number of structurally deficient bridges in the state had risen over the past year.

“One hundred more bridges are listed,” he said. “It made for an obvious story. It was just a matter of calling a few people.”

Isherwood used a spreadsheet do the calculations and spoke with state officials about the budget problems responsible for the increased number of structurally-deficient bridges.

“Having that data there, I was able to find a couple of bridges to focus on,” he said. “The data just made the numbers pop.”

I am proud to announce that purchases of datasets completely online, without the use of a phone or fax machine, is now available to members for most of the data we sell at the NICAR Database Library.

The new online store can be found via links on our main site at www.ire.org, or you can go directly to data.nicar.org.

To enforce the long-standing policy of only allowing journalists who are IRE Members to purchase data from the library, the online store requires visitors to register for the site. (It’s worth the effort; you can save it for accessing new features as we continue to improve our site.)

Within 24-48 hours, IRE staff members will verify whether a registrant is, in fact, an IRE Member.  Members in good standing will then be able to make purchases with a credit card using our new shopping cart.

On a personal note, it is my hope that this new Web application will enable us to better serve our members, who work daily in an ever-challenging environment where every minute counts.

Of course, you’ll still be able to purchase data the old fashioned way — and we certainly encourage people to call the library for our analysis and consultation services. But this new feature will hopefully be a boon to those members interested in purchasing datasets quickly.

Questions? Please feel free to email (datalib@nicar.org) or call us
573-884-7711.

- Jeremy Milarsky

U.S. roads and waterways get more dangerous over the summer months as vacationers hit the highway or fire up their boat motors.

According to the National Highway Traffic Safety Administration’s Fatality Analysis Reporting System (FARS), fatal vehicle accident rates typically inch up during the summer months and then decline during the fall.

It’s much the same on the waterways, according to the U.S. Coast Guard. The high number of boating accidents mirrors the summer vacation season.

So, if your news organization is reporting on these accidents, two databases offered by the IRE and NICAR Database Library that you can use for background research and more in-depth stories.

First is FARS, for details about U.S. fatal highway accidents back to 1975. FARS can be tricky to work with, but it is a trove of information about the accidents, and the people and vehicles involved in them. For instance, you can examine the data to see whether a road has had a problem with fatal accidents. Chris Halsne and his colleagues at KIRO in Seattle recently used the FARS data to help document flaws in luxury recreational vehicles.

Second is the Coast Guard Recreational Boat Accident database. This easy-to-use database goes all the way back to 1969 and can help you pinpoint trouble spots on the water. Former boot camper Marc Chase at The Times of Northwest Indiana got the data a couple of years ago and found that unprepared boaters commonly were to blame in fatal boating accidents.

Interested in either database? Contact the database library.

-David Herzog

With the sad news about flooding in the Midwestern United States this week, the Database Library received more than a few requests for the National Inventory of Dams. Not only does this dataset list the name and location of all federally-inspected dams in the United States, it has information useful for journalists, including when, exactly the last inspection date was for each dam and whether the dam is near a populated area.

More than a few good stories have come out of the dam data, which is still used for training purposes in IRE’s computer-assisted reporting class. In 1995, Dateline NBC aired a story showing the poor condition of many Wisconsin dams (the story included footage of a dam inspector who was able to stick an entire pole in one dam wall).

Unfortunately, federal officials closed this dataset to the public shortly after the Sept. 11 terrorist attacks, arguing that this information should be kept secret in the interests of national security.

This is a shame on a number of different levels, but especially in the context of the news last week. Journalists, as a result, had a tougher time providing the public with broader context as they reported the day’s news about flooding. Bigger questions like “how safe is the dam in your town?” obviously can still be posed, but getting at a real objective answer is now more difficult.

Despite these limitations, there is at least one good use for the data: since the actual locations of dams aren’t likely to change, it can give reporters a peek at where government-inspected dams are and what they are called.

For the record, we don’t plan on giving up on this data. We still request it annually — something at least one official has complained about — and we’ve asked for all correspondence leading up to the decision to close the dataset. We hope to at least get some better explanation of the government’s thinking in denying journalists basic information for reasons of “national security”.

- Jeremy Milarsky

Three years ago (in November 2005, to be exact, according to the Wayback Machine), Investigative Reporters and Editors updated the look and feel of its Web site from something resembling early Craig’s list to something looking more like, well, this.

At the time, the site administrators, who no doubt had their hearts in the right place, put this thing on the main page:

Some people are probably looking at the image on this post, which is a snapshot from the old home page (pre-today), and a link to the even older page (pre-2006) and thinking nothing’s wrong with that. The other 90 percent of you are snickering (no offense to the non-snickerers, I’m just making a point here).

Not only is this a bad idea from a design perspective (think of what it would be like if the New York Times sent you two copies of their paper every day: one that looks as it does now, another in its pre-1990’s black-and-white incarnation, just in case you are “feeling lost”) — but it also comes from a philosophical perspective that no longer makes sense if we want to truly be taken seriously on the Web.

Fortunately, though the past is the past, the future — the way I see it — looks bright. IRE and NICAR have some very exciting things planned for this Web site. We’ll be offering our members a chance to participate in a more personalized supportive online community (think NICAR-L, but with the messages filtered to the topics you work with every day), the ability to purchase data on-demand and more news about investigative journalism and the people who make it happen.

So with a lot of work, and good planning, IRE staff members hope to make this little space on the Web a better portal to our services for our members and the Web audience as a whole. And although it is easy to feel lost and hard, sometimes, to embrace change, there will be no links to the old home page from the new main page.

All that snickering was getting old.

– Jeremy Milarsky, Data Libarary Director

—————

On edit: Just to clarify here, this post is not meant to malign the people who made decisions before on this site. I am simply making the point that what may have made sense in the past doesn’t necessarily do so now. I hope the purpose of this space is to discuss ideas — good or bad — rather than the people who come up with them.