Msnbc.com has obtained and mapped the newest bridge safety records from the federal government. The data shows that thousands of bridges have not been inspected every two years as they are supposed to be. Msnbc.com is making the data available to all journalists.
Pulitzer Prize-winning journalist Bill Dedman answers Poynter Online's questions about
the project:Tompkins: What did the data reveal that surprised you most?
Dedman: All we assumed was that the records would show which states were inspecting bridges on time.
But as always happens, the records help explain how the system works, or where it doesn't work.
We had no idea that not all states were trying to inspect "every bridge at least every two years" -- the claim that states made after Minneapolis. We didn't know that states could get an exemption to delay inspection of some bridges, and that a few states would abuse that loophole by putting bad bridges on the slow list.
We also didn't know that states would vary so much in their inspections, that it's a patchwork of public safety as you drive from state to state. Many states have very clean records, with every bridge inspected on time. A few states have big problems.
We didn't know that the federal government has the authority to withhold funds for failing to inspect bridges on time, but hasn't used that authority in at least 15 years. We didn't know that the federal
government owns bridges, and that so many of them would be late for inspection.
And we didn't know that the Golden Gate Bridge would be late for inspection. That's what always happens with these public-records stories: You go in confident only that it will be news which states (or towns or police departments...) are better than others. The ranking is enough of a reason to do the story. But then you learn something about how the system works. It's anti-anecdotal reporting: Instead of starting with an anecdote, you start with the pattern, but trust that the anecdote will come. If you go out to one of those uninspected bridges at 7 in the morning and wait with your camera, a school bus will come along.
Tompkins: Why was it important to include a mapping function that allows the user to plot a route and see a bridge?
Dedman: We're just intermediaries.
The Federal Highway Administration hands out bridge records in this
format:
069 27 0052 1210010100404100000GOLDEN GATE-SAN FRAN BAY U.S.
HIGHWAY 101 04-MRN-101-L.01
999900000101000000010101374912001222836000641323212193706001080001999003
350001001121067109999A55313311003001018912805027903032032018902849999N00
00N000000 8 1657139492N585381027423040424N N N
0654860065490982291999 1N2
10000056000202Y817990020200000NN1190212 070722FC 119307209712 2
0589
That's the single row of data for the Golden Gate Bridge. We can make it useful.
While it's somewhat interesting to know generally that bridges are in poor condition or late for inspection, it's much more interesting to know whether one of those bridges is on your way to school or work.
Not every bridge is on the map -- because we had to limit it to bridges with at least 10,000 vehicles a day, and because states didn't provide the latitude and longitude of some bridges. (By the way, for a few years after 9/11, the locations of bridges were removed from the database, but they're back. Apparently the government has realized that a terrorist picking a bridge for a target is unlikely to start with the National Bridge Inventory.)
I didn't make the map -- a whole group of smart people at msnbc.com worked on it, including interactive producer Phil Zepeda and the designer, Julie Yokers. They turned our processed/analyzed data file into something elegant.
Clearly one challenge for the technical gurus is the scale. A map for a printed newspaper has to look right one time, and it should come out the same in every copy. But msnbc.com has about 100 million visitors a month -- about five times the traffic of the most-visited U.S. newspaper Web site,
The New York Times -- so a map has to work even when millions of people are using it at once. A lot of people worked behind the scenes to make this happen.
Tompkins: How have you learned to think interactively while working on projects like this one? How is that different from when you worked only with printed words and graphics for the newspaper?
Dedman: The work doesn't really feel that different from pre-Web days. We get an idea, we get the records, we read the documents that show how the system is supposed to work, we test out hunches, we share preliminary conclusions with the people we're writing about so they can tell us where we misunderstood, and we try to find a clear way to tell the story.
The map is part of the storytelling on this story. And it gives the reader something that lasts. Readers might see the story once, but they can come back to the map again and again.
We also wanted to give the readers a way to interact with officials, so you'll see that every chart has icons allowing readers to send e-mail to bridge officials in each state. And a copy of the e-mail goes to us, so we'll get an idea how often that feature is used and what readers are saying.
Tompkins: Think forward. How would you tell this same story three years from now in a way that will be even more effective?
Dedman: I wanted to do more with video -- take a ride with a family whose commute crosses several bad bridges and/or uninspected bridges. We did find a family who put a child on a school bus crossing a bridge -- just by following the school bus one morning and watching kids get on. But we could have done more with a family's road trip or daily routine.
Of course, we're providing the records to everyone, so any reporter can go find that family and take a drive.
Tompkins: Final thoughts?
Dedman: We were following the advice from former newspaper editor Gene Roberts, of
The Philadelphia Inquirer and
The New York Times. Paraphrasing: It's not enough to flood the coverage on a big story. We have to stay even after the others go home. That's what we tried to do after the Minneapolis collapse, getting the latest national information on bridge quality and inspections, trusting that those would be relevant to readers long after the "news peg" had passed.
And the story seemed obvious. As one of the state officials told us, "After Minneapolis, someone was going to ask the inspection question eventually."
I'm attracted to the stories that any of us could do. These new records on bridges were available to the first news organization that asked for them. If you beat me on a story because you have a secret source inside the executive branch, I'm jealous, but I don't really expect to meet Deep Throat in a parking garage. But when you beat me on something that I might have been able to do, if only I could have thought of it, that really makes me mad. (Like
The Washington Post's revelations on medical care for veterans at Walter Reed Army Medical Center.) Those are the ones that make me say, "Damn!"
