At OpenSpending we really want to make it easy and fun for journalists to write great stories with data. But what can we do to help?
There are already tons of ways for journalists to find newsworthy data. For example, in the U.K. there are daily email alerts for government data releases and even calendars pointing to upcoming releases months in advance, not to mention the maps that lead a journalist straight to data about a place. Clearly there’s no reason for a journalist to be short of data or receive it too late — when journalism is all about timing and working with data while it’s still fresh.
But there’s an essential extra step a journalists needs to perform once they have the data: Get the story.
The Data Is The Story
For some data sets, this step is simple: The data is the story. Take road traffic accidents or teenage pregnancy rates — these data sets need a display (map, chart or both), a link to a related news story, and some exploration of the worst-hit areas, and then bingo! You have a blog post. Doing all of this in a few hours is an art form that generally requires a clear head and little assistance.
There are more complicated data sets like this list of attendees of surveillance equipment conferences. Although the list stands by itself as an interesting talking point, there’s some serious work involved in finding relevant connections between attendance and government contracts for surveillance equipment and what this means.
Other data sets require yet more skill and time to get a story, but often that story will be so good it’s worth the effort.
It’s precisely these data sets that we want to help journalists tackle, and reduce the time it takes to get to the story.
What It Takes to Get the Story
As anyone who has visited the OpenSpending website will know, the project is firmly behind the idea that if you want to know what’s happening in your government then follow the money. Governments are sharing their spending habits like never before: You can download big chunks of government accounting databases, and more and more reports allow comparison of spending across countries. Sometimes the spending data isn’t good enough to work with, and sometimes it’s just not published at all. But there’s an awful lot available, and it takes some skill to extract good stories, which is exactly what we want to see.
So here’s the plan: We are monitoring some key data sets, the ones that require work to get the story and that show where government money is going. Journalists can select which data sets they’re interested in, and we’ll alert them when the data is due be released. When it’s released we’ll add a host of ways to get a story. The suggestions for getting a story will be very specific to the dataset, but examples of things we will be including are:
- Related datasets and replies to freedom-of-information requests and suggested ways to combine them
- Clean up the data if it’s not already in a usable format
- Step-by-step guides for how to interpret the data from our team of statisticians and software developers working on our sister project the School of Data
- Previous stories on the same topic that were a success
- Suggested people to talk to (e.g., in NGOs) who may have interesting comments
We will also highlight any delays to the data release as that can often be a story in itself.
We’ll be starting with U.K. and European Union data and then looking to roll out our open-source tools and methods as a package that people all over the world can use to run a service for their country.
This package will contain a mailing list and blog posts as our core datasets are released, along with its full potential laid out and a nice graphical calendar tool for future data releases.
Let us know what you think of our plans, and suggest data you think we should include by emailing us on our public mailing list at openspending.org or our team directly at [email protected].
Lisa Evans is a software engineer and journalist. After helping to create Where Does My Money Go, she worked with the Guardian’s datablog. She now works on OpenSpending.
