As I began my exploration of data journalism and its community of practice, I was cognizant that I needed to examine the craft from a broader perspective than I had in the past. My previous efforts were focused too narrowly on data journalism tools, which is perhaps understandable coming from someone with a lengthy reporting career who finds the journalism easier than the technology. When exploring new media forms, however, it is important to remember that while the medium is the message, the message or information can’t get lost in the medium. In other words, the goal is still journalism.
My approach to this assignment, therefore, was two-pronged: to continue my exploration of increasingly sophisticated data tools but to let the content decide the form in order to ensure that the end product conveys new and interesting information.
Infographic: Who earns what at Birmingham City Council?
It wasn’t long before my simultaneous search for a story and an appropriate data journalism tool yielded results. Acting on a directive from central government, Birmingham City Council released the salaries of its top executives just days before voting on an unprecedented austerity program that would cut approximately 10,000 jobs and lower the earnings of thousands of lesser paid employees.
It was a great story and I knew immediately that a flow chart visualizing how many executives earn top dollar would best convey the information to Birmingham voters. I searched for hours for a free data tool that would generate the flow chart automatically but found that my spreadsheet was too large and so I created the visualization by hand using Google Drawings. Although it was a new skill, it was not my first choice as it was laborious and didn’t completely meet my goal of learning to use more sophisticated tools. That being said, this infographic is the sixth most viewed story on our website Birmingham Budget Cuts and generated quite a bit of commentary on Twitter.
The benefit of belonging to communities of practice was immediately apparent, as Paul Bradshaw commented my blog post suggesting an alternate data tool. I also solicited feedback from Michelle Minkoff, a data journalist I had been following on the NICAR discussion group who had not only visual advice but suggested that I could go further with the journalism.
The next step I’d like to see on this would be to visualize the differences between the salaries. Maybe have a little bar chart in each box that shows each person’s salary in relationship to the top executive. It’s telling that Hughes makes about twice that of some of the people listed only two “levels” below him (ex: Sheila Rochester.) Another way to look at this data, that might fit even better with the story, would be to compare these salaries to the “poor” or what the “disabled” receive from the government. How disparate are the financial differences between the people making these decisions and those the decisions might affect?
My exploration into communities of practice in the field of data journalism has been an enlightening yet daunting experience, as I am a journalist with emerging skills trying to enter a world dominated by programmers and already cross-skilled reporters. It’s useful to read online forums, but difficult to contribute when the ethos of the group requires posting specific, practical questions and answers about issues beyond my skill set. Building a public profile is hard work and takes time and resourcefulness. My membership in the DataTO Google Group led to a recommendation to NICAR where I found Michelle Minkoff who has provided extensive advice via email which I then blogged and she then tweeted. I have also joined Hacks and Hackers Toronto and Toronto OpenStreetMap in the hope that upcoming meetups will help bridge the divide.
Toronto Council voting record
My search for communities of practice led directly to my second open data project as I found Socrata, a US website specializing in open government data that interestingly had an out of date spreadsheet on the spending of members of Toronto City Council.
I went to the Toronto open data site and there were no current expense reports for City Councillors but I found their voting records. The resulting data project met my journalism criteria because my ward has a new councillor with no track record at City Hall and I knew my community would be interested in how he’s voting. It was also a technological challenge because the file is csv, a format I have vowed to conquer.
You can view the spreadsheet here.
With the assistance of Paul Bradshaw I edited the file name to delete .xls and then used Google Refine to split the document into readable columns. I imported it to Google Docs and cleaned up the messy columns using the formula =split(G2, ”, ”) and then uploaded it to Google Fusion Tables. At first I was unable to post it on my blog and once I learned from my Fusion Table Users Group that spreadsheets were unembeddable I transferred it back to Google Docs and created a link that way.
This post, which I published on Facebook led to the most local reaction I’ve had yet on my blog and also had the desired effect of publicizing the fact that I’ll soon be launching a hyperlocal website in Birch Cliff.
My tweet was retweeted by Nik Garkush, the Open Source Stratagy Lead at Microsoft Canada who also made a very interesting comment on the blog in that it echoes the advice from Minkoff:
Looking forward to your hyperlocal news!! Re council voting & open data, are there any useful insights that you’ve derived from the data? Its what we do with the data and how we interpret it & share out that make the open data ‘come alive’
His comment is interesting because I made a very deliberate decision to simply post Gary Crawford’s voting record and allow users to peruse the data and draw their own conclusions, although I did colour code to help identify hot button issues. As my blog post indicates, my goal was to make “the invisible visible” and experiment with removing the mainstream media filter I’ve become accustomed to using for the last 30 years. If I was still in Birmingham, I’d be interested in having a class discussion on the value of this approach.