Providing Election Results

This year’s midterm elections saw a huge range of online campaigning innovations. None of them was truly revolutionary, but from candidate profiles on facebook to coordinated text messaging campaigns and use of youtube, the emphasis on reaching out to voters online continues to grow. With all the worries about polling irregularities, projects like Polling Place Photo Project (via Zeldman) were also a nice addition to the toolkit.

So it’s especially surprising that it’s so hard to get good information on the results, particularly on the night. We were at the home of a local state senatorial candidate switching back and forth between the local public access TV and the NBC affiliate, and seeing huge discrepancies between the two. Typically the public access programming, was displaying more of the local results, but it was never clear whether their results only applied to the city or covered a broader audience and their windows-based system frequently froze.

When we looked online it was similarly hard to get up to date results, not to mention the difficulty of discerning between projections and actual reports. When we heard that a candidate had won or lost it was difficult to know whether that was the expectation or if the result was actually mathematically conclusive.

Surely it wouldn’t be hard for each State’s Board of Elections to publish the results online? Since they must have some way of gathering the results into electronic systems, tallying them in a simple table needn’t be much extra work and, as ever, if the structure is worked out properly (perhaps with a sprinkling of microformats) it would quickly become easy for enterprising developers to build RSS/email/SMS alert systems, map the data, and aggregate it all in a variety of ways.

For all the recent emphasis in government circles on RSS and podcasts, this is the sort of open data issue that is unlikely to easily gain traction. Perhaps this is one of the ways that newspapers can make use of structured data, providing us with the underlying data before they overlay their own commentary?

Comments are closed.