Social media and news media analysis of the Australian federal election
Thanks to the backing of independents, Julia Gillard holds power with the narrowest possible margin, with 76 MPs out of the 150 member House of Representatives backing a minority Labor government.
The Three Amigos (independent MPs) – Bob Katter, Rob Oakeshott and Tony Windsor – took up an absorbent amount of the media spotlight over the last two and half weeks. We were interested to see if the public paid them as much heed as the media.
We were also interested to see what data would be gathered and analyzed and how it might be visualized to answer this question. We had a look at Google Trends and Twitter to see how people were searching for and talking about the independents.
Google Trends
Google Trends reported that search volumes spiked dramatically for the three independents when it became apparent Australia was facing a hung parliament – Katter’s comical outbursts and caricature cowboy appearance giving him an initial lead in the look-at-me race. Search volume for Windsor pre-election was high, but we attribute that to noise – Windsor is a common name. If you discount the Windsor search volume pre-hung parliament, it’s clear that it correlates roughly with the Katter search volume.
Unsurprisingly, news reference volume matched the initial election day fall-out, and began to peak again in anticipation of the independents’ decision.
What was surprising, was that Katter and Windsor continued to receive as much or more media coverage than Oakeshott, despite his position as apparent leader during the independents consultation and decision-making process.
Notably, Katter proved something of a home-town-hero.
Google search volume for Katter within the state of Queensland rose sharply following the election result, whilst Windsor and Oakeshott failed to even make the graph.
Twitter trend analysis tool, Trendistic, showed that the Twitter community snubbed the trio over the past few weeks, despite their moonlighting, until their respective announcements regarding their choice of party yesterday.
Interestingly, a closer inspection of the Twitter activity in the 24 hours surrounding yesterday’s announcements might reflect the electorates comparative affection for each of the independents.
When Twitter traffic over the past 30 days is analyzed, comparing Gillard and Abbott mentions with the independents, the results are as expected. Sort of. Gillard and Abbott shared the attention in the lead-up to the election, although Abbott got more (possibly pointing towards the substantial swing to the Coalition at the ballot box).
The independents then dominate the end of the graph as they finally unveiled the party they were siding with. Analysis of Twitter activity in the 24 hours following the independents respective press conferences showed that Gillard took over the Twitter-waves as the newly elected PM.
Social media vs news media
There is an interesting anomaly apparent between the social media data and news media data relating to Gillard, Abbott, Katter, Windsor and Oakeshott over the last 30 days.
Assuming that Twitter mentions represent the Australian electorate, Trendistic proved that voters only cared about Abbott and Gillard until election day. From there, interest waned. But most revealing was the fact that the three independents didn’t rate a mention until they announced their support for Abbott or Gillard, allowing a government to be formed.
Whilst Google Trends search volume index and news reference volume measured over the same period for the five political figures correlate with each other, and confirmed that the media and public ‘s interest focused on the two leaders in Abbot and Gillard and spiked around election day, they differed from Twitter mentions in one significant way.
The news media coverage of the independents remained steady and on par with Gillard and Abbott over the last 30 days (excluding election day). Twitter users did not care about the independents until yesterday’s decision.
So what does this tell us?
Is the social media hype just that, or is old media missing the mark?
And, what have we learnt?
That data visualization can help us understand trends and, with the right data, public opinion – very handy.
That the importance the media places on certain issues and people is not necessarily a good reflection of the importance the electorate places on those issues (if we can assume Twitter represents that).
And, that the disbanded-band-of-independent-brothers have had their fifteen minutes, and that Australia, is one heart attack away from a change of government.