UK General Elections, we are told have become increasingly presidential. And how each party leader comes across, particularly on Television, is important. Ed Miliband has faced criticism for his style and communication skills – It’s probably fair to comment that he’s not a natural TV performer. But what if it’s slightly worse than that? What if his television appearances have, overall, a slight negative effect on Labour’s vote share in the polls?
We should look at the evidence – What is the effect of Ed Miliband being on Television, overall?
There are 2 parts to this. First, Labour’s vote share in the polls. This is easy to track, and the data is freely available. How do we measure “Ed Miliband Television Time”? This is not obvious, without access to expensive media databases. We can however, use a proxy. It’s clumsy and it’s crude, but it might serve as a guide overall.
Sky News keep their main stories in a searchable archive.
You can search for stories featuring “Miliband” and see how many you get for given day.
Like with this link
We can see how many stories Sky News has archived, searchable by “Miliband”, on a given day.
This generates a fairly “noisy” sample. But if we look at the 3-line summary of each story, we can count the ones that explicitly feature “Miliband” in the summary.
This lets us focus on the stories that feature Ed Miliband prominently, rather then Labour generally. A quick scan of the stories suggests that this might be a useable if crude proxy for “Ed Miliband on Television” – The stories are concentrated at times when Ed Miliband is on the news, either “intentionally” – when he has a big-ticket press conference or announcement – or when he’s naturally part of the story, such as the Falkirk scandal.
It’s only one channel, but Sky News coverage will tend to correlate with coverage from other TV channels. And it’s a written archive, not an actual TV archive, but it will still broadly match TV Coverage. So, we can tabulate Ed Miliband’s high-profile Sky News appearances
Let’s look at the Sun You Gov polling of Labour’s vote share – 4 samples a week.
Every time Ed is reported on Sky News, we’ll give the message a day to sink in, then we’ll look at the You Gov Polling starting in 2 days (You Gov generally poll over 2 days, so we’re looking at the first day of polling, not the publication day).
Our Polls are only available 4 days a week, and many days will show no change in Labour share – So we need to fill in the blanks. We can attribute for every single day, a value “CurrentOrPendingPollChange” – So on any given day, we can say what the next change in the polls will be, when it comes. “Current” if it’s today, pending if it’s the next change.
In the Spreadsheet (click here to view the spreadsheet), we can see the effect on the polling 2 days later.
And of the 80 days that ED was mentioned, the effect on Labours Polling 2 days later was
“Current Or Next Poll Change Positive:” 34
“Current Or Next Poll Change Negative:” 46
So – does this mean anything? It’s not anything like conclusive, but it does add weight to the suggestion that Ed M on television has more often, a negative rather than positive effect on Labour’s share of the vote.
If we accept that (and it’s a big “if”), it must give Labour something to think about with less that a year to the GE.