The Optical Illusion of Net Ratings
"The Absurdity"
We asked a hundred people - "Who do you trust to buy a round for their friends in the pub, Boris Johnson or Sir Keir Starmer?" 42 said Boris, 24 Sir Keir, & the rest didn't know. They were asked who they thought would best organise a fun night out;. 42 said Boris, & 20 Sir Keir. In June 2020, another survey asked if the respondents thought either of the two "had personality" - Boris won by 34 points, 64 to 30. When asked again in September Boris increased his lead to 42, (67-25). The most recent Opinium poll asked whether the respondents found Boris or Sir Keir likeable - given that people would rather go for a night out with Boris, the choice they think has bags more personality of the two, it's not much of a surprise that he "got more likes" 43-38. So it would seem uncontroversial to say that Boris is the more likeable of the two men who want to be PM after the next GE. Unless...
Unless, instead of basing the results on who people actually like, you decided to ask who they disliked as well, and subtracted one from the other. Using this method, Sir Keir suddenly becomes more likeable than the more liked Boris, braver than the man more people consider brave of the two, and a stronger leader than the leader he trails in terms of people's judgement of who is strong. (Opinium March 2020)
Collectively these absurdities are one example why the use of net ratings rather than gross positives is a redder herring than a herring who just won first prize in a "Reddest Herring of the Year" competition.
"The Don't Know/Don't Care's"
The pandemic has made the last twelve months a strange old time in politics. Starmer took over the Labour Party leadership at the start of it, and his, and his party's, popularity has ebbed and flowed in line with the level of despair in the country - after a while of the government locking us in our houses whilst the news told us thousands were dying each day, Sir Keir & Labour polled well; when a Covid vaccine was deployed and the government announced a plan to let us live more normally in the near future, those ratings nosedived. What it gave him though, was the best opportunity a LotO has ever had to endear himself to the general public. Politics dominated the media, Sir Keir was given prime time slots to respond to the Prime Minister's Covid plans, the Conservatives refused to appear on ITV's breakfast show, leaving the door wide open for the opposition when all there was to do was watch tv and look at the internet; an unprecedented chance to woo the people he needs to get the keys to No10. Despite all this free publicity, in the latest Opinium Leadership Ratings, 37% of respondents didn't have a view on him one way or the other. This should depress Labour supporters, but the optical illusion of net ratings turns indifference into a positive, and makes the blanket comforting, not wet. It means Boris leads by just two point in terms of net satisfaction (45-38 playing Sir Keir's 34-29) - all to play for, "especially after he's just had the vaccine bounce!" cry Starmer's cheerleaders.
Well, no.
The smoke and mirrors of Net Ratings make this seems a close race, and Sir Keir still has 38% of the public to win over! But remember that only 70% of those with the vote use it, and the mist clears - the divisiveness of Boris Johnson is an asset, the blandness of Sir Keir a drag. Divide the Gross Positives by 70 and Boris is 15 points clear - the Don't Know's/Don't Care's don't vote; you want as few of them as possible.
"The Voting System"
In elections, only positive votes count - the strength of feeling against you is irrelevant so long as you get more votes than your competition, and so Voting Intention polls only ask respondents who they will vote for, not who they won't... because the second question doesn't matter. Lets imagine such a poll were taken, the question asked "Would you consider voting for the following party; Yes, No or Don't Know" and the results were as follows
Yes No DK Net
Con 30 70 0 -40
Lab 23 52 25 -29
LD 8 45 47 -37
SNP 3 38 59 -35
Green 2 40 58 -38
Brexit 1 48 51 -47
Only The Brexit Party are less popular than the Conservatives in net terms.
But of course, we have forgotten that 30% of voters, won't vote!So divide the positive number for each party by 70 (Possible voters minus stay at homes) and the result of an Election would be
Con 43%
Lab 33 %
LD 11%
SNP 4%
Green 3%
Brexit 1%
Of course there is a bit more to it than this. There is overlap -some people like Boris and Keir, some dislike both -some don't knows do vote. Better staticians than I will find flaws, and I am happy to listen to their corrections; this is a work in progress. But in the main, a system that rewards positive indicators, reflecting the electoral system, must be better than one that manages to find that people who are considered braver, stronger and more likeable, trail their less brave less strong, less likeable rivals