What do Londoners' second choices tell us?
The London mayoral elections constitute the biggest exercise in "personality politics" in the United Kingdom. They have so far encouraged between 1.7 to 2.4 million voters to participate in them.
But in another respect they are absolutely unique. Because they are counted by electronic imaging of individual ballot papers, we are able to track where every second preference choice goes to and comes from.
For those interested in voting statistics and trends this is the equivalent of gold dust, showing us the decisions of real voters rather than the hypothetical choices of respondents when answering questions about their second preferences in opinion polls.
While this information on how people cast their second preference votes is invaluable, it is, of course, important to remember that very few of these second votes played any part in the final outcome of the election.
Under the Supplementary Vote system used in the London mayoral elections, if no-one wins 50% or more votes at the first count, then only the top two candidates remain for the second round and all other candidates are eliminated.
It is only the second-preference votes of these eliminated candidates that are redistributed in the second round. Therefore, the 1.44 million second preference votes cast in 2012 by people who voted for Conservative Boris Johnson and Labour's Ken Livingstone with their first vote are not counted in the second round.
In 2012, these 1.44m second-preference votes cast by supporters of Johnson and Livingstone amounted to 82% of all second preference votes cast in the election.
This data from the May 2012 election has just been released by London Elects (the body that administers the mayoral and assembly elections).
So what are the main points to emerge from 2012 and the significant changes from comparable data for the 2004 and 2008 London Mayoral elections?
Firstly, the number of voters who choose not to cast a second-preference vote at all continues to rise: in 2004, this group amounted to 14.6% of all voters: in 2008, it was 17.0%; and in 2012, the figure rose to 20.2%.
It is also worth noting that 65,712 people whose first choice went to Boris Johnson, voted for him again as second choice, while 22% of Ken Livingstone's backers - 105,117 - gave him their second choice as well.
Coalition impact on the Liberal Democrats
It appears, following the Lib Dem decision to join the coalition government with the Conservatives, that fewer Labour voters gave their second preferences to the party. In 2004, 40.4% of Livingstone's second-preference votes went to the Lib Dems: in 2008, the figure was 36.3%: in 2012 it was 18.5%. Overall, the impact of joining the coalition by the Lib Dems can also be assessed by the total volume of second preferences they received. In 2004, the Lib Dems received 29.3% of all second preferences cast: in 2008, this share rose to 32% but by 2102, it fell to 20.6%.
Are UKIP supporters simply Conservatives voters on holiday?
In 2012, 15% of Johnson voters cast their second preferences for UKIP, compared with 9.8% in 2008. And these 2012 second preferences from Johnson supporters accounted for 70% of all second preferences received by the UKIP candidate. However, when we look at the second-preference choices of UKIP voters in 2012, we discover that 37.3% went to Johnson and 18% to the BNP. So far, so not too surprising. Yet Livingstone and the Greens received some 11% each of UKIP second preferences; and the Lib Dems received 8%. In fact the Conservative candidate received only one-third of UKIP second preference votes in 2004 and 2008.
Whereas in 2004, Green second preferences split 36.2% to Labour and 25.2% to the Lib Dems, in 2008 some 50% went to Labour and 18.8% to the Lib Dems. By 2012, 50% once again went to Labour but only 14.6% to the Lib Dems (almost the same number as those giving their second-preference votes to the Conservatives - 14.4%).
In 2004, when the London mayoral election was held on the same day as elections to the European Parliament, 44.6% of BNP second-preference votes were cast for UKIP and 20% for the Conservatives. In 2008, 36.2% went to the Conservatives, 17.8% to UKIP and 13.8% to the English Democrat candidate. By 2012, 29.9% went to the Conservatives and 29.8% to UKIP.
Siobhan Benita, Independent
The second preferences from Ms Benita's supporters were fairly evenly split between Johnson, 28.3%, and Livingstone, 30.9%. Almost one in five, 19.4%, went to the Greens and 13.1% to the Lib Dems. In return, the second-preference votes she received were also drawn from the supporters of these four parties - ranging from one in seven Johnson voters to, broadly one in 10 Labour, Green and Lib Dem voters.