"Voters might make choices based upon where candidates name are listed on the ballot paper. Names higher up the ballot appear more likely to receive benefits from this split voting"
There are natural advantages some candidates have other others, especially in local elections. One such advantage is the incumbency effect where the defending candidate has an advantage over rivals. Another such theorised advantage is candidates higher up the paper gain votes due to the fact their names are seen first by the voters as they unfold their ballot papers. This theorised alphabet phenomenon has been supported by recent academic research papers and does appear to give a small advantage (Webber, Rallings & Thrasher 2014). This blog post uses the 2018 Hastings borough council elections as a case study to test if this theory has weight.
The 2018 Hastings borough council elections provide an interesting natural experiment with regards to testing if candidates at the top of the ballot paper perform better. In these local elections, the boundaries were re-drawn, meaning that every elected official was up for re-election. Hastings Borough council has the unusual system where there are two members for every ward. Therefore, this election produced two candidates for all of the four main parties contesting the election, Conservative, Labour, Lib-Dem and Green. There are 16 wards, giving a total of 64 candidates, 32 pairs, of which to test if the order of the surnames provided a pattern. If a pattern did occur it would be expected that all four parties most popular candidate would have a surname higher up the ballot paper than the second most popular candidate for at least 8 out of 16 wards.
The findings:
With the winning party of a constituency, this occurred 10 out of 16 times. The party that finished second place this occurred 14 out of 16 times. The party who finished 3rd displayed this pattern 9 out of 16 times. Finally, with the party who finished 4th, this occurred 10 out of 15 times. Therefore, it can be said an alphabetical pattern did emerge. There are many reasons why this could be the case. For example, it could be by chance those who had been long-standing councillors, or who had a higher profile due to having a cabinet position, just happen to have surnames close to A than to Z. Yet, as we shall go onto see there are some interesting patterns to show that this might not be the case. This post argues this occurred because of split voting where voters who wished to vote for two parties did so according to which candidate was higher up the ballot paper as voters do not always spend time focusing upon individual candidates and instead chose to vote for parties.
Interesting examples:
Table 1 shows the Ashdown ward results. Here every party follows surname order. The Conservative Party case is the most interesting, where the winning party appears to receive votes which have been split. Edwards beats Marlow-Eastwood, perhaps if the second-placed candidate had dropped the double-barrelled part of their name to Eastwood they would have received the most amount of votes. The Green Party received more second preference votes and the candidate higher up the ballot paper appears to have gained more of this vote than the other Green option. This was the same for Silverhill ward, where Colin Sinden beat O’ Callaghan. Perhaps if Nigel had not put his middle name onto the ballot paper, and therefore would have been lower down the order, he would have finished in second place. Again, the 3rd party beneficiary to this split vote appears to be the Green Party candidate who was higher up the paper due to the doubled barrelled nature of his surname.
Table 2 shows two examples where long-standing candidates, who should receive a generous incumbency effect, do not appear to receive this well-established bonus. Instead, newer candidates, in one case a first-time candidate, received more votes than their long-standing party ally. Therefore, the winning candidate further down the ballot paper again appears to lose out from individuals who decided to split their votes between two parties. Moreover, the smaller party candidate who most benefited from this splitting of the vote appears to be the candidate higher up the ballot paper, again indicating there is a benefit to being higher up the ballot paper.
The curious case of Old Hastings:
Old Hastings, table 3, provides a very interesting case. In this ward, there was much anger directed towards the local Labour council against a proposed new development along the Old Town marina. However, at the same time, there was also much support for the Labour councillor who was due for re-election, Cllr Bacon, who was recognised as working very hard for the area. Therefore, those who wanted to protest could still support their Labour candidate, whilst at the same time letting Labour know how they felt by opting for a Green candidate in their second vote. As seen in table three this was the case and was so in a way that reflects alphabetical order, again indicating that the candidate who benefitted from this split vote may have done so as their name was higher up the ballot paper.
Exceptions to the rule:
The two examples where being higher up the ballot paper did not translate into a bonus occurred in seats that contained high-profile candidates. Table four shows the two wards where the council leader and Labour Party parliamentary candidate were running for re-election. These candidates received more votes than their party ally, meaning that high-profile candidates may make any alphabet effect redundant, showing limitations to the benefits of being higher up the ballot paper.
In summary:
Overall, the results from the 2018 HBC elections show that when voters are presented with options that contain choices where they can split their vote, and unusual thing in the UK’s political system, they might make choices based upon where candidates name are listed on the ballot paper. Names higher up the ballot appear more likely to receive benefits from this split voting, whereas names lower down the order appear to suffer from this. My advice? If you run for local government change your name to A. Andrews.
References:
Webber, R., C. Rallings, G. Borisyuk, and M. Thrasher. 2014. “Ballot Order Positional Effects in British Local Elections, 1973-2011.” Parliamentary Affairs 67: 119-36. http://www.electionscentre.co.uk/?page_id=3196
Comments