Ordering effects, luck & rationality

I point out here that people's choices can be sensitive to arbitary differences in the order in which options appear. In the Eurovision song contest, for example, the first or later performers have more chance of winning than those appearing in the middle of the show. I omitted to add that this ordering effect also matters in politics. A new paper concludes:

Alphabetic ordering effects exist in the 1977 2011 Irish general elections. The effect is significant, in both a statistical and substantive sense. The estimated effect of being listed first on an alphabetical ballot paper in an Irish general election is approximately 544 first preference votes or 1.27 percentage points for the average candidate.

This corroborates findings from the US (pdf) and Australia (pdf).

Experimental evidence (pdf) suggests such an effect might exist in UK council elections if not general elections.However, it's worth noting that 191 of our 650 MPs (29.4%) have names beginning with A-E, compared to only 54 (8.3%) with names beginning U-Z. 

And it's not just in politics that an early name has an advantage; in football, acting and academia (pdf), early names are more likely to scoop big prizes.

All this might sound quirky. But it's not. It has two implications.

One is that it confirms the message of Ed Smith's Luck. Success depends in (large?) part not (just) upon ability and hard work, but upon fortune. Clearly, the successful have an interest in denying this. But if luck operates through subtle, smallish and non-obvious mechanisms such as naming effects, they might not be fully aware of it. 

Secondly, ballot order effects suggest that electoral outcomes are not entirely the rational will of the people, but are also influenced by luck. And if something as arbitrary as alphabetical order influences elections, how many other arbitrary or non-rational factors might also do so? 

Comments

[Chris]

'However, it's worth noting that 191 of our 650 MPs (29.4%) have names beginning with A-E, compared to only 54 (8.3%) with names beginning U-Z.'

Without information on the prevalence of names beginning with letters A-E or U-Z in society as a whole, the above quote is not particularly informative. It may be the case that there are simply that many more people with names beginning with letters A-E rather than U-Z generally.

[chris]

@ Chris - such information is hard to find. Judging by the Leicester phone book, A-Es are 27.6% of the population, and U-Zs are 9.7%, suggesting a slight bias to early names.

[Account Deleted]

The implication is that a subset of voters make their mind up with pencil in hand, and that this favours those at the top of the ballot. (The solution is quite simple: print the names in random order on the ballot papers - it's amazing what computers can do these days.)

While the idea that votes are cast on an arbitrary basis may offend some, I'm cheered to think that there are those who consider it responsible to vote even when they have no initial preference. We should never give an inch to the idea that there are some of us who are not fit to vote.

That said, I'm a staunch supporter of all that lies between A and E, so I don't have a big problem with this anyway.

[Edward Harkins]

I seem to recall that in one of the early -post-devolution elections in Scotland, the SNP leader Alex Salmond garnered significant advantage from the use of his name - rather than just the party name - on the ballot card.

The advantage was perceived to have been significant enough for a change in the electoral rules to be imposed. This reform was aimed at re-instating more of a 'randomness' in the listing on ballot cards.

[Bruce]

"@ Chris - such information is hard to find. Judging by the Leicester phone book, A-Es are 27.6% of the population, and U-Zs are 9.7%, suggesting a slight bias to early names."

Could this be evolution at play? The A-Es, with that inate advantage, have been more successfully, attracted more mates, had more children and so over time have come to dominate, while the end of the alpabet has withered...?

[Ilham]

I would like to comment on the Gwyneth Paltrow YOU TUBE voebfromatroad. I don’t really care about the privileged Gwyneth and I REALLY don’t care about who she votes for, but I do care about the military overseas.They have been neglected in regards to voting. If Gwyneth actually cared about everyone voting, not just her candidate, she would try to help the military get a chance to vote.Read this and decide. I have checked the facts and theyr'e right on.WASHINGTON, D.C. — Rep. Roy Blunt, the House Republican whip, on July 8 introduced a resolution demanding that the Defense Department better enable U.S. military personnel overseas to vote in the November elections. That act was followed by silence. Democrats normally leap on an opportunity to find fault with the Bush Pentagon. But not a single Democrat joined Blunt as a co-sponsor, and an all-Republican proposal cannot pass in the Democratic-controlled House. Analysis by the federal Election Assistance Commission, rejecting inflated Defense Department voting claims, estimated overseas and absentee military voting for the 2006 midterm elections at a disgracefully low 5.5 percent. The quality of voting statistics is so poor that there is no way to tell how many of the slightly over 330,000 votes actually were sent in by the absentee military voters and their dependents and how many by civilian Americans living abroad — 6 million all total.Nobody who has studied the question objectively sees any improvement since 2006, and that is a scandal. Retired U.S. Marine Corps Capt. Charles Henry wrote in the July issue of the U.S. Naval Institute Proceedings: “While virtually everyone involved … seems to agree that military people deserve at least equal opportunity when it comes to having their votes counted, indications are that in November 2008, many thousands of service members who try to vote will do so in vain.”So Gwyneth you be sure to vote!

[Federica]

It's because vontig is not compulsory in the USA. So their typical voter turnout is 30-40%. Last time people were amazed that there was more than 50% vontig.It works so well in Australia because we know everyone's going to turn up (well 98-99% of registered voters) so we prepare. Why would the Yanks bother setting up so many polling places for only 30% of those elegible?Their system doesn't seem very democratic from an Australian perspective where the pollies are always claiming a mandate. The US has a first past the post system so, if there's a couple of 3rd party candidates, the winner might get only 40% of the votes cast by that 30% of the elegible population hardly a mandate for anything.