Is AV fairer than our current system?
No - AV is not a proportional system and can lead to more disproportionate results than our current system. Research suggests that in 3 out of the last 4 UK elections AV would have produced more disproportionate results.
It is surprising to see other electoral reformers so readily endorsing AV. Most electoral reformers, including AV2011, believe proportional representation to be the best democratic solution and yet AV is not a proportional system. Even the Electoral Reform Society (see its report A Better Alternative?) claims that AV tends to produce more proportional outcomes than our current voting system. This is mistaken. There is no evidence that AV is more proportional. In fact, the evidence suggests that the two systems are similarly disproportional.
By forecasting how recent elections may have turned out under AV (by use of voter surveys at each election), we can compare the disproportionality of our current system with the projected AV results. Political scientists measure the disproportionality of elections by the use of an index. If the index is zero we have perfect alignment of seats to votes; the higher the index the more disproportional the election. PR systems tend to have a value of 4.0 or less. The table below shows the index for the past four UK elections:
| Election | Current System | Projected AV | AV Better? |
|---|---|---|---|
| 2010 | 14.6 | 11.9 | Yes |
| 2005 | 13.1 | 15.1 | No |
| 2001 | 17.8 | 19.0 | No |
| 1997 | 16.5 | 22.3 | No |
| Average | 15.5 | 17.1 | No |
The indices are calculated using the Gallagher Least Squares method for the three main parties only.
If we go even further back, to 1983, the average indices are virtually identical: 16.27 and 16.34. However, these forecasts are only as good as the survey data on which they are based. Is there any real data we can use? The only major democracy to use AV for its national elections is Australia, which has strong two-party politics. In the UK, since the mid-1970s, we have had a three party system. If we take account of this and then compare Australia and the UK what do we find? Statistically, Australian elections under AV are no more proportional than the UK.
In addition, if we look at regional elections in Western Canada and Queensland we find extraordinary disproportionality with AV. Two examples: in 1948 the Social Credit party in Alberta hit the jackpot when it scooped 97% of the seats by winning 53% of first preference votes (in the 47 constituencies that used AV); and in 2004 the ALP in Queensland grabbed 71% of the seats on 47% of first preferences.
AV is no more proportional than our current system. Those who seek a proportional system should seek alternatives to the Alternative Vote. See A Plan For Reform.
