How should we assess the local election results when we have a sense of things on Friday morning? The gains/ losses figures are the most popular measure as far as the media is concerned, because perhaps the best and most comparable measure – National Equivalent Vote share – is complex to calculate and can be worked out in more ways than one.
Benchmarks for local election gains and losses vary year by year because:
- Different numbers of seats are available each year; for instance there were 4,104 seats in the 2008 locals (the baseline for 2012) and over 10,000 in the big year for English local elections that is the 2007/11 year of the cycle. 2012 is a fairly small year for seat numbers, so that massive gains and losses like Labour’s haul of around 2000 gains in 1995 are impossible.
- Different sorts of area are contested in different years – Scotland and Wales this year, other years England only, and in general terms rural England in 2011 and 2013, and urban England in 2011 and 2012.
- Different starting points. This year’s starting point (except in Scotland) is 2008, which was a very good local election year for the Conservatives. Labour therefore have to win a biggish number of seats to be respectable, and the Conservatives can afford to shed a number of seats won at their high water mark.
- Electoral system – the seats up in 2011 (and 2014 in London boroughs) will tend to magnify changes, because the multi-member first past the post system often involves large turnovers of seats if there is a swing. The STV PR system used in Scotland will produce smaller changes for a given swing.
- Expectation management. This works in a couple of ways. The most obvious is that parties will try to under-claim their expected gains or exaggerate potential losses, to make the results on the day look ‘better than expected’. The more subtle is that they are affected by opinion polls and the general climate. A party that is, say, 10 points ahead in the national polls but whose local results are in line with a 5-point lead may be said to have had a ‘disappointing’ set of results, even though they are ‘objectively’ better than those of a party that is level in the polls but gets a two point lead on the local result and claims a triumph.
Figures for gains and losses are fuzzy around the edges, for several reasons.
They are affected by a few or more councils each year having boundary changes and therefore not being comparable with past years (this year: Hartlepool, Rugby, Daventry, Broxbourne) and affecting the overall party numbers.
Another factor is by-elections; some of the changes, particularly in a cycle like 2008/12 when there has been a big change in political climate, will have been discounted because the seats have already changed hands in by-elections. Take Walsall, for instance, where two wards that were Conservative in 2008 have already been won by Labour in by-elections. There are also defections, which tend to pattern wider political trends and can be important in some individual authorities – for instance the collapse of the Liberal Democrats in Rochdale or the Conservative split in Sefton. In some such cases a party can make apparent ‘gains’ by just recapturing seats from defectors. Another complication is sometimes that by-elections run concurrently with the local elections. In Bolton for instance there is a triple vacancy in a safe Conservative area, Bradshaw, giving the Conservatives an apparent extra 2 seats.
Seat totals are also skewed against urban England. The number of electors in each seat is much smaller in rural and suburban authorities, and seats are particularly small in the most rural areas. In terms of seats coming up in 2012, Claverdon ward (Stratford on Avon DC, Warwickshire) with fewer than 2,000 electors is weighted the same as over 20,000 in some Birmingham wards. A party doing differentially well in big cities and poorly in smaller towns and rural areas (as Labour did in 2011) will find its seat gains looking unflattering.
With all these warnings, the starting point needs to be the previous set of local elections in England and Wales in 2008. This was a very good Conservative year and a Labour disaster (national equivalent vote Conservative 43 per cent, Labour 24 per cent and Lib Dem 23 per cent). The Conservatives gained 300 seats and Labour lost 434. Reversing these figures would more or less restore the position that existed in 2004. Allowing for the fact that there are fewer seats available (large councils in Cheshire and Durham were contested in 2008), by-elections, and a possible slippage in Scotland, a Labour gain overall of around 400 would be something like a repeat of 2004. Now this would be a poor result. The local elections in 2004 were not a good year for Labour, with urban voters in particular giving the party a ‘kicking’ over the Iraq war and general ennui, and the Tories picking up in a number of marginals like Tamworth and Swindon. It was not a drubbing like 2008, but it was still a defeat. Translated into a national vote share, the Conservatives were ahead with 37 per cent, Lib Dems on 27 per cent and Labour on 26 per cent.
So, a net Labour gain of fewer than 400 seats is bad, particularly as a repeat of the 2011 local elections would mean a net gain of over 250 seats in the 36 metropolitan boroughs alone, another 70 in the 16 comparable unitary authorities, and about another 130 (very approximately) in comparable district councils. Fewer than 400 net gains in England would suggest slippage from 2011 (Labour can expect to make more gains in Wales than losses suffered in Scotland).
A ‘par’ result for Labour would involve doing a bit better than 2011 in urban England, consistent with a clear (but not landslide) lead over the Conservatives in national vote share, and picking up considerably in Wales while slipping a bit in Scotland. Perhaps about 100 gains in Wales, 500 in England, and 30 losses in Scotland, gives us a net gain of 570. (Having examined the local detail, I have revised this down a bit).
Translating recent trends in by-elections and polls, the leading experts in local elections, Colin Rallings and Michael Thrasher of Plymouth University, project about 700 Labour gains which feels a bit high. Anything from 550 to 700 should probably be regarded as more or less par.
Over 700 is therefore extremely good for Labour. The party cannot exceed this total much, because its ability to win seats is already ‘maxed out’ in many areas on the basis of the 2011 results (e.g. Manchester) or would be on a modest swing from 2011, and there are simply not all that many seats available. Also bear in mind the caveats about seats that have already switched in by-elections.
The Conservative picture is a mirror image of Labour’s. They must expect losses to Labour, and unlike in 2011 there will not be many places where they can make compensating gains from Liberal Democrats and Independents.
A repeat of 2011 in England would involve net Conservative losses of the order of 180-200 seats. They should hold reasonably steady in Scotland, and lose a bit in Wales – perhaps 30 seats? A net loss of 200 must be accounted a success for the Conservatives if they manage it, given the poor context for the 2012 elections.
A par result would be a bit worse than this (300 odd), and a genuinely bad result would be something like 400 down.
The Liberal Democrats must expect to lose, possibly very severely. A repeat of 2011 would involve around 280 losses in England, perhaps 30 in Wales and something of a caning in Scotland, losing around half of their 166 seats from 2007. This adds up to a little short of 500 for the expected level of losses, assuming that they are pretty much where they were in May 2011. If they have recovered a bit since then, though local activism or a hint of recovery in the national position, they will do better. Because a repeat of 2011 would put them at rock bottom in many areas (the reverse phenomenon to Labour being maxed out in the big cities), there is not a lot of downside risk for the Lib Dems – the main forecast is pretty brutal. Much worse than this and they are still losing support compared to 2011.