The ultimate political reference book for amateur psephologists and political obsessives everywhere. From the team which brought you the Politico’s Guide to the General Election.
Posted on 13 May 2010 by admin
The ultimate political reference book for amateur psephologists and political obsessives everywhere. From the team which brought you the Politico’s Guide to the General Election.
Posted on 01 October 2009 by admin
The Tories are doing better in marginal seats than the national polls suggest, warns Lewis Baston
Even now, unbelievably, some Labour people seem to be complacent about the next general election. The argument goes that the Conservatives, because of electoral system bias against them, need to be 11-points clear of Labour in the national share of the vote to have a majority. This is true only if the swing is uniform, ie the same across the country. While uniform national swing is usually the best rule of thumb for translating poll figures into seats in the House of Commons, it is only an assumption, not a rule. For instance, Labour did significantly better in 2001 than uniform swing predicted because Labour MPs first elected in 1997 often boosted their majorities.
The local elections in June 2009 were a test of how far ahead the Tories really need to be to win an election. The ‘national equivalent vote’ of the parties (ie the local results translated into what they would mean in an election across the whole country) was, depending on whose projection you look at, the Conservatives on either 35% or 38% and Labour on 22% or 23%. This means a swing of 8% or 9% from Labour to Conservative, slightly more than the 7% they need to win a majority under the uniform swing assumption. Given that governments rarely repeat their worst mid-term performance in a general election, some people assume that an overall Conservative majority is unlikely.
The results in the key marginal constituencies where there were local elections in June should explode any such complacency. While the national swing appears to have been 8-9%, it is much higher in most of the marginals.
In the constituencies where more or less any swing will switch the seat to the Tories or LibDems, it seems about average – although 8% or 9% is easily enough to do the job. The ominous finding is from the constituencies where the Conservatives need a bit more of a swing to gain from Labour. In these cases the average swing is 13% or thereabouts, which would cut a swathe through Labour’s parliamentary representation. There were 61 Labour-held seats with county elections in June. Only four would have survived an election like the county elections. This is because the Conservatives seem to be getting the big swings where they need them.
In some of the target seats, the Conservatives are simply blowing Labour away – swings of 18% in South Ribble and 17% in Tamworth are extremely large by any comparison, and reflect a particular loss of support in areas where New Labour did particularly well in 1997. In others, Labour’s traditional vote has also melted away, as in Leicestershire North West where the BNP won what had been the safe Labour ward of Coalville, while the Conservatives have stood still or gained slightly. In this set of elections in the new towns, where Labour has done poorly for years in local elections, the swing may not appear quite so bad, but this often reflects the Conservatives losing votes to the right – UKIP, BNP and English Democrats – which might not help in general election conditions. Some coastal areas where Labour prospered in 1997 also have high swings – Dover, Morecambe and Waveney all have swings in the 15-16% bracket.
The Conservatives are not stupid in matters of political strategy, and know that they need either a 7%-plus national swing, or to do better in the marginals. They have focused their energies, campaigning messages and money (from Michael Ashcroft and elsewhere) on the marginals they need, and it seems to be paying dividends.
Local elections, although they are strong evidence, do not automatically reflect what would happen in a general election. People sometimes vote differently in local and national elections, and a different range of parties and candidates stand in each election. Turnout is also a lot lower, and the voters who stay at home in local elections but vote in general elections may not share the views of those who vote in council elections.
Labour needs to do two things in the short term – recover ground in the national polls, and raise its game in the marginal seats. In the longer term, Labour also needs to scrap an electoral system where pouring resources into a tiny number of seats can win party control over the government, and replace it with one where there is a genuine national dialogue.
Lewis Baston is from the Labour Campaign for Electoral Reform and author of Politico’s Guide to the General Election. To read the full research see here.
Posted on 07 November 2008 by admin
Glenrothes was a triumph for Labour. But to call a general election now would be folly
Labour’s win in Glenrothes is astonishing, all the more so for being so unexpected. The Scottish National Party were confident and Labour nervous even as the polls closed.
There are several ways in which this election was extremely good for Labour. The party not only increased its share of the vote, a rare enough feat for a party in government, particularly in a seat it already holds. Labour’s share of the vote increased by 3.2 percentage points. But it actually polled a few hundred more votes in absolute terms than it did in 2005, because there was only a small fall in turnout. There is simply no way of spinning that into anything other than a triumph.
The swing to the SNP was kept down to 5% in Glenrothes, even though in campaigning terms the party had thrown the kitchen sink at the constituency. A 5% swing might sound a reasonable result for the SNP until one looks at the party’s record in byelections. While the 22.6% swing in Glasgow East earlier this year was very high, 10% is usual. In contests between Labour and the SNP, there is a double-digit swing more or less whatever the political circumstances. This happened in Livingston in 2005, Paisley South in 1997, the two Paisley elections in 1990 and Falkirk West in 2000. The only previous occasion in recent years that the SNP has fallen short of a 10% swing in a two-party contest for a Labour seat was in Glasgow Anniesland, following the death of Donald Dewar in 2000. There are some similarities between that contest and this – it took place at almost exactly the same point in the parliament, during a Labour recovery from a low point (although 2000′s nadir around the fuel crisis was shallow and very short in comparison to the recent trough), and was prompted by the death of a well-respected local MP.
Another historical parallel reaches back a little further. The SNP dominated Scottish politics in the mid-1970s, surging in the two 1974 elections and scoring some impressive results in the 1977 district council elections, although the supply of byelections had dried up. When two Labour seats fell vacant in 1978, at Glasgow Garscadden and then at Hamilton, the SNP contested them vigorously but Labour easily fought off the competition, electing Donald Dewar and George Robertson to parliament.
While Glenrothes was an extremely good result for Labour, and for Gordon Brown, Labour should not be tempted by a quick general election. The campaign focused on the shortcomings of the SNP-run Fife council, which effectively put the SNP candidate and council leader Peter Grant on the defensive. A general election would focus discontent at the national level.
Fife also has a particular regard for Gordon Brown, a a local MP, and for people like Glenrothes’ new MP, Lindsay Roy. The result is extremely satisfying for him, and also for Sarah Brown – who seems to have decided after her turn at the party conference that she has a taste for political campaigning. However, it is fair to say that Fife’s warmth for the PM still runs ahead of the nation’s, even though Brown has recovered a lot of ground during the financial crisis.
While Fife council bore the brunt of the SNP’s new found and uncomfortable accountability, the Scottish government cannot escape some damage. While Alex Salmond’s government is not unpopular, exactly, the gloss has come off and the voters are in a more questioning mood about what it has delivered and where it is taking Scotland. For Salmond himself, there will be momentary embarrassment at having predicted an SNP victory and come a cropper, but confidence and swagger are part of his political persona and he can absorb being proved wrong every now and then.
The byelection, crucially, said nothing about how competitive the Conservatives are in their target seats in England and Wales, although it was more evidence that the Liberal Democrats’ strong vote in Scotland in 2005 will not be repeated next time. To call a general election now, when it is cold, dark and wet, the mood of the country is apprehensive, and the extent of Labour’s national recovery remains uncertain, would be folly. There is time to wait and see if the national polls improve further before the next feasible election date comes round in spring 2009. Labour should celebrate Glenrothes, and the palpable turning of the tide, but keep cool and not let election speculation run out of control, as they did in September 2007. Glenrothes reminded us that a Labour win at the election is possible (particularly if one counts being the largest party in a hung parliament as a win), but that is not to say it is probable.
Posted on 25 July 2008 by admin
With a high turnout and a dramatic result, the byelection was a disaster for Labour, but what does it mean for the other parties?
Glasgow East was an astonishing result. The SNP win in this seat, however squeaky the majority, was remarkable not just for its high swing, but for the relatively good turnout, the strong Labour candidate, and the previously intractable nature of Labour support in Glasgow’s east end. It seemed a solid Labour firewall, but in east Glasgow it is far from unknown for solid-looking structures to “go on fire”.
Scottish byelections in apparently safe Labour seats have often seen large swings from Labour to the SNP – 22% in Hamilton South in 1999, 19% in John Smith’s old seat of Monklands East in 1994, and 11% in Paisley South in 1997. In each of these years, Labour had a comfortable lead in the national polls. The scale of the swing often reflects the first serious campaigning in such a seat, and this must surely be the case in a seat that was considered as safe as Glasgow East. While all of these near misses faded rapidly, so too did the victories in Hamilton and Govan.
The 42% turnout in the byelection is pretty impressive. A loss of only six points since the 2005 general election suggests that the first serious campaign for a seat in the east end of Glasgow in the last 50 years has motivated voters. The low turnout in byelections between 1997 and 2001 presaged the steep fall in 2001. Following the small drop in Crewe and Nantwich, it is clear that byelections have started to bring voters to the polls and this suggests that turnout will rise at the general election.
A previous rule of thumb in west of Scotland byelections was that the Catholic Labour vote was more solid than the party’s support in “Protestant” seats like Govan and Monklands East (ie Airdrie). But Glasgow East rather disproves this. Labour’s nightmare must be that the Catholic Labour vote is going the way of the Protestant working-class Tory vote that used to be so strong in the west of Scotland before it collapsed between about 1960 and 1980.
I never shared the view that John Mason was an unconvincing candidate – people in a large chunk of the constituency, Baillieston ward, knew him and obviously liked him as he obtained an enormous personal vote in the 2007 council elections. Glasgow East was quite blessed with some good candidates – Mason and Margaret Curran obviously, and the also-ran Conservative and Lib Dem candidates Davena Rankin and Ian Robertson were worthy of notice and perhaps something better in future.
The Conservatives’ relatively good result in Glasgow East, unlike their wins in Crewe and Henley, should not be taken too seriously. If you are a Tory in Glasgow East, you are accustomed to adversity and the vote, although small, is very solid and remained so (on the slightly lower turnout) in the byelection. But the result is mildly encouraging. The Lib Dem vote proved, in the absence of much local organisation, squeezable.
Among the effects of Glasgow East will be to achieve something scarcely possible in politics – to make Alex Salmond even more self-satisfied. He took something of a risk in predicting victory in this byelection, and while he has the admirable quality of being able to brazen out over-optimistic remarks, triumph is not good for his political style. First minister’s questions is a preening, posturing spectacle that even a hardened observer of Westminster may find difficult to take. Sooner or later, the elastic will snap, as it has in the past for confident Scottish heroes such as Ally McLeod, who found adulation turned rapidly into – equally overdone – rejection in 1978.
The obvious point is that Glasgow East was a truly shocking result for Labour, a sign that the party is very deep in the trough, like the Major government from 1992 to 1997 and the Wilson government in 1967-69. Perhaps, like severe government defeats such as Dudley (both in 1968 and 1994) and Staffordshire South East (1996) it is a sign that the some sort of realignment is happening. The Tories flirted with disaster between 1993 and 2003, in that even some of their southern English strongholds were crumbling. Now, in Glasgow East, a Labour fortress has fallen.
Posted on 16 July 2008 by admin
Like Thomas Strathclyde, I am dismayed by the proposed Lords’ reforms, but it’s the issue of recall that is truly indefensible
There are many disappointing aspects to yesterday’s government white paper on the reform of the House of Lords, as well as some useful proposals. On this, at least, I am in agreement with Lord Strathclyde, although we identify different disappointments. He valiantly defends the indefensible proposition that a new second chamber would, by virtue of being elected by a system of first past the post, be dominated by the hand-picked candidates of the two largest political parties, almost regardless of how low their vote may fall, and tend to have a government majority. This would make for a poor revising, scrutinising chamber. Postponing any action until after the general election is also disappointing, not least because it minimises the chance of anything actually happening for another few decades, as is normal in the world of Lords’ reform. But the most surprising let-down in the small print of the white paper is the suggestion that members of the new Senate (or whatever it ends up being designated) might be subject to recall votes.
Recall is the procedure whereby an elected official can be subject to a special election, triggered by a petition signed by a proportion of the electorate demanding one. If the majority (or some specified proportion of the voters or the electorate) decide to give the incumbent the boot, then out he or she goes. Recall was originally part of the progressive government reform package in United States, in the early decades of the 20th century, along with primary elections and the direct election of senators. But in effect, recall is thoroughly reactionary rather than progressive. It gives well-organised and powerful interest groups the ability to knock out public figures who oppose their agenda, as shown by the most famous recall in recent years, the overthrow of governor Gray Davis of California in 2003.
Recall has other weird properties. It is perfectly possible, particularly if the bizarre suggestion of introducing first past the post becomes reality, for the replacement candidate to have fewer votes than the recalled incumbent.
Recall of members of the UK Senate would, according to the white paper, be only for serious dereliction of duty, corruption or misconduct, and perhaps only after five years in office. Recall is in the white paper for an understandable reason – that the members of the Senate would have long terms of office, of around 12 years, and that in extreme cases there should be some sanction to get rid of members who are not doing their duty. But there is a terrible dilemma. Too broad a definition of the grounds for recall will lead to politically motivated recalls. Too narrow, and it will either not be used or end up being an excuse for the Senate not to take action itself against errant members.
The white paper tentatively proposes a tighter set of rules about standards of conduct in the Senate than the current Lords’ rules, and an attendance requirement. Instead of recall, the new chamber should have a really rigorous set of rules, enforced by proper independent scrutiny bodies. It should also – as should the Commons – be subject to the criminal laws of fraud and corruption. Where serious cases arise, the legal process – with its standards of proof and definition of terms – is a more appropriate way of coping than a heated and potentially highly politicised recall vote. For lesser evils, recall seems excessive – although it should be made as hard as possible for the chamber to avoid effective sanctions against those who commit them.
Even if recall were hardly used in the Senate, and its inclusion motivated by concern about long terms of office, its enshrinement in British constitutional law would be a dangerous precedent. Recall is not a scalpel that cuts out only the moribund and corrupt from the body politic, it is a very blunt instrument. Even more than the referendum, it is vulnerable to exploitation by demagogues, anti-politics populism and the malice of bitter defeated candidates and special interests who are losing out in the policy process. Recall is purely negative – it is about accumulating discontents, not putting forward a positive agenda. There is too much of that already in political life.
Posted on 02 May 2006 by admin
A Labour rout on Thursday may owe more to the disillusion of its voters than a surge to the opposition. The thing to watch for is turnout.
Labour councillors nervously anticipating Thursday’s local elections must wonder how much harder the government could work to mess things up for them. One probably has to look back to 1968 to find a parallel. Devaluation, financial crisis, tax rises, spending cuts, Cabinet resignations and lurid press coverage of immigration and Enoch Powell’s rivers of blood were bad enough, but just before polling day the government put up NHS charges (despite having promised not to). The result was a massacre at the polls, with nearly 800 losses in London alone and many cities including Birmingham where Labour won not a single seat.
Local election results tend to go consistently against the party nationally in power, particularly when the government is a Labour government. Even during the honeymoon period of the 1998 local elections, the Labour lead was lower than in the 1997 general election or in the national polls. There is always a turnout differential that makes it difficult to get Labour supporters to the local polls while Labour hold office nationally even at the best of times. These are self-evidently not the best of times. Labour must expect a bad result, but how bad?
Assessing what is a reasonable benchmark for success or disaster is difficult and bedevilled both by the complexities of local elections and the expectations management practised by all the parties. Election night will see spin in its purest form, as each party claims to have out-performed what could reasonably be expected of it. Those with long memories will recall 1990, when Conservative success in Wandsworth and Westminster distracted attention from poor national results, and 1996 when the Tory disaster wasn’t quite as complete as the year before.
The most consequential measure of performance is in terms of council control. This matters because it gives (or takes away) a party’s ability to put policies into practice at a local level. Because every seat in London is up for election, this is where the most dramatic changes will take place. Labour did well from the electoral system in 2002 (the last time the seats were fought), winning 15 boroughs (including 4 where the party actually polled fewer votes than the Conservatives). Even before last week, their chances of holding Bexley and Hammersmith & Fulham looked vanishingly small, and it would be no surprise if Croydon and Merton also flipped to the Conservatives. Labour’s vote has eroded both in ethnically mixed areas and liberal middle class areas to the Lib Dems and others, and the loss of at least Brent (and quite possibly Hounslow, Camden and Tower Hamlets) is likely. If any of these boroughs survive under Labour control, it is a tribute to the local councillors’ management of services rather than an endorsement of the government’s recent record. It would be a less expected, and very serious, blow if any of Haringey, Ealing or Lewisham fell.
The Conservatives will be hoping to pick up the four boroughs from Labour, and also take another three where they didn’t quite make it in 2002 (Harrow, Havering and Hillingdon). There is also an interesting confrontation with the Lib Dems in three middle class south-west London boroughs – Kingston and Sutton are run by the Lib Dems and Richmond by the Conservatives, but there is a lively contest in all three. If the Lib Dems carry off Richmond and defend the other two, they will gain in confidence about repelling the electoral challenge of David Cameron’s liberal conservatism.
Outside London, fewer changes are likely – Labour did so badly in 2004 that even unexpectedly good results would not be enough to recapture power in cities such as Newcastle, Leeds and Birmingham. Labour control of Derby and Newcastle-under-Lyme hangs by a thread, but it would take a real meltdown to lose Manchester.
The measure of party performance that will probably attract most attention is the net Labour loss of seats over the night. In assessing what this might be, it is important to disentangle the different starting points. If Labour were to do pretty much as badly as in 2004, which was a rotten local election year for the party, the party would lose something of the order of 350 seats. This would be composed of no change in the metropolitan boroughs (because these seats were last contested in 2004) and substantial losses in the areas last fought in 2002. There was a swing of 5 or 6% away from Labour in the metropolitan boroughs between 2002 and 2004, and if all that happens is that the London boroughs catch up with this movement London alone would produce a loss of about 200 Labour seats. Adding in the seats in the district and unitary authorities which Labour are defending from 2002 and comparison with results in 2004 is possible, Labour would lose 131 more seats. Labour losses of 300-350 seats would therefore signal an overall result on a par with 2004.
Although 2004 was a bad result for Labour, the party went on to win a general election with a perfectly adequate majority only 11 months later. In the current circumstances, a similar result should be the occasion for a certain amount of relief and it certainly wouldn’t count as “meltdown”. Given the recent hellish run of bad publicity for the government, something a little worse than 2004 should be expected – perhaps total losses around the 420 mark. Anything much over that would be a sign that Labour’s position is a lot worse than in any other set of local elections under this government, and put its future into question.
The gains are unlikely to all be in one direction – the Conservatives will have to share the spoils with the Lib Dems and a host of others – Greens, Ukip, the BNP, independents and a variety of local and single-issue parties. If the Conservatives are more than 250 up, they will have done well.
The better measure of how well or badly the parties are doing nationally is their share of the vote. One indicator of this is the national equivalent vote share projection that the broadcasters will do on election night, but that is at best approximate and trends may be distorted by the dominance of London and other urban areas in this round of elections. Labour came third in this measure in 2004 and must be braced to do so again.
However, the detailed voting numbers are more interesting and reliable. In every election in London since 1994 – for parliament, boroughs, mayor, Europe and Assembly – the Conservatives have been more or less flatlining on around 30% and have varied only between 27% (2004 Euro election) and 34% (2002 London boroughs). If the Conservatives break out of this range into the high 30s, they can claim to be making real progress in the capital, and if they get over 40% in London they are entitled to savour a triumph. Labour’s vote, on the other hand, has fluctuated wildly – although third place, or anything under 25%, would be bad news.
Although the metropolitan boroughs are unlikely to see much drama in terms of seats and councils changing hands, the share of the vote will be interesting. The Conservative share has been incredibly stable in these elections, holding steady at 26% in every set of borough elections since 1998 with the exception of William Hague’s best year, 2000, when they won 31%. If the Conservatives are still stuck on 26%, this would be a disappointing result for Cameron, but anything above 30% would be pretty good, as would overtaking Labour. It is also possible, but a long shot, that the Lib Dems could win the largest share of the vote in the metropolitan authorities this year – the results in 2004 had Labour on only 33% and both Lib Dems and Conservatives on 26%.
There are several notes of caution to enter about interpreting local elections. One is that local issues do matter, and seem to be increasing in importance in recent years. Some councils (such as Conservative Wandsworth) have a good relationship with their electorates and seem insulated from national trends. Others fall foul of local issues and suffer the consequences, as the Plymouth Conservatives did in 2003 when they lost badly despite a favourable national trend.
The decay of the traditional system is more advanced locally than nationally. The Lib Dems have long capitalised on local issues and done better than expected in local elections, even in areas such as Southampton that tend to be Lab-Con fights at general elections. Smaller parties and independents can also expect to do well. In some areas local politics has become bewilderingly plural, for instance in Kirklees in Yorkshire where no party won more than 25% of the vote in 2004.
Variations in local election turnout can be important. A Labour rout on Thursday may owe more to the disillusion and alienation of Labour voters than any great surge to the opposition. If turnout drops significantly from the 33% reached in urban areas in 2002, it is questionable how significant the gains of the other parties will turn out to be. If, however, turnout is in the high 30s and Labour are trounced, then the government should be seriously worried about a real voter revolt.
The multi-party nature of modern politics will save Labour from a 1968-style wipe-out even if the party’s vote is just as low. Back then, voters who turned against Labour went by default to the Conservatives, but now there is more choice for the disaffected and local elections are rarely as uniform. There may be a few crumbs of comfort for Labour in the actual results – but, ironically, the party has probably got the national spin wrong. As noted in the Guardian today: “If Mr Blair loses more than 200 seats nationally he will be in serious trouble.” If he loses only 200, he’s probably actually in better shape with the voters than he was two years ago.