Labour's Heartland Revival

MORI's monthly poll for The Times published last week received a lot of attention and secondary reporting, mostly concentrating on the four-point drop in the Conservative voting intention share, to its lowest level since April. But the finding that may be by far the most politically significant was much less widely reported. Over the last four months there has been what may amount to a sea-change in the attitudes of Labour supporters to voting. One reading of the figures would suggest it may presage the difference between significant Tory gains that may secure William Hague's position to fight a second election, and another Labour landslide.

Labour has led in the polls throughout the Parliament, but in the last 18 months the Tories have been able to crow at some high-profile election victories, in the European elections and across England in local government. What all these victories have had in common, however, is that they were achieved on very low turnouts, with Tory supporters being disproportionately prepared to go to the polling booth. Commentators plausibly linked this to disillusionment in Labour's heartlands, and the polls backed up this diagnosis and suggested that a similar effect might work to the Conservatives' advantage in the next general election. Labour supporters, when asked, were substantially less likely to say that they were certain, or even very likely, to vote at the next election.

As recently as March, two-thirds of Conservatives, 66%, said that they were certain to vote, but only 54% of Labour's adherents said the same. This is a very substantial difference, and meant that although Labour led by 48% to 29% on the voting intentions of all adults, the lead was cut to 45%:33% among those "certain to vote". (MORI's past experience has suggested that "certain to vote" on election eve is much the best predictor of eventual turnout). The difference between these two figures is seven points in the lead, equivalent to a swing of 3.5%. To put that figure in perspective, that is bigger than the national swing in 10 of the last 14 general elections. Put another way, the effect of the apparent increase in Tory support which was shown in the polls in the spring and early summer would have been doubled by the effects of differential turnout.

But our newest poll shows that that differential has almost melted away. In August, 72% of Conservatives said that they were certain to vote, but so did 70% of Labour's supporters, a difference so small that the, rounded to the nearest whole number, the voting intention among those certain to vote is the same as that among all adults. The ace up Mr Hague's sleeve has unaccountably turned into a joker.

So what has changed? It is clear that something has enthused Labour's supporters into feeling that their vote is worthwhile and, assuming that this is not simply a wave of panic sweeping the country that it might get a 14-pint-a-day Prime Minister if it doesn't vote against him, it is presumably a reaction to the government's policy initiatives in general and probably the spending review in particular. This certainly seems plausible if we note where the biggest increases in vote certainty since March have taken place.

Q Can you tell me how likely you are to get along to vote at the next election?

23-28 March 2000

160	Voting Intention			Constituency
160	Con	Lab	LD	Con held	Lab held	Lab marg	Lab medium	Lab core
160	%	%	%	%	%	%	%	%
Certain not/not very/quite	16	24	16	25	35	32	33	40
Very likely	18	21	23	16	16	15	15	17
Certain to vote	66	54	61	58	45	52	50	37
Don't know	*	1	*	1	4	1	2	6

17-21 August 2000

160	Voting Intention			Constituency
160	Con	Lab	LD	Con held	Lab held	Lab marg	Lab medium	Lab core
160	%	%	%	%	%	%	%	%
Certain not/not very/quite	13	17	17	26	27	31	22	33
Very likely	14	12	11	14	10	15	9	8
Certain to vote	72	70	72	57	61	52	67	54
Don't know	1	1	0	3	2	2	2	5

There has been no significant change in Conservative-held seats, where 57% are now certain to vote, as were 58% in March; nor has there been in Labour "marginal" seats on the conventional definition (a lead of 10% or less, seats which the Conservatives will recapture easily if they achieve even a respectable defeat nationally, let alone a victory), where the figure has stayed steady at 52% certain to vote. But in Labour's safer seats, there has been a sharp jump: from 50% to 67% in the "medium" seats, where Labour's lead is between 10% and 35%, and from 37% to 54% in Labour's core heartland seats, where the majority is more than 35%. If the rise had been only in the medium seats, we might have assumed that it was simply a reaction to Tory resurgence - these are mostly seats that have been won by the Tories at some time in the past ten years, and will be the battleground in a close election, so votes here might matter if the Tories are back on the winning trail. But with an equal increase in the heartlands, it seems clear that Labour is getting through to its supporters who, up until a few months ago, were showing strongly stay-at-home tendencies.

The proportion of the whole electorate who say that they are certain to vote is now 59%, higher than the 55% who said they were in September 1996 (7 months before the 1997 election, probably about the same period, give or take a few weeks, that we have to wait until the next). So perhaps we will not, as some pundits seem to expect, have an even lower turnout in 2001 than in 1997; perhaps Labour's supporters will turn out after all. That looks like very bad news for the Tories - far worse than what is, perhaps, a temporary setback to the recovery of their own support.