Jargon and buzzwords are all over election coverage like a cheap suit. Here is hopefully a helpful glossary to some of the terms you’ll encounter.


Basildon – Was a constituency in Essex that has been a bellweather since 1974 and since 2010 has formed part of the Basildon and Billericay constituency.

Earned a place in British electoral history at the 1992 General Election. At this time still a bellweather, it was a constituency with a high proportion of lower middle class voters and home of the ‘Essex Man’, whose support was seen as crucial in keeping the Tories in power between 1979 and 1997.

Defending a majority of 2,649, sitting Conservative MP David Amess (of cake fame) held off the Labour challenge fairly comfortably. The declaration of this result became a defining moment of Election Night 1992 as it made it clear the Tories were to win another majority.

During every election journalists are keen to look for a ‘Basildon’. They found one in 2015 in the Midlands marginal of Nuneaton, where a Tory defence there made it clear that Ed Miliband was not to be Prime Minister.

(see also – ‘Bellweather’, ‘Essex Man’, ‘Marginal’, ‘Portillo Moment’) 

Bellweather – A constituency that always s votes the same way as the country. E.g., x constituency voted Labour in 2005, who won the most seats nationwide and voted for the Conservatives in 2010, who won the most seats nationwide.

Bellweather constituencies are less cast-iron than bellweather states in US presidential elections like Ohio. This is because constituencies are subject to regular reviews and boundary changes, meaning its demographics constantly change.

The longest running bellweather in the UK is Dartford in Kent, which has voted for the largest party at every election since 1964.

(see also – ‘Marginal’, ‘Basildon’)


Churn – The movement of voters between different parties. In a country that is obsessed with the concept of ‘swing’, people often assume that voters only move in one direction to another.

E.g. 2010 result in Burnley 2015 result in Burnley

       LD 35.7%                         LAB 37.6%  (+6.3%)

       LAB 31.3%                      LD 29.5%  (-6.2%)

       CON 16.6%                     UKIP 17.3% (+15.1%)

       UKIP 2.2%                      CON 13.5% (-3.1%)

       OTH: 15%                       OTH 2.1% (-12.9%)

Here many would be very quick to assume “Ah easy! Look the Lib Dems fell by almost the same as Labour rose, clearly all they all went straight to Labour. Also UKIP clearly picked up the big decline in others and the small fall for the Tories – just look at the the numbers!”

In reality voters don’t act that straightforward and are motivated by all sorts of reasons. The numbers can equally be read as Labour picking up all the Tory defectors and some of the fall in the Others’ share with UKIP picking up the rest of them and former Lib Dems.

Equally, just because the Labour vote has risen in 2015 doesn’t mean they retained everyone who voted Labour in 2010. A big number might have voted Tory in 2015 but they are made up for by defectors from the Lib Dems and the Tories losing more to UKIP.

Voters are moving parties all over the shop all the time, they are constantly churning around.

(see also – ‘Swing’, ‘Uniform National Swing’)


Essex Man – A reference to a particular demographic of lower middle class and working class voters that supporting the Conservatives between 1979 and 1997.

The Essex Man typically referred to individuals whose parents had moved from inner-city London after the war to Essex and with the decline of manufacturing were now generally employed in middle management or self-employed as tradesmen. Less wedded to trade unions than their parents, they were attracted to Conservative policies such as lower taxation and Right to Buy.

New Labour would base their entire strategy to reenter Government around this sort of voter. The term is similar to the concept of ‘Middle England’.


Herding – Refers to the idea that rather than coming up with genuinely independent results, pollsters tend to copy the results of other pollsters.

The theory is that if your polls are wrong and everybody else is right your polls are going to be seen as useless. Your radically different poll could be correct, but could also be completely wrong. It’s much safer to be part of an industry-wide error than an individual one.

Consequently, there is some belief that pollsters weight their polls according to what other polls are saying.

(See also – ‘Outlier’, ‘Weighting’)


Marginal – A marginal refers to a constituency where the swing required for it to change hands is relatively small. Marginals come in all sorts of shapes and sizes and is where elections are largely fought and won.

Areas with lots of marginals include Labour-Conservative marginals in the Midlands, North West, and Northwest London and Lib-Dem-Conservative marginals in the West Country and Southwest London.

There is no strict criteria for what makes a seat a marginal, but it is generally where the swing required to change hands is five percent or less.

Marginals can be incredibly tight, for example the Conservatives hold the seat of Gower with a majority of just 27 votes over Labour.

You can get marginals that can be three-way, so to speak. For example, the 2015 result in Thurrock saw the Conservatives, Labour and UKIP within two percent of each other. Northern Ireland often sees four-way marginals, with the SDLP winning Belfast South in 2015 with only 24.5% of the vote. However, since the decline of the Liberal Democrats, these sorts of seats are relatively rare in Great Britain.

(see also – ‘Basildon’, ‘Safe seats’, ‘Swing’) P

Portillo Moment – A memorable moment on election night that is generally a source of delight for one party’s supporters.

Before he got into trains Michael Portillo was a leading  figure of Conservative right of the 1990s and Defence Secretary at the time of the 1997 General Election. An unpopular figure, his loss of his safe seat of Enfield Southgate to the young Stephen Twigg was hailed by Labour supporters. Portillo has since himself remarked, “My name is now synonymous with eating a bucketload of shit in public”.

The Tories returned the favour in 2015 as they revelled in the delight in beating Labour Shadow Chancellor Ed Balls in his Morley and Outwood constituency. 

(see also – ‘Basildon’, ‘Safe seat’),


Safe seats – Seats which require very large swings to change hands. Safe seats are generally head by important party figures or those that have ambitions to be. For example, Theresa May enjoys a rather large majority of 29,059 votes in Maidenhead, this means key party figures are protected.

Safe seats see little campaigning during elections and are the most sought after by candidates. Conservative safe seats tend to be in wealthier rural areas in places like the Home Counties and North Yorkshire. Labour safe seats tend to be in poorer urban areas or in rural areas with a strong industrial past like the North East and South Wales. The Lib Dems don’t really have safe seats anymore, and conversely pretty much all the SNP’s seats can be regarded as safe. 

There is no strict criteria for a safe seat, but it is generally one where the swing required for it to change hands is at least 10%.

Safe seats can’t be taken for granted though. They can often be swept away in landslides as Michael Portillo learnt in 1997 and indeed the entire Scottish Labour Party did in 2015.

(see also – ‘Portillo Moment’, ‘Marginal’)

Sampling – It is impossible for polls to survey 65m people for cost and time purposes. Instead, they survey between 1000 to 2000 individuals.

But these can’t be any old 1000 people. They have to be a representative sample of the population. There are two ways to gather samples.

One is to do son randomly. This involves randomly dialling people until the number of required respondents is met.

The other is through quotas. Pollsters will know, for example, 52% of the population are women so they will go out and try and find 520 women. Nowadays this is mainly done through internet panels of volunteers that sign up for polls.

(See also – ‘Weighting’)

Shy Tory – The 1992 Election saw a big error in the polls that significantly underestimated the Tory vote.

Part of the problem were 1987 Conservative voters answering voting intention questions with ‘Don’t Know’. In reality they were highly likely to vote Conservative.

Pollsters have made adjustments to take account of this. Some reallocate ‘Don’t Knows’ according to past voting intention, on the basis individuals are more likely than not to stick with their previous choice. Others exclude ‘Don’t Knows’. Some try to squeeze a voting intention out of them.

Snap election – An election that is called prior to the completion of the five year parliamentary term.

Prior to the Fixed Term Parliament Act 2011, Governments could call elections whenever they liked. They would often choose to do so when political conditions seemed favourable either to protect or increase their majority. Technically 2005, 2001, 1987 and 1983 were snap elections.

However, a snap election is more accurately seen as one that is unexpectedly called by the Government very quickly into their term. Examples include October 1974 (eight months after the previous election), 1966 (two years after the previous election) and 1951 (one year after the previous election).

Swing – A very rough way of showing the change in one party’s support against the change in another. It is perhaps the quintessence of election night in the UK, featuring in almost every BBC broadcast since 1955.

It can be used in the context of illustrating changes on a national scale or, more commonly on a constituency basis.

A swing is calculated by adding the change in one party’s vote share to the share in another’s and dividing it by two.

E.g. 2015 result in Copeland 2017 result in Copeland

LAB   42%                                CON     44% (+8%)

CON  36%                                 LAB    38% (-4%)

UKIP  16%                                LD      7%   (+3%)

LD      4%                                  UKIP  7%    (-9%)

GRE   3%                                   GRE  2%   (-1%)

Swing will be used to reflect the changes in the position of the top two parties, in this case Labour and the Conservatives. Next add the two changes together (ignore the minus) so 4+8=12. Divide this by this by two, 12/2=6.. The swing in Copeland was then 6% from Labour to the Conservatives. In theory this means that six in every 100 Labour voters in Copeland switched to the Conservatives.

Swing was created in an era of two party politics so is less useful in a more fractured system with voters switching between lots of different parties. Indeed, the advent of the Alliance saw the BBC stop using its Swingometer for its 1983 and 1987 election coverage. It does however provide a useful rough guide.

(See also – ‘Churn’, Uniform National Swing)

Uniform National Swing (UNS) – Once a swing for a constituency has been calculated after a by-election, or indeed during Election night, pundits will try to see what this means for nationwide results as a whole.

So take the example of Copeland, with its swing of 6% from Labour to Conservatives in the 2017 by-election. Lets assume this is an accurate reflection of nationwide feeling and in an election nationwide this swing in uniform across the country. This means in every constituency six in every 100 Labour voters switch to the Conservatives. This gives an estimation of how many seats switch between Labour and the Conservatives.

UNS will never be 100% accurate. Constituencies all have their own unique politics and demographics meaning voters will act in different ways up and down the country giving different swings from seat to seat. We also live in a multi-party democracy that is increasingly regionalised. For example, a Labour to Conservative swing will tell us very little about what will happen in SNP-dominated Scotland.

(See also – ‘Churn, ‘Swing’)


Weighting – No matter how hard they try, it is impossible for pollsters to get a truly representative sample. Pollsters have to weight various groups within their sample to ensure they are representative.

For example, a sample for 1000 survey has 540 women, but we know women are 52% of the population. To weight this correctly to give a 52% female proportion, every individual female respondent is given a weight of 54/52 = 1.04. Literally they are multiplied by a factor of 1.07 and count as 1.07 of a person when the totals are tallied up.

A more controversial part is political weighting. You may have the right gender and regional balance, but your sample could be unrepresentative of wider political opinion. Pollsters weight for this in the same way with demographics, but it is difficult to assess what a representative level of each political party support is.

Working Majority – As there are 650 seats in Parliament, technically parties need 326 seats to form a majority.

However, in reality the figure is lower. One MP, currently John Bercow, sits as the Speaker who does not vote in Parliament except in the extremely rare occasions of a tie-break. This reduces the figure to 649, and thus a ‘winning line’ for a working majority of 325 seats.

The Irish Republican party, Sinn Féin, does not take its seats in Parliament and is does not believe it is a legitimate body. This means its four seats are taken out of the equation meaning a true winning line of 323 seats.

This explains the differing explanations for the current Conservative majority. The Tories won 330 seats at the election, giving them an overall majority of 10. Taking away the speaking and Sinn Féin, however, gives them a working majority of 14.

(See also, ‘Hung Parliament’, ‘Overall Majority’)