UK

Sky News poll tracker: How are political parties performing as an election looms?

Sky News poll tracker: How are political parties performing as an election looms?

The live poll tracker from Sky News collates the results of opinion surveys carried out by all the main polling organisations – and allows you to see how the political parties are performing.

By charting changing voting intentions from January 2020 to now, the tracker allows you to monitor the evolving picture as we head towards the next general election.

Below you can learn more about the methodology, and how to read the data.

The tool you need as the election looms

Sam Coates

Deputy political editor

@SamCoatesSky

Bookmark this page, remember this tool. Sky News has launched its own, authoritative version of one of the most important indicators available ahead of a general election next year.

Almost every day between now and the election, there will be new opinion polls by a clutch of different pollsters – each using different methodologies and all asking who voters will support on polling day.

Which pollster will be closest, which method is the right one, who should you look at? Those questions will always be unanswerable until the morning after election day, with the past only a broad guide to the future.

There is a tendency for political professionals to seize on every one of the polls, magnify every percentage point of movement, and draw dramatic headline conclusions. No doubt I will at times be guilty of this, but it will also put you at risk of over interpreting a single outlier poll.

Every poll has a margin of error of two or three percentage points either side. This isn’t just ignorable small print, it’s a big challenge for all of us – and a warning for all of us not to impatiently rewrite political narratives based on a single number change.

So the best way to use opinion polling reliably requires patience – and a lot more data. That is where this tool comes in.

How does one pollster, with its (usually) consistent methodology, move over weeks and months? Is there a discernible pattern from several different pollsters over a matter of days? Those of us with our noses pressed firmly up against the glass don’t want to wait for this.

This sort of analysis is only available through a “poll of polls”, which takes data from every single pollster that is asking voting intention questions and signed up to the industry standards body, the British Polling Council.

It is drawn up by Sky election analyst Will Jennings and Sky data and elections editor Isla Glaister – and supported by a team of Sky data scientists and designers. It’s an important piece of work for us, and a lot of thought has gone into it.

The poll of polls seeks to give an answer to the most important question of all – the direction of travel of public opinion over time. Are the closing months of this parliament, the declining state of the economy and the emergence of Labour’s policy platform making any difference? Keep coming back to this page.

There are limits. Crude attempts to turn the polling averages for the main parties into a number of seats for each party will always be just that: rough and ready and probably ultimately unhelpful (not that people will stop trying). This is a GB poll so the level of support for the SNP necessarily reflects how they fare comparatively across Great Britain, not just in Scotland.

Likewise, there is nothing here about Northern Ireland. Liberal Democrats might say they perform better in target seats where they focus resources, rather than nationally where they rely on air war alone.

Nevertheless, this is the page – and a tool – which will tell you the biggest picture story about the main parties and their comparative level of support as we hurl towards a general election where anything could happen. See you back here soon.

How does the tracker work?

The main line

The main line travelling from left to right shows the average support that each party was recording on a given date. The average is a simple mean of each of the most recent polls from all pollsters recognised by the British Polling Council.

Pollsters have slightly different methodologies in how they interpret raw results from the sample of people they ask. Our average uses a maximum of one poll per pollster, which means it is not skewed by pollsters who happen to publish surveys more regularly than others.

If the most recent poll by a given pollster was more than 28 days ago, we exclude it from the average.

The dots

The dots on the chart represent results from individual polls. If you click on a dot you can see the details of that particular poll for each party, including the name of the pollster who carried it out and the date they finished asking people.

The pollsters

The polls we include are all those by pollsters recognised by the British Polling Council (BPC).

The BPC is an association of polling organisations that publish polls, with a commitment to promoting transparency.

It is concerned only with polls and surveys that set out to measure the opinions of representative samples – such as the views of all adults, or all voters.

Membership is limited to organisations who can show to the satisfaction of the BPC that the sampling methods and weighting procedures used are designed to accurately represent the views of all people within designated target groups.

How are polls carried out?

Most polls these days are carried out online. Pollsters use a panel of people whom they know demographic information about – such as age, gender, education and where they live – so they can pick a sample that best represents the whole UK.

If polls are carried out over the phone, they will ask people this information at the time so that they can factor it into calculations.

Over the course of a few days, they ask these people their political preference and then take into account how many people of different demographics they’ve asked – and adjust the results according to what each pollster thinks is the best way to make the sample most representative of the country as a whole.

In general, pollsters should ask at least 1,000 people to get a reliable result. Statistical theory indicates that you are unlikely to get much more reliable results by asking any more than a couple of thousand people – even in a country of almost 70 million – but too many fewer than 1,000 could make the poll less likely to accurately reflect the views of the population.

More detail from the BPC.

Credits

Chart design and implementation:
Dr Will Jennings, Sky News election analyst
Daniel Dunford, senior data journalist
Jenai Edwards, designer

Production:
Przemyslaw Pluta, lead data engineer


The Data and Forensics team is a multi-skilled unit dedicated to providing transparent journalism from Sky News. We gather, analyse and visualise data to tell data-driven stories. We combine traditional reporting skills with advanced analysis of satellite images, social media and other open source information. Through multimedia storytelling we aim to better explain the world while also showing how our journalism is done.

Why data journalism matters to Sky News