How to find a fact

"5% of British businesses export to the EU."

"5% of British VAT-registered businesses export to the EU."

"5% of British VAT-registered businesses export to other VAT-registered companies in the EU."

All of these claims are partly right, and none of them tell you the whole story about British trade with the EU.

During the EU referendum, Vote Leave claimed that 6% of British businesses export to the EU. It said this was based on an update of analysis by Business for Britain, which estimated that up to 5% of UK companies exported to the EU in the period 2006-2012.

When we wanted to check Vote Leave’s claim we did what we always do: we went to the primary source.

Business for Britain had done their analysis based on Freedom of Information requests sent to HMRC. We don’t know what it requested, but the response from HMRC (which HMRC sent to us) gave an estimate of the total number of businesses that had submitted ‘EC sales’ lists. An EC sales list is a form that VAT-registered businesses exporting goods or services to VAT-registered customers elsewhere in the EU have to submit to HMRC.

Here we found our first hurdle: the figures excluded UK firms that were not VAT registered. Only companies that have a taxable turnover of more than £85,000 per year have to register for VAT.

The second hurdle? The EC sales list data also excludes trade with businesses or customers, elsewhere in the EU, that are not VAT registered.

So potentially there are a lot of businesses that aren’t included in the data behind the 5% claim. We asked the HMRC press office if, as Business for Britain argued, the majority of exporters would be VAT-registered anyway, meaning that the conclusion was still sound. HMRC told us it didn’t know.

It’s worth saying at this point that this isn’t the only set of data on the number of businesses exporting, but none of the data we have provides us with a complete answer.

We prefer not to write a factcheck that says “we don’t know”. We want, wherever possible, to give you the information you need to make up your own mind. So we needed more answers, and more time. Factchecking is often about a trade-off between speed and helpful information. We’d always rather take longer to publish something helpful, than rush to publication.

We asked the HMRC Head of Profession for Statistics, who didn’t know. Then we tried the Head of Profession at the Department for Business, Innovation and Skills (BIS), who also didn’t know. She suggested we contacted a BIS statistician, who gave us the name of someone in the Office of National Statistics (ONS) who they said was looking into using the EC sales list data.

So we called the ONS. Who referred us back to another person at HMRC. This time, the statistician we spoke to did have an answer for us. They suggested that the exclusion of non-VAT registered businesses could mean that the data misses quite a lot of trade, such as individuals buying or selling things like apps and games online. But no-one knows what the exact or even approximate number of businesses this might be.

Do only 5% of British businesses export to the EU? It’s probably more than that —it could be up to 11% if you’re willing to make enough assumptions — but we don’t know for sure (sorry).

Fail to prepare, prepare to fail

A few weeks later, about a month before the EU referendum vote, BIS produced this analysis, which estimated that 8% of all UK small and medium sized businesses (SMEs) export to the EU, and a further 15% are in the supply chains of other businesses that export to the EU. This was the data behind Stronger In’s claim that 1.2 million SMEs rely on trade with the EU.

There’s a question here about why it took until a month before the vote for any public information provider to put together an estimate for the overall number of businesses exporting to the EU. And when an estimate was produced, why it only focused on SMEs.

We had to speak to three different government departments to get an idea of how accurate the figures being used by the campaigns were, and none of them had the whole picture. And this wasn’t exactly a niche request—it wouldn’t have been hard to anticipate years in advance that questions about UK trade with the EU might have been a source of contention during the EU referendum.

This was just one of many similar experiences we’ve had over the years. And we think that everyone deserves better.

That’s why we’ve started the Need to Know project with the House of Commons Library, the UK Statistics Authority, and the Economic and Social Research Council, which is bringing together government statisticians, academic researchers, policy experts, social researchers and journalists to anticipate what questions will need answering over the next few years.

Together we’ll be mapping out what information already exists, what gaps in the data need filling, and the most effective means of communicating existing data. From this, we’ll be putting together a joint action plan for how we can begin to address those gaps. Robust data takes time to gather, analyse, and digest. We want to make sure this work is done properly, before the data has a chance to become political football.