Fact check: Can we take Danny Alexander's "3 million jobs" threat seriously?

In June, Liberal Democrat Chief Secretary to the Treasury Danny Alexander said that, according to Treasury analysis, more than three million British jobs would be at risk if Britain left the European Union. This is what he said:

“Indeed, the latest Treasury analysis shows that 3.3 million British jobs are connected to Britain’s place in Europe. That is the measure of the risk that isolationists would have us take.”

What new analysis could this be? So we asked the Treasury. You can read the full response here, but this is perhaps the key part:

“As set out by the Chief Secretary to the Treasury, the Treasury estimate that 3.3 million jobs in the UK may be related to exports to other European Union countries. This figure is based on the assumption that the share of UK employment associated with UK exports to the EU is equal to the share of output that is exported to the EU, making allowance for the composition of the UK economy. It is not an estimate of the impact of EU membership on employment.”

The Times also reported on the FOI response today. Now, there are at least three things that are problematic with Mr Alexander's claims:

1. First of all, this is not in fact 'new' analysis. As the response to our FOI makes clear, Mr Alexander's remarks were based on the following methodology, which has been used and cited countless times before. In 2003, Ruth Kelly told Parliament that:

"The Treasury estimates that 3 million jobs in the UK are linked, directly and indirectly, to the export of goods and services to the European Union. This figure is based on the assumption that the share of total UK employment associated with UK exports to the EU is equal to the share of total UK value added (GVA) generated by UK exports to the EU. The information necessary to apply the same method to derive comparable estimates for England, Scotland, Wales and Northern Ireland is not available."

In February 2014, Lord Livingston again confirmed this methodology as follows:

"The estimate of 3.5 million jobs linked to trade with the European Union is based on the assumption that the share of UK employment linked to trade with the EU is equal to the share of total UK value added (GDP) generated in the production of goods and services exported to the EU." 

"The calculation uses data from UK Input-Output tables to estimate the proportion of UK value-added content generated in exports of goods and services and applies this to the values of UK exports to the EU. This is then divided by total UK GDP and the resultant proportion then applied to the total UK labour force to estimate the proportion of the labour force linked to EU exports on a value-added basis."

In short, the methodology dates back more than ten years and does not seem to have been updated at all.

2. Perhaps most crucially, as the Treasury's response makes clear, this study "is not an estimate of the impact of EU membership on employment." So it is rather misleading, to say the least, of Mr Alexander to suggest that these jobs are at "risk". To say that X number of jobs are linked (directly or indirectly) to exports to the EU is clearly not the same thing as suggesting that they are dependent on EU membership. To be fair to Mr Alexander the Treasury's position on this is somewhat confused itself - it claims that this is not an estimate of the impact of EU membership on employment, however, that is exactly what the simplistic calculation and, importantly, its flawed counterfactual seem to suggest (see below). 

3. Thirdly,  is immediately apparent, it is a very simple approximation, not detailed analysis which one might have expected from HMT on such an important question. The methodology used is very simplistic, for a number of reasons:

• A flawed counter-factual: We have always stressed the importance of the counter-factual when assessing the future of the EU/UK relationship. The counter-factual here is essentially that these jobs would not exist without the EU.  This is odd for at least two reasons: first, it effectively assumes an end to all exports to Europe should the UK withdraw from the EU. We can argue about the level of market access a post-Brexit UK may be granted (we've done a lot of work on this) but one thing is clear: there will be exports from the UK to the EU under any scenario. Secondly, by definition the analysis assumes that all value added by jobs related to EU exports would not exist without the EU membership. A simple common sense check suggests that, actually many of these jobs may still produce some value even if the goods did not find their way to the EU and that the resources could be alternatively employed.
• Assumes productivity is the same across the UK economy: While it is claimed that the calculation "takes account of the composition of the UK economy" it is not clear exactly how this is done. On the surface the calculation also seems to implicitly assume that labour productivity (broadly output per worker or per hour) is the same across the entire economy (by saying the basic proportion of output corresponds to the same proportion of employment). Fundamentally we know this is not true - on the most basic level, we know that skilled and unskilled jobs will have different productivity levels. A quick glance at the most recent ONS labour productivity statistics confirms this and highlights that over the decade since this methodology was created, different sectors' and regions' productivity rates have grown in different ways. Unless this is accounted for in a detailed way in the methodology it is likely to distort the figure.

The final point above suggests that there is either some continuing lack of transparency with regards to the calculation or it really is just overly simplistic (if not both). This is all quite ironic given that UKIP and Better Off Outers in general are often criticised - and often justifiably so - for not establishing a credible counter-factual whilst relying on heroic assumptions.

Mr Alexander should know better.