Saturday, 8 November 2014

Correlation does not equal causation

One of the most cliched statistical truths has been particularly pertinent this week, argues Andrew Fisher

Migration and low pay

It's unsurprising that in the current xenophobic climate that migrants are getting blamed for the low pay crisis in the UK economy as recent reports show over five million people are paid less than the living wage.

This is the inevitable UKIP and xenophobia-lite response to the University College London report this week that showed, very clearly, that migrants from inside and outside the EU have been net contributors to the UK economy in the last decade or so.

So some of the xenophobes took to the media (social and corporate) to argue that migrants must drive wages down for others even if they are net contributors themselves. While some reports show a small decrease in wages at the lower paid end of the labour market correlated to migration, the research is far from conclusive. As one robust report from 2008 states, "The above correlations offer little support to standard theory predictions that migration inflows exert downwards pressure on wages and employment".

In fact wages have been falling as a share of GDP since the late 1970s and the deregulation of the labour market and increasing legal restrictions on trade unions - which have combined to reduce the coverage of collective bargaining agreements. There is also the systematic hollowing out of skilled manual labour from the UK economy, which has depressed wages and the laissez-faire policies that have expanded our low skill economy.

Migration is almost certainly a marginal, if at all significant, factor in low wages. And low wages won't be resolved by capping or stopping migration, but by stronger employment and trade union rights (i.e. increasing the power of labour vs capital) - and by building a more skilled economy, which would require some level of industrial strategy (AWOL since the 1970s).

Dodgy stats on the benefit cap

The DWP press office twitter account, which has seemingly opted-out  of the political neutrality that the civil service demands, produced more evidence this week when it pumped out more government spin on the 'success' of the benefit cap (see our myths of the benefit cap post for background)

But the UK Statistics Authority (UKSA) has already chastised the DWP and Iain Duncan Smith for mistaking correlation for causation before (see Guardian report from 2013). In an open letter Andrew Dilnot of UKSA told Duncan Smith the benefit cap figures were "not intended to show the additional numbers entering work as a direct result of the contact" and therefore such claims are "unsupported by the official statistics published by the department".

In short, the fact that some people who were affected by the benefit cap are now in work could be due to the natural turnover of unemployed people finding work, and nothing to do with the fact there is a benefit cap.

Aside: For more on dodgy stats and Iain Duncan Smith, see this dossier produced by PCS

Osborne and the EU

And finally, just because our EU bill has halved from £1.7 billion to £850 million - and George Osborne visited Europe - it doesn't mean George Osborne negotiated the halving of our EU bill.

No comments: