Cooking the Books 2: Statistical errors

There is a silly argument going on at the moment between the government and an organisation called Migrationwatch which favours tougher controls on immigration. The government claims that people born abroad working in the UK have caused a small but positive increase to gross domestic product per capita. Migrationwatch claims the opposite and argues that in future only immigrants whose work contribution raises GDP per head should be allowed in.

GDP per head, i.e., total annual production of goods and services divided by total population, is simply a statistic; it doesnt cause anything but is a measure or reflection of a situation caused by real facts. If GDP per head falls because GDP has fallen or has remained unchanged while population has gone up this might indicate a deterioration in general living standards (though even then most people could be unaffected since a fall in GDP per head does not mean that everybody is worse off any more than a rise means everybody is better off). But GDP per head is not falling but rising. So, what the government and Migrationwatch are arguing about is the hypothetical question of whether it would have risen faster if there had been fewer immigrants.

But how do you measure what a worker contributes to GDP, i.e., to total annual production? Migrationwatch explicitly, and the government implicitly, assume that a worker contributes only the equivalent of their wages. As Migrationwatch argue in a recent research paper:

In the calendar year 2003 the UKs GDP was £1.099 billion. £613 billion of this amount was compensation of employees. So, apportioning this amount of GDP generated by employment earnings amongst the working population of 27.6 million people this gives average earnings per worker of £22,200 a year (“Selection criteria for immigrant workers”, www.migrationwatch.org).

But if workers produced only £613 billion of a total production of £1,099 billion who produced the rest? The same statistics show that the rest is made up of profits (25 percent), mixed income, i.e., the profits and the labour contribution of the self-employed, (6 percent), and the difference between taxes and subsidies.

Since work is the only possible source of new wealth, a more accurate calculation would be to divide £1,099 billion by the working population; which gives a contribution of £39,800 per worker. This would reflect the fact that all productive workers, whether native-born or born abroad, contribute to GDP considerably more than their earnings but what they produce above this goes as profits to their employers.

This rather demolishes Migrationwatchs convoluted calculations to reach the conclusion that a worker must earn about £27,000 a year to make, on average, a positive contribution to GDP per head and that only migrants earning this or more should be allowed in.

Migrationwatchs stigmatising of any worker, native-born ones too, earning less than £27,000 as a burden since they contribute less to GDP than average so dragging GDP per head down is just plain ridiculous. As GDP per head is an average there will always be some above and some below it. Migrationwatchs proposal to raise the average by eliminating some of those below it would achieve this but it would reduce GDP (since even Migrationwatch admits that any immigrant who works contributes something to GDP). A bit like cutting off your nose to spite your face. But then, Migrationwatch is only deploying apparently sophisticated statistical arguments to back up its already-decided policy of keep the riff-raff out.

Leave a Reply