MigrationWatch - bunch of charlatans

The Mail story of a couple of days ago, ''Migrants bring only 4p a week in financial benefit', says report' has been bugging me.

I'm not an economist or a statistician. If anyone reading this knows better than I do and sees any mistakes or can clarify what I'm about to go on to say, please say so in the comments and I'll make necessary amendments.

I'm a little confused about why MigrationWatch is concerned, or why this should be considered a scandal at all. I'm also a bit suspicious about how MigrationWatch came up with its '4p a week' figure.

To start with - why is it a problem that immigrants are responsible for the same amount of production as everyone else plus an extra £126m per year on top? I just don't get it. Surely, the anti-immigration lobby argue that immigrants are a burden, which this doesn't prove.

Carrying on from that though - I don't like the way MigrationWatch translates the £126m per year to 4p a week per head. Their report, 'The impact of immigration on GDP per head' shows the working. It also deals with other reports, and there are problems with the way it addresses the figures in those - like counting total immigration as raising population by the same amount rather than using net immigration (which gives double the net immigration figure in one instance), or counting UK-born children of immigrants as immigrants, or burying some figures in the notes because they're higher than the ones used - but it doesn't do that with the Government figures.

What it does is take a figure it gets from a Parliamentary debate made by Joan Ryan, which says, "migration has increased output by at least £4 billion," - and expresses that as a percentage of total GDP - which is 0.32%. It then takes the figure for net immigration from 2005, and expresses that as a percentage of the total population, which is 0.31%. The 0.01% is deemed to be the extra production. They then work out GDP per head of the 60m population, which is around £21,000 per head, and work out what 0.01% of that would be and divide it by the number of weeks per year to get the 4p, which is the smallest number possible.

I have a couple of problems with this. Firstly, I'm not sure what showing the overall extra contribution on the entire population actually proves. Why would anyone expect 185,000 people to produce so much more than the rest of the population that dividing their extra contribution by 60 million would produce a massively high number? Has anybody ever argued that it does, or should? In short, this is all sleight of hand which is addressing a bit of a strawman.

Surely, if you wanted to show how much more per head immigrants contribute than the rest of the population, you'd take the extra benefit and divide it by the number of immigrants.

Now, 0.01% of £1250bn is £125m. Dividing that up between the number of immigrants (185,000) gives about £676 per head per year that immigrants produce compared to the rest of the population. Expressing that per week comes to £12.98. Of course, the net immigration figure is a bit of a problem with dividing in this way, and you'd really want to divide it by the total number of immigrants to get an accurate answer - but it's still going to be more than 4p a week. Something MigrationWatch will never admit is that the reason this figure is so low when you divide it by sixty bloody million and then by fifty again is because the number of immigrants compared to the main population is so low to start with.

The second problem I have is that the MigrationWatch figure might be a mistake in the first place. The £4bn figure is taken from a Parliamentary debate, which is not a proper study that shows its working and fully explains how figures are arrived at. It could well be that the Treasury figure it mentions was calculated by starting with the percentage of immigrants in the population, working out what the same percentage of GDP that would be and rounding up to the nearest billion. It wouldn't be a very good way to arrive at an accurate figure, but the Treasury's argument could be that the population has risen by 0.31% and as a proportion of total GDP, they would represent about £4bn that wouldn't otherwise be there. MW's '4p a week' stuff could be just the result of the Treasury already assuming that the output of migrants is equal to everyone else and including a rounding error. MW could well be jumping at shadows. I've emailed the Treasury and the Home Office (Joan Ryan's Department) to find out how the figure is calculated or what report it came from. Because I am a plank.

The other figure quoted by the Mail and MigrationWatch is the figure for the amount in remittances to foreign countries made by migrants. There are problems with this, too. It's taken from a different debate in Parliament, this time from November 2004. The juxtaposition of this figure with the contribution made by one set if immigrants from one single year creates the false impression that the immigrants responsible for contributing £4bn in 2005 are also responsible for sending around £4bn to developing countries in 2004. This is, of course, nonsense.

Remittances are sent by far more people than actually arrive in the country in that same year. In putting the level of remittances next to the amount produced by one set of arrivals from one single year is misleading. MigrationWatch are playing fast and loose with the definition of 'immigrant' here.

Of course, the Mail are all over this story. It looks great to say that immigrants send £10 million a day home but only contribute 4p a day - but it's all a familiar Daily Mail smoke and mirrors show.

It's weird. Before I started this blog I knew that the Daily Mail was dodgy and not to be trusted, but because I hardly ever read it I had no idea exactly how much it distorts, exaggerates and lies. Reading the Mail a lot gave me the impression that MigrationWatch were sort of dodgy too, but looking closely at one report has made me realise exactly how dodgy they are. MigrationWatch UK? BunchOfCharlatans UK more like.