MigrationWatch - last post

Okay, last post about MigrationWatch for a while, I promise. Skip this post if you don't fancy going through a bunch of numbers and 'no it isn't' style contrariness. See if I care.

It's important before I go on that I think MW's whole premise is based on a strawman argument. As far as I've been able to tell, people generally talk about the growth of GDP contributed by migrants to show how much would not have been contributed if they were not here, not how much more they contribute than anyone else. The argument is usually that immigrants tend to either do low paid jobs nobody else wants or fill skills gaps, so the inclusion of the addition to GDP is normally shown for at least two reasons:

1. To show how much would have been missed had these positions not been filled.

2. To show that migrants are not responsible for a drop in GDP.

Given that a main pillar of the pro-immigration argument is that migrants disproportionally do low paid work, why would anyone interpret that argument as saying that migrants produce far more per head than anyone else? The very argument is based on the assumption that a great number of migrants will produce less on average than everyone else. The benefit is that shit jobs that nobody wants get done, and vacancies that we can't fill get filled. There is a secondary argument to this one, which is that migrants often do jobs that allow members of the host population to make a greater contribution than they otherwise could. The extreme example I've seen is to take an au pair who works for two doctors, allowing them both to work rather than have one stay at home. MW seem to be attacking an argument that nobody has made.

I could probably leave things there, and not dignify the report with any more attention, but it's worth looking at because it's pretty damn shonky - and because enough people just dismiss MigrationWatch without explaining why, which is why I've had to spend time going through this. The report is split into sections and the paragraphs are (mis)numbered, so I'll go through section by section. You lucky people.

It doesn't begin well. There are problems with the summary. With the very fist sentence, in fact:
Most immigrants add to production but, in the long term, the host population will only benefit if there is a resultant increase in GDP per head.
No. As I said above, the host population will benefit if things get done that wouldn't otherwise get done. People will benefit by the recruitment of new nurses regardless of how little they're paid or what their individual contribution to GDP per head is. This is particularly true if the level of migration is small in comparison to the population. If three nurses are recruited, that represents sod all GDP even before it's divided by 60 million, but there's definitely a benefit. Plus, migration can indirectly increase GDP per head as well as directly from their own production.

Indeed, the main benefit accrues to the immigrants who are able to send home about £10 million a day.
Notice how MW characterise a benefit to immigrants as a Bad Thing. Every pro-immigration study and argument I've read do not ignore the benefit to immigrants, but actually use it as a positive argument for immigration. It's often referred to as unofficial international aid. The very answer in Hansard that MW get the figure from talks about the remittances to poorer countries as a benefit of migration. But how dare immigrants think they can use their own money in any way they like. The cheek of it! (This indignation, minus the figures, is also in the other release. It says, 'The major benefit goes to immigrants which is why they come,' as if it's a surprise that migrants come here because it benefits them, and as if it's bad that it does).

As I mentioned in my last couple of posts, there is a problem in that the figure they're quoting is of total remittances sent in a year. More people can send remittances than people who have arrived in one single year. Comparing the two is misleading.

There's some attempt to address a couple of studies that suggest immigrants contribute more in taxes than they receive. It says:
The government's first effort which showed a net benefit of £2.5bn was based on a year in which the budget was in surplus so everybody was making a positive fiscal contribution.
So these figures are dismissed because the budget was in surplus. MW make it sound like this is unusual, but the budget has been in surplus every year for the past few years so any recent study that produces results MW don't like can be dismissed out of hand. Which is handy. (Thanks very much, I'm here all week). It also says:
This study was superseded by an IPPR study which also showed a positive contribution. However the result was distorted by the inclusion of all children of mixed households (one parent an immigrant, the other not) in the host community. Correcting for this by splitting the cost of these children 50/50 produced a small negative outcome.
Why add the children of mixed households to the immigrant community? They're not immigrants, they were born here. Adding them is not 'correcting' at all. It's doing the opposite. What MW are saying is that if you add people who shouldn't be added to the immigrant community, the results are different. No shit, Sherlock.

This lead in is in itself quite misleading. The summary mentions effect on GDP per head, as do most of the other sections, but this intro gives the impression that MW will be talking about tax contributions, which they're not.

The National Institute report
From paragraph 3:
It concluded that immigrants who have arrived since 1998 have raised GDP by 3.1%. According to the Labour Force Survey figures in the report, of the 58.987m population, 2.249m have come to the UK in the 8 years since 1997. The crude addition to the population (excluding UK-born children) is therefore 3.8%. So the benefit is negative in terms of GDP per head.
No, the addition to the population is not 3.8%. MW have ignored the number of people who have left the country or died in this period. The addition to the population is however many of the 2.249m are above the number of those who have left the country or died in the same period. That works out at 1.349m (follow the links and do the maths). Net migration for this period works out at about 2.3% of the population. So that would be a positive contribution of 0.8% then.

There's a great bit of information buried away in the notes for this section:
However, the 3.1% is just the earnings contribution to GDP. The government normally use this as a proxy for their contribution to GDP as a whole. On this basis the contribution to GDP would be 4.5%.
So, using Government measuring techniques would actually produce a positive even if we use MW's false figure for the rise in population, but it buries this in the notes. (Remember this, it's important in paragraph 4). Also, using the correct number for the rise in population shows that migrant contribution using this measure is actually just under double the population increase. That'd be slightly better than a negative, I'd say. It follows this up with:
On the other hand, UK-born children should have been included with the immigrant community. When these two factors are taken into account the overall impact on GDP is likely to be neutral.
So their actual argument here is that migrant contribution is neutral, not negative as it says in the actual body of the report. I also like the detailed workings MW include to explain how they arrived at this conclusion - oh wait, they didn't include any workings! Maybe that's because to actually say, "by reaching around and pulling the figure out of our arse, we conclude..." might be admitting too much. Again, why should UK-born children of immigrants count as immigrants anyway? They're not bloody immigrants. What MW is basically doing here is saying that if you pretend some people who are not immigrants actually are, and then you decide that there are just enough of these people to cancel out any positive contribution immigrants might make, then they cancel out the contribution if you include them. Jesus! These people are statistical wizards!

From paragraph 4:
[The report makes claims about the contribution made in 2004-2005]. This claim was based on 815,000 migrants arriving in 2004 and 2005. They therefore added 1.4% to the population and 0.9% to earnings GDP. [Emphasis mine]
No they didn't. Net immigration in 2004 and 2005 combined is actually 408,000, meaning that about 0.7% was added to the population. That would create a positive contribution.

If earnings are taken as a proxy for overall GDP, as is the government’s practice, the addition to GDP would be 1.3% so the impact on GDP per head would be slightly negative.
Remember, this measurement was buried in the notes in the last paragragraph when it would have shown a positive contribution, but it's included in the main body of the report here, where it shows a negative. Must just be a coincidence, eh? But again, this measurement shows a gain to GDP of almost double the actual gain to the population.

The Government calculation
I dealt with this one in my last couple of posts. It's really difficult to address properly without knowing where the £4bn comes from. The truth is that if you make the same assumption about the figures as MW (the second time around, rather than the first) migrants contribute £66.46 GDP to overall GDP per head, which raises GDP per head by the amount MW claims.

The important thing to note here is that MW actually use net immigration in this section, which proves that they do actually know that only net immigration shows how much the population has increased because of migrants.

So why would they not use this measure in the rest of the report? It couldn't be because that would consistently show positive contributions, could it? And why use net immigration here? Because these are the figures MW are really interested in getting into the papers. That's why they're the ones MW have tried to release before, and why they're the ones MW use their simplistic 'Mars bar' analogy with. The other sections are just bluff to make it look as though other studies support MW's findings in this chapter.

The Item Club report
From paragraph 7:
It remarked on their wide dispersal around the country and across a range of industries. The largest number are in administration, business and management which "certainly contradicts, the impression that workers come to the UK to take up low-skilled occupations.” (In fact the Workers Registration Scheme shows that 80% of employed East Europeans are earning less than £6 and 95% are earning less than £8 an hour).
The implication of the bit in brackets is that the report says what it does to suggest something about the wages of Eastern European immigrants. It doesn't. The argument about the wide variety of migrants' skills is usually made to suggest that migrants won't compete for all the same jobs. See, that's why the report points out the dispersal around the country. And the claim about what employment section the largest number are from is actually hinted at by the Accession Monitoring Reoprt (of the Worker Registration Scheme) that MW themselves quote for figures, which includes a table of the top 10 employment sectors with administration, business and management at the top. The hint's very subtle, so it's easy to miss. [Ahem]. The way that MW have worded this paragraph makes it look as though the Item Club report has made a massive blunder in claiming the largest number are in administration, business and management when this is in fact the case.

I'd have to see the Item Club report to comment properly on paragraph 8. Do they really assume that Eastern Europeans registered under the Worker Registration Scheme earn the same average as everyone else, or are they including the self-employed or others not covered by the scheme as well, or something else? Like, for instance, migrants working in more than one job. The Accession Monitoring report MW use to get their 'just over half the UK average' figure for migrant wages says:
The total since May 2004 includes [...] 4,000 multiple registrations (from those
working for more than one employer simultaneously), and 34,000 multiple re-registrations (from those registering for subsequent, additional jobs, or those who have left their employer and are re-registering for more than one job).
As well as assuming that migrants work no overtime, and that those who have been in the UK for over a year and no longer have to register have never been promoted, or moved from a temporary or casual position into a permanent post with better pay.

From paragraph 9:

An extra 300,000 people adds 0.5% to the population.
Does it? Does it really? Even withouth the net immigration rule, MW are using their other favourite trick of ignoring the number of those who return. There are no figures to show how many will stay, but we have an idea from the workers who register their intentions. 59% registered their intended length of stay, and almost half stated less than a year (45% said less than three months), so around 28% say they intend to stay for less than a year. Allowing for the net effect of immigration and the number of those who leave the UK would reduce the total significantly. Also:
This excludes accompanying dependants who add at least 17% to the migrant population.
Why drop the Worker Registration Scheme stats now? MW fucking loved them before. The Accession Monitoring Report of spring 2006 shows that the number of dependants so far amounts to around 8% (374,555 total migrants with 30,080 dependants), and there's a note that says:
Note: it is likely that there is some ‘double counting’ of dependants, in the sense that some of those recorded as dependants (particularly older children and spouses) may also have registered in their own right to work in the UK.
So the number is probably lower than 8% because some people listed as dependants are actually not dependants.

Instead of this figure, MW use one they get from one of their own shonky reports. Less than 8% becomes 17%. Almost double again. Nice that, eh? There's no need to trash the methodology used in that report, as we have some nice direct numbers right here we can use. Sweet. But I will anyway, because it exposes a certain amount of dishonesty here. Hurrah! Their other report 'Economic Contribution of A8 Migrants' says:
We assume that dependants will mainly accompany migrant workers who intend to stay in the UK for some time and that any dependants who have left are balanced by additional dependants who joined migrant workers after they became established in the UK. For the purpose of this analysis we have assumed that half of the 427,000 migrants have remained in the UK and that all their dependants remain in the UK. [The number of migrants they use here is from a later report than the one the Item Club must have used - in this later report, the number of dependants is around 9%].
That's quite a dodgy way to work out the percentage of dependants, isn't it? We don't know how many will return, but we'll assume it's a high number so that when we arbitrarily decide that the number of dependants will reman the same even after we halve the number of adults, we'll get a high percentage. Could you think of anywhere else that MW would assume that the number of workers who have returned is as many as half? Remember that when you see their stats of 600,000 total migration from Eastern Europe quoted. MW themselves actually think the real number is half that. Plus, they completely ignore the note about duplication.
UK-born dependants and dependants who later join migrants already established in the UK are additional to this.
Not sure about the immigration staus of UK-born dependants in this case. But the report carries on with some more numbers with meticulously recorded working:
The total addition to the population is therefore likely to be in the region of 0.6% to 0.7% of the population and the impact of A8 migration on the UK’s GDP per head will be significantly negative.
No it isn't, because MW have exaggerated the initial rise in population as well as the proportion of dependants.

Now, there's a problem in not being able to see the report (Item Club reports only go as far back as June on the site) becuase the 300,000 may well have already been adjusted for net immigration and the number who might have returned. If this is true though, MW's assumption for the percentage of dependants would need adjusting - and they still ignore the note about the number of dependants actually being lower. They just seem to have assumed the level is double the figure actually reported - which is depressingly familiar.

International Experience and Conclusion
Conclusion shmonclusion.

The striking thing about this report is the ability to shift arguments when the numbers don't fit MW's assumption that immigration is a Bad, Bad Thing. Don't like the offhand reference in a parliamentary debate and prefer a full report? Never mind, we'll just shift from using net immigration to make the addition to the population look larger. That still doesn't work? We'll include people who are actually not immigrants into our calculations. Don't know the actual number of these people? We'll just assume it's enough to support our conclusion. None of these things work? We'll just say the whole study is void because of the year it was produced. That'll take care of it.


septicisle said...

Don't despair, these are brilliant posts. Thanks for doing them.

Five Chinese Crackers said...

Cheers for that. As a reward, you get another one!