Tuesday, 11 February 2014

Jonathan Portes's Stats-'n'-Graphs Games on Immigration

Rather than introduce Jonathan Portes myself, I will save myself the trouble and let Liberty GB's Chairman, Paul Weston, do so instead:

"Portes is currently Director of the National Institute of Economic and Social Research, Britain's longest established independent economic research institute. Prior to that, he was Chief Economist at the Cabinet Office where he advised the Cabinet Secretary and Number 10 Downing Street on economic and financial issues ... . Portes was also associated with the Labour Party which has admitted that flooding Britain with the Third World was part and parcel of a 'social objective' to transform the cultural landscape of Britain and to rub the nose of the right in 'diversity'." – From "Is Tommy Robinson Anti-Semitic?".

Lies, damned lies, statistics ... and ideology

Statistics is a funny science because one of the first things statisticians usually do – when speaking to novices or introducing their subject – is warn you about their discipline. So even though statisticians don't always quote the well-known line from Disraeli themselves, here it is anyway:

There are three kinds of lies: lies, damned lies, and statistics.
One of the reasons why people should be suspicious of stats is that, yes, many people, including Leftists like Portes, have told us to be so. Nonetheless, in the case of Leftists, they are usually only talking about the stats which work against their own ideologies or causes. When they work in their favour, then, of course, 'stats never lie'.

The thing about Jonathan Portes's endless citation of stats and graphs is that another 'expert', or even another twenty experts, could quite easily come along – and they often do – with just as many stats and graphs which work directly against all the positions and causes Portes is trying to advance or support with his own. That's the thing about statistics – they can be made to say just about anything.

The fact is that Portes is clearly using stats and graphs politically and ideologically. What I mean by that is that he starts off with a set of ideological and political positions (on immigration, racism, employment, poverty, etc.), and then applies stats to them. The ideology or politics comes first and then the stats and graphs dress up or justify them. After all, this man is – or was – a supporter and paid employee of the Labour Party and has done much work for it.

This isn't an argument that 'all statistics lie' or even that there's no place for statistics in political debate. It's just an argument for being sceptical about them. As I said earlier, Leftists like Portes often say pretty much the same thing – if only when it comes to the stats which work against their causes or political positions. And Leftists, not just self-described Marxists, often give a similar warning about theory and ideology. That is, they advise us to take note of the 'ideological or theoretical underpinnings' of what people (their enemies) say. Of course the implicit assumption here, again, is that the ideological or theoretical underpinnings of what Leftists say should be either ignored or embraced.

Leftists are right on this. Citations of stats or graphs will often – though not always – come with various political or ideological underpinnings. Jonathan Portes, the Grand Wizard of stats and graphs, is no exception to that general rule.

There is of course another big problem. What if the statistics are simply false? I don't mean misleading or open to interpretation – I mean untrue. Indeed what if the statistician is well aware that they are untrue and is simply using them to further certain political goals? The thing is, if people aren't checking Jonathan Portes's large output of stats and graphs on a regular basis, and I will assume that very few people – if any – are, then he could be getting away with statistical murder. Indeed even if people are checking Portes's goods, there's no guarantee that they'll spot his duplicities or even his outright lies. After all, we're talking about stats here! And what if a Leftist or left-liberal statistician or 'expert' notes a false or duplicitous use of stats – will he automatically inform the public about it even if he otherwise agrees with Portes's politics generally or, more specifically, if he agrees with the political position being advanced by that particular stat or graph? A person would need to spend a lot of time and brain-power on Portes's statistical and graph-laden effluvia in order to make sure he wasn't playing silly buggers. I bet there are very few people – if any – who are actually doing that.

Of course telling a lie verbally is far more obvious and seemingly reprehensible than telling a lie with numbers or graphs. For example, if Portes were to say, "I won the prize for statistical objectivity from the London School of Hard Knocks", then that lie would both be easy to spot and easy to refute. Not so with stats and graphs.

'Most groomers and paedophiles are white, not Muslim'

Hah! A classic soundbite and a classic statistical deceit.

Take a fictional town or city in which only one percent of the population is Muslim. Of course most groomers and paedophiles in that town or city will be white. Now say there are five percent Muslims in that town or city; the same will likely be true in that case too. You can keep this argument up even until the overall percentage of Muslims in the town is as much as 30 percent. Beyond that, though, the situation will indeed change.

If in that first example of 1% Muslims in the overall population, Muslims also made up 1% of the figures for sexual grooming, that would be what you would expect. However, the chances are that Muslim groomers, as a percentage of all groomers, would greatly exceed 1%.

Consider the real cases of places where Muslims make up from 1.7% (Colchester), to 24.7% (Bradford), all the way to 36% (Tower Hamlets) of the overall population; often the ratios are far higher when it comes to sexual offenders (especially in the case of sexual grooming). For example, there are cases in which Muslims make up 5% percent of a town or city yet account for nearly all sexual groomers.

In others words, simply citing the percentage of the overall population of a town or city who are Muslim groomers or paedophiles will be misleading, and the same goes for the entire UK. What has to be stated is the percentage of groomers or paedophiles who are Muslims. That makes a big difference.
For example, according to Al Jazeera, this was the reality in 2013 (29th June):
"[A study] done in 2011 looked at 2,300 potential offenders who had been caught grooming. Of the 940 whose race could be identified, 26 percent were Asian, while 38 percent were white."
Now here we see another problem – that ubiquitous and insulting term (to the Chinese, Indians, Japanese, etc.) – 'Asian'. It is clear that almost all of these 'Asian' groomers were Muslims. In addition we can ask: What about Arabic, Kurdish, Turkish, North African, etc. Muslim groomers – were they classed as 'Asian' too?

Instead of percentages, we can express the reality in terms of numbers. For example, another survey found that 367 groomers were white and 346 were Asian (yes, that term again!). Again, even though the numbers and proportion of Muslims are very high, the areas or towns and cities we are talking about will still need to be addressed too.

Jonathan Portes's Most Recent Stat

The following stat was published on the Internet on 15th August 2013 by Jonathan Portes:

"The increase in Bulgarian/ Romanian workers represents less than 1 in 1000 UK jobs. Hardly 'unsustainable'."

That figure of 1 in 1000 UK jobs, even if true, depends on how evenly spread in the UK these Bulgarian and Romanian workers are. The assumption here – or what Portes wants people to assume – is that these workers are spread throughout the UK. But they're not! (And this applies to nearly all arguments about immigrants.) There are many more Bulgarian/ Romanian workers in some places than in others, which makes the figure of 1 in 1000 highly misleading.

In these areas at least, such an intake will indeed be unsustainable (as Portes sarcastically puts it). Admittedly, I'm not entirely sure what 'unsustainable' means in this context. What I do know is that British workers, in the areas of the UK ignored by Portes, will be finding it very hard to find work. Sometimes they will find it almost impossible. In addition, large immigrant populations (not only the Bulgarians and Rumanians) will be causing various problems for the community, such as crime, a shortage of housing, etc. I'm not saying that all immigrants are criminals, but the other problems arise by virtue of the fact that they are a large group of new people in areas so far unaccustomed to such large intakes of immigrants. (Perhaps Portes has a cheap immigrant nanny and/ or cleaner, as many bourgeois Labour and Leftist bigwigs have, especially in London.)

Let's think about the propagandistic power of the numbers themselves. The phrase '1 in 1000' sounds – sort of – impressive. The number 1 sounds small and the number 1000 sounds large. Nonetheless, one active and persistent paedophile in 1000 is a very significant number. It's 100 active paedophiles per 100,000 of the population, or around 150 active and persistent paedophiles in an average-sized town.

Now what about one active terrorist for every thousand people? Do you see what I mean? Stats alone tell don't tell you the whole story. Every citation of numbers and stats needs to be considered in context, along with non-statistical analyses taking account of other factors. In fact it's often the case that citations of stats require other stats which make sense of them (as was seen in the Muslim-grooming case mentioned earlier).
It's clear that Portes was only talking about very recent Bulgarian and Rumanian immigrants. What about all the other immigrants in specific areas of high concentration? What if, besides the Bulgarians and Rumanians, there are also large numbers of immigrants from many other countries? Again, that 1 in 1000 stat makes even less sense in these contexts and is a clear piece of dissimulation or outright propaganda.

Finally, his claim that "The increase in Bulgarian/ Romanian workers represents less than 1 in 1000 UK jobs" makes no mention of timescale. If Portes is referring to the rise of "almost 26% on the 112,000 in the previous three months", as reported in a Huffington Post article, then that is a large rise in a short period. And as with Portes's "1 in 1000" stat, this 26 percent increase will not be evenly spread throughout the UK. Portes is playing with percentages, playing with timescales and playing with numbers, all strategically chosen to make it seem that there isn't a problem at all.

The Office for National Statistics also said that "141,000 Bulgarians and Romanians were employed during the period from April to June". Does "were employed" mean they had or that they found jobs? In any case, a rise of 29,000 Bulgarian and Rumanian workers in just three months is very large. If that trend continued, it would result in an increase of almost 120,000 Bulgarian and Romanian workers in one single year. Again, will they be evenly spread throughout the UK? Not a chance.

Not only all that. We need to make distinctions between very recent immigrants (those who have come within the last year or so), fairly recent immigrants (those who have come within, say, the last one to five years), and recent immigrants (between, say, five and fifteen years). These distinctions are important because, for example, very recent immigrants can create problems which are serious and pressing and which aren't associated to such a degree with immigrants who've been in the UK longer.

There are also the obvious distinctions to make between the various types of immigrant. For example: the many immigrants who end up claiming benefits; those with genuine skills and qualifications; those who are likely to end up engaging in crime; those who'll find it hard to assimilate or who refuse to do so; and, perhaps most problematical of all, those who are Islamists or militant/ fundamentalist Muslims. This latter group will include individuals who will end up being unpredictable (in terms of 'community cohesion') even if they find work and are skilled or qualified. Indeed even if they become professors at our universities or journalists who write for our newspapers.



Of course statistics, numbers and graphs look sexy. Or at least they look academic and serious. (All along I've been assuming that Jonathan Portes is a statistician – but maybe he isn't. He may not even be any good at arithmetic!) So what can happen is that stats, numbers and graphs are often – though not always – used to impress us or, in Portes's case, to bolster a pre-existing political or ideological case that was otherwise – and up until their use – completely free of stats and numbers. This is a kind of fake objectivity or even a mild case of scientism. You find this phenomenon all over the place: from the legions of commonplace philosophy students (as well as some professors) who tart up their work with logical symbols, Ps and Qs and various visual 'schemata' (when they're simply not needed), to other academics who place at least ten footnotes at the bottom of each page and a thousand references at the end of each paper. These things can of course be necessary and useful in academic literature (or even outside academia); however, they are too often used to bowl us over or, in Jonathan Portes's case, to hoodwink us.

No comments:

Post a Comment