Steyn and his numbers
Sully links to a review by Johann Hari of Mark Steyn’s book ‘America Alone: The End Of The World As We Know It’. I haven’t read the book, probably won’t and don’t necessarily agree with Hari. What does interest me, however, is Hari’s consideration of Steyn’s demographic projections for Europe, because I’ve wondered about them before.
Back in early 2005, Steyn wrote an article entitled US can sit back and watch Europe implode. In it, he quotes ‘some projections’ that the EU will be 40% moslem by 2025; bizarrely, he wrote this article after 10 new states acceded to the EU, adding a population of 74 million, nearly all of whom are non-moslem, which would about cancel out the effect of eventual Turkish accession in terms of moslems added to the EU population; Turkey, remember, is far and away the biggest moslem population at all likely to join the EU by accession. He also mentions ‘CIA projections’ that Europe will break up within 15 years. He doesn’t provide any evidence or citations, however, however, so google is our friend.
There was an NIC report released in December 2004, a couple of months before the article was written, where four ‘fictional scenarios’ were considered; one of them, starting on page 83, is ‘A New Caliphate’ in which the moslem population of Europe is projected to go as high as between 22 and 37 million; that is not the same thing, of course, as heading towards 40 percent. Could he have made that mistake? It doesn’t seem reasonable that he could have, but maybe he did, and in company; there is a Scotsman article at about the same time (a little over a month before Steyn’s article) that talks of a CIA report (that clearly seems to refer to the NIC2020 project from which the fictional scenario I mentioned was taken) where it is projected that the moslem population will climb to 22 to 37% instead of the 22-37 million in the report. So it seems that the Scotsman author, at least, did make the mistake of switching percent in for millions. The only collapse hypothesised within 15 years, in the NIC report, though, is that of US-EU relations, not of the EU itself (as correctly reported by the Scotsman article). Probably unconnected, but in 2005 both the Scotsman and the Telegraph (for whom Steyn at that time wrote a column) were owned by the same people, the Barclay brothers.
At the beginning of 2004, the population of the EU was 456 million. 40% of that figure is 182 million. Where are they going to come from in twenty years, assuming anything close to a constant EU population, from a 2005 muslim EU population of 15 million or so? Even if the population drops as the babyboomers do, there’s still going to be, ballpark, over a hundred million new moslems needed to make that figure true.
Did Steyn misunderstand the figures he was quoting? I have no idea; he should really have referenced them so we could check. Did he retract the claims? Maybe; I’ve not seen it, but he seems to be peddling the same crap in his book. But then, it’s more fun to make huge claims than do boring old sourced journalism.
personally I’m not a big fan of either Steyn or Hari. Steyn is a polemist and Hari is just plain annoying.
Comment by dizzy — 3/10/2007 @ 3:05 am
I don’t really know anything about Hari, but I’m certainly not about to spend money on Steyn’s book, particularly given that I don’t trust Steyn’s numbers in general; without his numbers, he has no case for impending doom other than, perhaps, wishful thinking.
Comment by Adam — 3/10/2007 @ 7:20 am
40% by 2025 is ridiculous. But 25% in countries such as France or Holland is quite possible. This will make a difference, considering that concentration will mean Muslim populations in certain urban areas will be 40% or more. Enough to change elections.
Steyn’s larger point is that *something* will happen if there are two populations in an area, one well below replacement rate fertility and one well above. Especially when, as you say, a nation is unable (due to the self-censorship of multiculturalism) to state what it means to be to be a citizen of that nation. The threat of puritan Islam (if assimiltation doesn’t weaken it) is the same as the threat of wacko Christians in the US who want to place religon above science, human rights, and democracry.
Comment by spacebar — 3/10/2007 @ 5:11 pm
I do think that there’s a lot of interesting discussion to be had on the issue; I just don’t think that Steyn’s the guy to lead it. After all, what happens will depend on the numbers more than it depends on anything else. I don’t have a problem with the idea that there may be a problem, just with Steyn’s populist hackery and apparent fact avoidance.
The other thing worth mentioning is that family sizes can quite likely decrease amongst the descendents of migrant families, so projections that require birthrates amongst 3rd generation that are the same as from previous generations would require some strong justification. I guess that is a matter, in part, of assimilation, but some of the pressures that lead to large families in their family’s countries of origin are significantly lessened in the West.
Comment by Adam — 3/10/2007 @ 5:25 pm
Steyn should be part of the debate; no one said he should lead it. Numbers are not his strong point; he started life as a music critic, not a statistician. I haven’t read America Alone, but his columns on Europe cover three themes: demography, European over-reliance on feeble institutions like the UN, and the effect of political correctness in not even allowing debate on minorities. Numbers are not the key factor here. Islamic minorities are already the #1 problem minority (in terms of violence and alienation) in Britain and France, and they are only, what, 7% of the population. Given the events of 9/11 and 7/7, the cycle of demonization and alienation will most likely worsen. Closing militant mosques and deporting imams may improve things, or make them worse. To say there “may” be a problem seems terribly out of date. There is a problem now. And to note that this same minority is also the fastest growing population group is to say something important about the problem.
You may very well be right and this will all peter out in the next five years. But if it doesn’t, due to wars ocurring in Israel, Lebannon, Iraq, or Iran, then Europe 15 years out will in deep trouble.
Comment by spacebar — 3/12/2007 @ 10:58 am
You can’t say that numbers are not the problem and then give a bunch of statements the truth of which depends on the numbers. Numbers are not the be- and end-all but they are the foundation on which all the other discussion must rest. Otherwise, we might as well be discussing the chance of an extra-solar body hitting the planet; much less likely than moslem-inspired disaster, one assumes, but if we’re not crunching numbers, why not pick what would be, if it happened, a much bigger problem?
Steyn needn’t be a statistician; all he has to do, minimally, is give the source of his numbers. For extra points, if his analytic capability is lacking or he feels that it is, he could ask someone that is a better numbercruncher than him to do a sanity check and help with projections.
I don’t know if it will peter out or not; demographic projections are not my field of expertise. I just think that we should pretty much ignore contributions whose foundation in reality is at best unclear. Otherwise, let me be the first to warn against the dangers of an invasion by intelligent wolverines bent on World Domination.
Comment by Adam — 3/12/2007 @ 12:41 pm
The point of all this is to say that my position is that we have to know the numbers and that the numbers that we use in the debate have to be sourced. If Steyn had said where he got his numbers from then we could look at the source and evaluate the quality of the projections he quotes. As he didn’t, what are we supposed to do? Assume that the numbers are right?
I will reiterate that I haven’t read his book; I’m talking about the article from 2005, which was brought to my mind by Hari’s review (linked through Sully) of the book.
Comment by Adam — 3/12/2007 @ 1:14 pm
40% by 2025 is ridiculous. But 25% in countries such as France or Holland is quite possible.
I am not even sure that this makes sense. Polling in France shows that only 3% of the population identifies “most closely” with Islam.
Éléments d’analyse géographique de l’implantation des religions en France – Dec 2006
The poll asks which religion do you “most closely” identify with:
- 64% said Catholicism.
- 27.6% said no religion.
- 3% said Islam.
- 2.1% said Protestantism.
- 0.6% said Judaism.
It is forbidden for the government to ask people what religion they practise and the lack of data leads some (mostly in the U.S. strangely) to erroneous conclusions. People outside of France seem to confuse being of north African or Middle Eastern decent with practising Islam. 10% of the population comes from north Africa or the Middle East but most don’t identify with any religion.
The United States will become a Latino Spanish speaking nation long before French becomes a Muslim country.
Comment by toujoursdan — 3/12/2007 @ 9:13 pm
Your poll results of 3% seem on the low side. Wikipedia reports a range between 3 and 10. And the BBC (http://news.bbc.co.uk/2/hi/europe/4385768.stm) reports 5-6 million Muslims in France (8 – 9.6%).
Adam. “You can’t say that numbers are not the problem and then give a bunch of statements the truth of which depends on the numbers.”
No I didn’t. I stated that Steyn uses numbers in only one of his three ideas on Europe. Then the only number I used was Muslims make up 7% of Britain and France. How can you refute this? If you say the actual number is 3% then you’re making my point because a minority that is a serious problem at 3% can only get worse as it grows. And if you say the actual number is 12% then you’re supporting Steyn’s sorts of numbers.
You live in Britain I think. What is your memory of the days after 7/7? When it came to light that the killers were not foreign terrorists, but home-grown, cricket-playing Northern England boys. I remember (seen from Canada) a huge national shock in Britain.
You seem partly in denial. Surely you would agree that
a) Britain is #1 or 2 country most likely to be targeted by islamic terrorists.
b) Britain’s #1 current terrorist threat is not the IRA or PETA. It’s Islamo-terrorists.
c) Britain has had multiple attacks or near attacks. 7/7, the plot on Heathrow, the plot on the airliners this summer with liquid explosives, and the recent plot to kidnap and behead British soldiers.
Given this, and the probability of a major mid-East war in the next ten years as at least 50%, it would seem that serious trouble is a serious probability.
Maybe I’m missing something. How do you see all this as not a problem?
Comment by spacebar — 3/13/2007 @ 9:44 am
In the absence of any official survey different forms of calculation are used.
The figures cited by the BBC and many American sources seem to take the total population of migrants from predominately Muslim countries and assume that everyone who is from one of these countries is a [self identified or practising] Muslim. The wording of the BBC paragraph suggests that this is the criteria used.
The problem with that is that many from North Africa and the Middle East came from mixed Muslim and Christian countries – like Lebanon, Egypt and Syria or represent the most Francophile [secular or converted to Catholicism] segments of nations like Morocco, Algeria and Tunisia who left when France pulled out or (like the Turks) may be very secular even at home. Counting these as practising Muslims (particularly of the fundamentalist variety, which seems to be the alarm in books like these) doesn’t make much sense. They population is much more diverse than this. Islam itself is much more diverse than this.
The French poll I cite samples people from across the country and actually asks people which religion they feel closest to, (which still doesn’t necessarily mean they practise it.) This takes into account that people from predominately Muslim countries that may not practise or feel tied to their religion in their new home.
I understand that the US has an industry of these alarmist books that seem to play on Islamiphobia and contempt for European “socialism” to sell lots of books, but they don’t seem to take into account that Islam itself is a diverse religion with a growing progressive movement in the West (represented here in Canada with groups like the Muslim Canadian Congress and Canadian Muslim Union) and that secularization is happening with many 2nd and 3rd generation immigrants from Muslim countries, just like it does with every other immigrant community.
Comment by toujoursdan — 3/13/2007 @ 10:30 am
Spacebar, you said:
Islamic minorities are already the #1 problem minority (in terms of violence and alienation) in Britain and France, and they are only, what, 7% of the population.
That’s a statement that can be examined and justified numerically. I’m not saying that it’s wrong (as I have no idea), just that it has to be justified and that we have to agree on figures of merit with which to measure problems. There is truth to be had, subject to agreeing metrics, but that truth is going to have to be collected and it’ll inevitably be numerical. Once we have a better handle on that, we can certainly talk about problems and their solution.
What comprises a ‘major mid-East war’? From where did you source your 50% probability (clearly, the probability depends on what constitutes a ‘major war’, if it can be assessed at all).
Anyhow, my main point was that Steyn shouldn’t use numbers to support his case and not provide their source. That’s particularly true when his numerical claims are so enormously spectacular and on which the arguments so critically rely.
My examination of the NIC report and the Scotsman article and others was driven by my desire to find his figures, but all I found were similar looking figures, in a context consistent with the one he described in his article, that had ‘millions’ instead of ‘percent’; this was in a ‘fictional scenario’, and was followed by an imaginary letter written by Bin Laden’s grandson. If I failed to find his actual sources, that’s because he didn’t list them.
Comment by Adam — 3/13/2007 @ 10:47 am
“I’m not saying that it’s wrong (as I have no idea)” … “subject to agreeing metrics, but that truth is going to have to be collected and it’ll inevitably be numerical.”
Wonderfully bizarre. You claim no opinion on the terrorist threat in your country. And you demand a fully quantitative discussion complete with agreed metrics. High standards! So how can it be that the author of the following highly un-numerical statement elsewhere on this blog is yourself:
“That adultery is not illegal in civilian society is not because it’s condoned by the Government, even though it appears to be somewhat extensively practiced by many of them.”
Metrics for ‘extensively’ or ‘many’ are missing, as is a definition for adultery. Is oral sex adultery??? But of course I get your point; just as you get mine :)
Anyway, we should probably read Steyn’s book before further debating the quality of its contents. The strangest fact of all is that France doesn’t even track the numbers of Muslims in its population, leading to the wide discrepancy between the BBS and toujoursdan’s poll.
Toujoursdan: don’t you think there’s a teeny chance that people will under-report their Islamic beliefs given today’s climate? It would take more than one poll to convince me that all the other sites are wrong by a factor of 3. There was a British poll showing 25% of young Muslims agreed with the tactic of terrorism. Do you believe that poll? I don’t.
Comment by spacebar — 3/13/2007 @ 6:34 pm
I don’t live in the UK anymore, but more importantly I am not prepared to generalise from anecdotal evidence to the country as a whole (and, in fact, if I did, most of the anecdotal evidence/experience I had would suggest that the overwhelming majority of moslems in the UK were providing a net benefit to the UK). My point is that numbers are key. We can wave our hands all day about what we want, or fear, or feel to be true but the actual truth, whatever it is, will be in the numbers.
I’m not going to pay for Steyn’s book, however. I’m not even posting about Steyn’s book, for that matter; I’m posting about an article he wrote in 2005, that was brought to mind by Hari writing about Steyn’s book. Andrew Sullivan linked this article on his blog and Hari linked it on his own website. I don’t know whether Steyn has seen it yet or whether someone has asked him for an answer. I hope that he will provide the source of his numbers so that we can look at them.
Comment by Adam — 3/13/2007 @ 8:00 pm
Toujoursdan: don’t you think there’s a teeny chance that people will under-report their Islamic beliefs given today’s climate? It would take more than one poll to convince me that all the other sites are wrong by a factor of 3. There was a British poll showing 25% of young Muslims agreed with the tactic of terrorism. Do you believe that poll? I don’t.
All this poll asked is which religion you felt closest to. Comparing that with a politically leading question like your UK poll doesn’t even make sense.
Secondly, I already gave the reason why there is a huge difference with some other polls and the assumptions built into both. If you want to take issue with it and argue that everyone who comes from a predominately Muslim country is a practising Muslim, make that case.
I don’t think they would under report it. The climate in France isn’t the same as in the US or the UK.
Comment by toujoursdan — 3/14/2007 @ 9:59 am
[...] few days ago, I made this post questioning the identity of Mark Steyn’s source for some numbers in a 2005 article which seem [...]
Pingback by The Crossed Pond » ‘Steyn and his numbers’ update — 3/15/2007 @ 8:13 am
As I have said, I haven’t read the book and am not commenting on it (unlike Hari, who read the book and was commenting on it). Reading his article, as linked from Sullivan’s page, was the thing that reminded me of the article.
My quibble was with the entire article because its relevance was clearly based on the figures given in the single sentence; if those figures were way wrong, why all this talk of a problem?
I haven’t actually said that his figures are wrong, although I did find a contemporeaneous account that fitted the sorts of things that he and the Scotsman article were saying but which had millions instead of percentages. I am merely wondering from where he got them, so I could make a guess as to how seriously he was really looking at the numbers.
I would also say that a 5.3% population increase (you may be aiming highish there as it’s over 18 years) is unlikely; nearly all that population base in your projection is from Turkey and the current Turkish population growth rate is estimated at 1.06% (2006) as you can find here. So, absent something that will cause a 500% increase in the Turkish birthrate, I would suggest that perhaps you are wrong. There will be a certain compounding effect as a growing population become old enough to have children, so the figure of 5.3% would probably be higher than is needed to make up those additional millions, but I don’t know what the demographic is of the population likely to be having children over the next 18 years.
Another issue is that the projection depends on Turkey joining the EU (which may or may not happen) but doesn’t consider whether other countries will join the EU. Romania and Bulgaria joined in January this year and added 30 million people who were overwhelmingly non-moslem, for example, and there’s strong rationale for further expansion (particularly if, like me, you favour a large trading bloc with a weak central governance). In particular, there is a strong movement in Ukraine, a non-moslem nation with a population of 46 million, to join the EU, as well.
As an aside, the point that toujoursdan has made, which I have completely neglected because I’ve been adressing what I imagine Steyn means by ‘moslem’, is a good one; the number of people that are ‘moslem’ in the sense of believing or professing belief in Islam is likely to be somewhat smaller than the overall population of, say, Turkey. That’s a matter for discussing implications of population numbers, though; I’m just dealing, narrowly, with the article to which I have access and to which I have posted a link.
Comment by Adam — 3/15/2007 @ 5:03 pm
You say that you “don’t necessarily agree with Hari”. That’s wise, since Hari puts Steyn to shame in the playing-fast-and-loose-with-statistics department. Hari brazenly asserts that Steyn predicts a European Muslim population of 200 million by 2020; in fact that prediction is Hari’s own manufactured straw-man of Steyn’s thesis.
Your own quibble appears to be limited to a single sentence in a single two-year-old column. You are correct that Steyn does not cite the source of “some projections”. Have you read any footnoted, annotated op-eds recently? Me neither. I doubt that the 40% figure is for the entire EU population. Could be working age, or child-bearing age, or under 21. It is quite possible that some such qualifier failed to make it into the published version (ask Mr. Hari about editorial redactions; the editor at the New Statesman excised two-thirds of his review for publication).
But let’s say that Steyn really meant 40% of the EU total. Your deconstruction of that number is extremely simplistic. You simply take 40% of the EU’s total 2005 population, come up with 182 million and ask rhetorically where they are all to come from. This ignores several factors. First, the EU countries are entering a period of steep population decline (see here). 456 million may be optimistic; and, whatever the number is then, the replacement population being born or immigrating now to make up for deaths is disproportionately Muslim. In other words, it’s not just rising numbers of Muslims contributing to that percentage but declining numbers of native Europeans. Second, “15 million or so” (your number) are already there, so actually it is 167 million needed, not 182 million. And finally, if Turkey joins the EU, that adds 68 million Muslims (2005 figure; 2006 population is 70.14 million).
So let’s take the EU plus Turkey in 2005: 83 million Muslims, 443,000,000 non-Muslim, 526 million total. That’s 15.7% Muslim. Now let’s take your assumption that the total remains static: 526 million times 40% yields 210.4 million. 83 million were already there in 2005. So to get to 40% requires an increase of 127.4 million, or about 7.1 million per year for the next 18 years, or about 5.3% annual increase in the Muslim population. Is that impossible?
Comment by SnidelySnark — 3/15/2007 @ 4:32 pm