In this post, I would like to dissect another claim in Christopher Caldwell’s book „Reflections on the Revolution In Europe: Immigration, Islam, and the West“ which was published in 2009. As I have already explained elsewhere („Enoch Powell’s Secret Formula“), Caldwell’s main technique is to frame the issue early on with some strong images. In this case, it is the Muslim takeover of Europe. If you buy into this scenario, you don’t need proof, all you need is some anecdotal evidence that is not totally at odds with the main storyline. Here is a paragraph where Caldwell explains how the Muslim takeover of Europe is already under way:
„Muslims now either dominate or vie for domination of certain important European cities. A partial list of them would include Amsterdam and Rotterdam in Holland; Strasbourg and Marseille (and many of the Paris suburbs) in France; Duisburg, Cologne, and the Berlin neighborhoods of Kreuzberg and Neukölln in Germany; and Blackburn, Bradford, Dewsbury, Leicester, East London, and the periphery of Manchester in England. Such places may, as immigration continues and the voting power and political savvy of the Muslims already there increases, take on an increasingly Muslim character.“
Those are long sentences. So let me boil the paragraph down to what a casual reader might understand:
„Muslims now … dominate … important European cities. A partial list of them would include Amsterdam … in Holland; … Paris … in France; … Berlin … in Germany; and … London … in England. Such places … take on an increasingly Muslim character.“
Of course, Caldwell also knows of many important European cities that you have never heard of: e.g. the metropolis of Blackburn is teeming with 105,085 inhabitants, and the metropolis of Dewsbury boasts 62,945 residents. 153,887 and 167,248 people call the Berlin neighborhoods of Kreuzberg and Neukölln their home. Strasbourg has 276,170, Leicester 337,653 inhabitants. Towns that might qualify as major cities are Rotterdam at 633,471, Duisburg at 491,231, and Bradford at 528,155. Cologne is the only town with slightly more than a million residents, Amsterdam at 838,338 and Marseille at 855,393 at least come close. As for the really big cities, Caldwell carefully speaks of „many of the Paris suburbs“, „East London“, „the periphery of Manchester“, and „Berlin neighborhoods.“
Let me put this in perspective. The impressive list of „important European cities“ would be on a par with the tenth-largest city in the US, San Jose (Cologne), the 14th largest city, Indianapolis (Marseille), the 16th largest city, Fort Worth (Amsterdam), the 29th largest city, Baltimore (Rotterdam), the 33th largest city, Tucson (Bradford), the 35th largest city, Sacramento (Duisburg), the 57th largest city, Santa Ana (Leicester), and the 66th largest city, Cincinnati (Strasbourg). Blackburn and Dewsbury would not even make it into the top 200. And for the really big cities: the metropolitan areas of London and Paris are in the same ballpark as New York, Berlin as Los Angeles, and Manchester as Chicago. But there we are only talking about some suburbs, neighborhoods, or the periphery.
How impressed would you be if I posed as an expert on what is going on in the US and came up with this list of American cities as typical for the US? And remember that Europe has about half a billion inhabitants, the US only somewhat more than 320 million. If I add up only the cities (not the neighborhoods), then all Caldwell is talking about is a population of 5.5 million people, or slightly more than 1 percent of the total population of Europe. As for the neighborhoods in the big cities, I would have to add perhaps a few million, and that might bring it up to 2%. That’s all it takes to conquer Europe. Like after the Normandy landings, the Allied troops stayed on the beach and said to themselves: „That’s good enough, let’s call it V-Day.“
Now, you may object that I am being very unfair with Christopher Caldwell here. He wrote that this is just a „partial list“ and he surely could name lots of other cities with ease. Well, no, I’m not being unfair, I am actually pretty lenient with him. It is is indeed a „partial list“, but only in the sense that “partial” is the opposite of “impartial.” Christopher Caldwell does not give you a random selection from a longer list. For each country, he took the cities that have the highest share of Muslims. Any further examples would either have to be rather small cities, or have a lower share of Muslims, or both. He already had to resort to Dewsbury for this list. No offense to people from Dewsbury who read this, but I had never heard of your metropolis before I read Caldwell’s book.
But then at least those are dominated by Muslims?
Well, how can I say this, I am so sorry. It’s not going well with the Islamization of Europe. Hope you haven’t bet on it yet. I won’t go through all the examples in detail, but will concentrate on the claim for Germany. To this end, I downloaded data from the Federal Statistics Office for 2014.
There are 424 districts in Germany with roughly 200,000 inhabitants on average (technically: “Regierungsbezirke”, “Kreise”, and for small states the whole state) . Most Muslims come from Turkey. The district with the highest share of Turkish citizens was Duisburg with 6.5% of the population. Cologne was the next major city with 4.9%. As I said, Caldwell’s selection is anything but representative.
The percentages are for Turkish citizens only, but there are also German citizens of Turkish descent. On the whole, about half of all those of Turkish descent have German citizenship (roughly a third of the latter have also Turkish citizenship). I don’t know whether those with dual citizenship were also counted as Turkish citizens, but I am on the safe side if I assume they were not. So you would have to double the share of Turkish citizens to get a rough estimate for all those of Turkish descent.
In other words: Even for the two cities with the highest share of Turkish citizens, the share of all those of Turkish descent only amounts to something like 13% and 10%, respectively, a far cry from a majority. You may add a few percent for those who are Muslims, but are from other countries. Strictly speaking, though, you would also have to subtract those who do not consider themselves Muslims although they are of from a Muslim country. No matter how you slice it: There is not a single district in Germany out of 424 where Muslims come even close to a majority.
The list I am working with does not include the neighborhoods of Kreuzberg and Neukölln (only Berlin as whole). However, for both the share of all immigrants and their children (not only from Muslim countries) is well below 50%. No part of Germany of any remarkable size has a Muslim majority, not by a far stretch. None.
As I said, I concentrate on Germany here. But I have also checked the numbers for other cities on Caldwell’s list. Here are the shares for Muslims that I have found: Amsterdam 14%, Rotterdam 13.3%, Strasbourg 19%, Marseilles 23.5%, Blackburn 25.7%, Bradford 24.7%, Leicester 18.6%. I could not find data for Dewsbury which is perhaps not surprising for such a small town. As for East London, Tower Hamlets has a share of 34.5% and Newham of 32% Muslims. The suburb of Paris with the most immigrants is Seine-Saint Denis which has a share of 17.3% immigrants born in non-EU countries (not all of whom are Muslims). Even if you double this percentage to include their descendants, it does not look like you could find a Muslim majority even there.
Literally in no city on Caldwell’s list is there a Muslim majority. Surely, you will find some part of a neighborhood with a Muslim majority if you drill down even further. But that means you are now looking at a fraction of 1% or 2% of the total population of Europe.
But doesn’t Christopher Caldwell only say that: „Such places may, as immigration continues and the voting power and political savvy of the Muslims already there increases, take on an increasingly Muslim character.“ Of course, he does. After selecting the most extreme 1% or 2%, if you can also introduce additional qualifications like “more immigration in the future that only ends up in these locations,” and “increasing political savvy of those already there,” then you are able to construct a case where a tiny sliver of Europe „may take on an increasingly Muslim character.“ Note how carefully worded that is.
It all only works because of framing: You first have to subscribe to Caldwell’s overarching story how this is only a start in an ongoing takeover, and then you can extrapolate to this totally unimpressive result. And on top of this, he tones even this claim down with a further qualification: “Muslims now either dominate or vie for domination of certain important European cities.”
Last defense line: But Europe is dying out, and Muslims have extremely high fertility rates.
Sorry, that doesn’t work either. Fertility for Turkish immigrants in Germany went down to about 2.5 already in the 1980s. For the second generation, it looks like it is already below replacement level. (Cf. footnote  below.) And then fertility data for Germany are heavily distorted downwards. With rising age at birth, total fertility rates undercount actual fertility. If you correct for this so-called “tempo effect,” you already had fertility of about 1.65 for Germany a decade ago. (Cf. footnote  below.)
And it has gone up a little since then, so the current number should be more like 1.7 or above. But that is not all. Since mean age at birth is about 31 years in Germany, you would have to make this figure comparable with shorter generation lengths for other groups. If you do this for a mean age at birth of 25 years, you can add another 0.05 or so, and you are safely in a range from 1.7 to 1.8 which is actually only marginally lower than for countries like the US, and also for those of Turkish descent in Germany. Just do the math how long it would take to turn a minority of about 6% into a majority with such marginal differences.
Here is the takeaway of my post:
- Christopher Caldwell selects a list of cities that is presented as if it were a representative sample, but which consists only of extreme cases for what he wants to show. Talk about selection bias.
- Even in those extreme cases, which make up only 1% or 2% of the total population of Europe, Muslims are not a majority, mostly not even close. For Germany, the assertion is resoundingly false everywhere.
- Caldwell must know this. That’s why he equivocates between “dominating” and “vying for domination,” and that’s also why he introduces further assumptions like more immigration in the future and increasing political savvy to end up with a lame conclusion that some cities „may take on an increasingly Muslim character.“
- The fiction of a dying Europe doesn’t save his argument either.
- All it amounts to is good framing, so you accept on faith that a Muslim takeover is already under way. And then Caldwell throws carefully selected data at you and leaves out relevant information, so a casual reader, especially with little knowledge about Europe, can feel vindicated.
- There is no ongoing Muslim takeover of Europe.
 Katharina Wolf: „Fertility of Turkish migrants in Germany: Duration of stay matters“, MPIDR Working Paper WP 2014-001, 2014, table on page 17. — Werner Haug, Paul Compton, Youssef Courbage: The demographic characteristics of immigrant populations, table on page 227. — Susanne Schmid & Martin Kohls: „Generatives Verhalten und Migration“, 2011, table on page 189, where the fertility for women of Turkish descent around 2007 is estimated as 1.80 or 1.85 depending on the data sets used.
 Marc Luy & Olga Pötzsch: „Schätzung der tempobereinigten Geburtenziffer für West- und Ostdeutschland, 1955-2008“, Comparative Population Studies – Zeitschrift für Bevölkerungswissenschaft, Jahrgang 35, 3/2010, p. 569–604, cf. page 585.