Christopher Caldwell: Untrustworthy on Facts

Christopher Caldwell, senior editor of neocon house journal the Weekly Standard, once wrote a six-page feature in the New York Times magazine in which he claimed that Robert Kilroy-Silk would “transform European politics”. Despite this, he is still taken seriously by some people; disturbingly, this includes the editors of the Financial Times. In his column this weekend, he issues a rant against trades unions and specifically French ones. I am not going to trouble my readers by taking issue with his ideological position; this is well-known, hence there’s no informational gain in arguing with it.

Instead, I’m sticking to his factual assertions.

Sixty per cent of SNCF cancellations are due to strikes.

The only source for this statement I can find is the French Government’s spokesman; anyway, as the overall 10-minute punctuality rate is of the order of 90 per cent (source: SNCF Annual Report 2006), this is equivalent to saying that 3 trains in every hundred are affected by industrial action. In fact that is a considerable overstatement itself, as not all trains that run 10 minutes late are cancellations. Anyway, this is a theoretical issue; Le Canard Enchaine published the actual figures, according to which strikes accounted for 140 out of 6,043 delays recorded in 2006 – about 2 per cent. Caldwell is wrong.

The young anti-union orator Sabine Herold drew tens of thousands to her speeches during the strikes of 2003.

Fortunately, I’d recently seen some of her old election posters, so I actually knew who she was, which puts me ahead of the vast bulk of the French public. It is actually possible that Mme Herold pulled in at least 10 kilodemonstrators; French Wikipedia claims she did, citing Le Monde as saying she got 30,000, but I can’t find a root-source for this anywhere; just a lot of wingnuts clapping each other on the back. It hasn’t stopped her claiming 100,000 in order to sell books. But it’s hard to be sure, as her political party didn’t get enough votes to be broken out independently in the official results of the 2007 parliamentary election. Neither could they find 500 local councillors willing to sign their presidential nomination. To place a lower bound on her popularity, though, we can say with certainty that she pulled some 345 votes on her home turf, the very bourgeois 16th arrondissement of Paris. That is, 1.4 per cent of the vote. Her fellow leader, Edouard Fillias, pulled a whacking 228 votes in the 12th – 0.52 per cent.

This didn’t stop various right-wing anglophone papers lionising her; fortunately she kept the tributes on her own website. Here’s Matthew Campbell of the Sunday Times predicting that if Segolene Royal wasn’t elected, she might be. Here’s the Daily Telegraph asking whether she really did speak for millions. I think you got your answer, son.

Anyway, moving swiftly on:

They rest on government-accorded privileges, particularly that of compelling membership, whether formally or informally – a privilege that, if it were exercised by a church or a political party, would horrify the public.

“They” are trade unions; it’s a pity Caldwell appears not to know that the closed shop has been illegal in France since 1956.

9 thoughts on “Christopher Caldwell: Untrustworthy on Facts

  1. strikes accounted for 140 out of 6,043 delays recorded in 2006 – about 2 per cent. Caldwell is wrong.

    And how many cancellations? I hope for SNCF’s sake that only a small fraction of delays are cancellations.

  2. I hope for SNCF’s sake that only a small fraction of delays are cancellations.

    Undoubtedly.

    Trains, unlike aircraft, generally aren’t affected by adverse weather, so there aren’t all that many reasons why trains would be cancelled outright (as opposed to merely delayed). It therefore is not at all surprising that 60% of outright cancellations result from labor actions. I am presuming that scheduled cancellations, for example those caused by weekend trackwork, are not being counted.

    As for the size of Herold’s rallies, let me point out that it is very difficult to determine the attendance of a non-ticketed rally or demonstration with any degree of accuracy. Crowds tend to be fluid, making a physical count nearly impossible in most circumstances, and in any event many people have an incentive to overestimate or underestimate the size of a particular rally.

    As an amusing illustration, I read not long ago that if all the people who supposedly watched the Tournament of Roses Parade (held in Pasadena, California each New Year’s Day) in person actually lined the parade route, given the length of the route and the depth of the sidewalks, there would be tens of thousands of deaths from suffocation and trampling.

  3. However, Caldwell is evidently trying to suggest that large number of trains are cancelled due to uppity proles. It’s quite possible to suggestio falsi by asserting something accurate in itself but misrepresented.

    Regarding rally size, I always like the Malatesta estimator. Total the estimates, then take the mean, then subtract 30 per cent; this gives a value of 29,761, but I suspect it’s skewed by the bullshit 100,000 claim.

  4. Regarding rally size, I always like the Malatesta estimator. Total the estimates, then take the mean, then subtract 30 per cent; this gives a value of 29,761, but I suspect it’s skewed by the bullshit 100,000 claim.

    Sounds sensible.

    Another problem with counting attendance at a rally or demonstration, especially one held in a public place, is determining how many of the people on the scene acutally are participating, and how many of them just happened to be in the area and merely have come over to see what all the fuss is about.

  5. In this case, there was also the point that the demo had been called by far better-known rightwing organisations, whose supporters most of the protestors were.

  6. My comment is slightly off topic, but I like that you are checking so carefully the original sources of the information that you are using in this post.

    I believe that doing this is becoming increasingly important in the blogging world in comparison with traditional media. This is because sources are even more diversified and the threshold to communicate is lower. We are thus less likely to know the trustworthiness of the sources that we are using.

    It should not surprise us then that companies like Datops have even developed software especially for that purpose (tracing the original source of information).

  7. Crowds tend to be fluid, making a physical count nearly impossible in most circumstances, and in any event many people have an incentive to overestimate or underestimate the size of a particular rally.

  8. Pingback: ArtikelverzeichnisStars und Sternchen

Comments are closed.