The ‘lump of terrorism’ fallacy

by Henry Farrell on September 7, 2003

The “Washington Post”: reports today that

bq. The occupation of Iraq — once the home of the caliph, or universal leader, of Muslims — is a galvanizing symbol for radical Islamic groups. On Internet sites and in mosques across the Islamic world, thousands of potential fighters are hearing — and heeding — calls to go to Iraq to fight the infidel, according to European and Arab intelligence sources who have tracked some of the movements of the recruits.

Dunno how true this is – “Juan Cole”: thinks that the Post is exaggerating wildly – but it got me to thinking about how the “flypaper” theory beloved of “Glenn Reynolds”: and his crowd is based on a fundamental error of logic. If you look at it closely, it distinctly resembles a fundamental mistake that economists call the “lump of labour fallacy”:

Much bad economic punditry starts from the premise that there’s a ‘lump of labour’ – a fixed amount of work to be done in the economy. On this argument, if you want to reduce unemployment, you can do it by lowering people’s working hours, so that work is ‘freed up’ to be shared with the unemployed. Of course, this argument doesn’t hold water – the demand for labour isn’t a fixed constant in real economies. Instead, it varies, depending on a host of other factors (which themselves are likely to be affected, perhaps in perverse ways, by any ham-handed efforts to ‘share the jobs around’).

Similarly, the ‘flypaper’ theory implicitly assumes that there’s a fixed amount of al Qaeda terrorism sloshing around in the international system, so that it’s a good^1^ idea to divert it from the US to Iraq – more terrorists attacking troops in Iraq would mean less terrorists attacking the homeland. But there isn’t a fixed amount – instead, US actions in Iraq are almost certain to affect the ‘supply’ of al Qaeda terrorists. Indeed, the WP article suggests that the US occupation is leading to a substantial increase in the willingness of potential fighters to take up arms, so that the invasion isn’t just drawing existing al Qaeda combatants to Iraq; it’s creating new recruits.

The jury is still out on whether the Post is right or not on the facts – but it’s demonstrably true that the actions of the US (in invading Iraq, in how it behaves within Iraq) are going to affect terrorism on the supply-side. Pro-war types can still try to make the case that the US invasion is going to decrease terrorism in the long run (they have their work cut out for them), but ‘lump of terrorism’ theories like the flypaper argument are bogus, and should be treated with the ridicule that they deserve.

^1^ Of course, ‘Good’ here means ‘good for Americans,’ not ‘good for Iraqis.’


by Brian on September 7, 2003

Geoffrey Nunberg has good column in TAP about the strange history of the word liberal in America. Maybe I should have expected the following data, but I was really stunned by how strong the race and class connotations of liberal have become over here.

bq. From a semantic point of view, this negative branding campaign has largely succeeded in changing the meaning of the word liberal itself. In major newspapers, for example, the phrases “middle-class liberals” and “middle-class Democrats” are used with about the same frequency. But the phrase “working-class liberals” is almost nonexistent; it’s outnumbered by “working-class Democrats” by about 30-to-1. And while “white liberals” is used about as frequently as “white Democrats,” the phrase “black Democrats” outnumbers “black liberals” by better than 15-to-1. The patterns are similar if you plug in “African American,” “Latino” and the like.

bq. By contrast, the press refers to working-class conservatives as frequently as it does to working-class Republicans — and far more frequently than it refers to working-class liberals. And there are five times as many references to black conservatives as to black liberals, though references to black Democrats vastly outnumber references to black Republicans. The implication is that unlike conservatives, liberals are rarely found among minorities or the working class. When those groups vote Democratic, it’s presumably out of narrow self-interest or traditional party loyalty rather than because of any underlying ideological commitment. From that point of view, the political attitudes that make someone a liberal are simply the outward expression of a particular social identity, no different from a predilection for granite countertops or bottled water. For all intents and purposes, liberal has become as much a referential term as Bolshevik was.