Despite being a leftist, I’m going to be fair-minded and make some kinder comments about David Cameron’s speech on immigration (initially). Firstly, he is of course entering into a legitimate debate about migration. While the question of asylum is separate, economic migration is a debatable good: it’s perfectly fair to ask how much migration is desirable and what the overall pros and cons are for Britain. Furthermore, Cameron was explicit that migrants can create jobs as well as filling them, taking on the idea that there’s some sort of lump sum of labour in an economy. And he did attempt to draw some key distinctions (e.g. about the speed of change, rather than necessarily the principle): this was not on a par with Michael Howard’s 2005 campaign.

The fact remains, though, that there are lots of straw men lurking in the speech. Anyone who’s listened to Jack Straw, David Blunkett or Phil Woolas will be a bit surprised to hear that the last Labour government gave the impression that it was racist to talk about immigration, for instance. Nor is it particularly helpful to claim that the government which brought in a points system promoted the view that any attempt to control immigration was madness. And was there really a ‘mass relativism’ about sham marriages? This isn’t on the scale of 2005, but there are still nasty undertones here – in the run-up to elections, at a time when the economy is in a grim state and voters are angry. Politicians who declaim loudly about how ‘we aren’t allowed to say anything about immigration’ are playing a dangerous game.

Cameron was also making dubious use of statistics at best. For a start, while looking at EU migration, he got his figures wrong. Long-term net EU immigration ran at 57,000 from June 2009 to June 2010, not 27,000. The choice of year was also misleading at best, at a time when there has been partial unwind of the post-2004 migration from eastern Europe. Note, I’m not complaining about that migration: but in this instance, Cameron is underplaying it in order to argue that his proposed cap will be effective when, in recent years, it would have usually had little effect. And by the way, if you’re worried about destabilisation and impacts on local services, net long-term migration is not the only possible concern: the ‘churn’ of migration, causing rapid fluctuations for councils, can also cause problems – not to mention internal migration within the UK.

The statistical details matter because the misuse of figures contributes to a wider argument: that we can meaningfully control immigration, in the sense of reliably determining a net total migration figure for the UK. For the purposes of this post, I’m going to sidestep my views on preferred migration levels (fluctuating depending on economic circumstances, but probably on the liberal end of the spectrum, if you’re interested). As a simple matter of fact, irrespective of what we’d like to do, it is not the case that we can control immigration. We can, perhaps, manage migration, in the sense of knowing (broadly) who’s coming in and making sure the system works better. But EU migration is a very large part of the whole, and we can’t restrict that (transitional controls on new members are just that – transitional). Free movement of labour is, rightly, central to the EU Single Market – and we can’t have one without the other. So economically, we haven’t got much alternative in terms of EU migration.

Furthermore, our borders are always going to be fairly porous: how could it be otherwise in a country with quite so many visa waiver agreements, for example? (And one of the biggest categories of overstayers is from those countries, by the way.) The fact that we have something like 400,000 illegal residents in the UK says something about the difficulty of enforcing border controls. I’m sure we can do a fair bit to make them more effective, but I also suspect that there is a direct relationship between how restrictive our overall policy is and how much illegal migration we end up with.

Ultimately, Britain is a small island near a bigger continent in a world of mass travel. We are lying to our voters if we pretend that immigration can be fundamentally controlled: or at least, controlled without unacceptable consequences in terms of civil liberties and day-to-day life. We can affect levels at the margins, we can monitor what’s happening, we can try and affect the drivers of demand for migration – beefing up HMRC’s minimum wage compliance would be a start. But if we’re being honest, that’s about it.

That is a very unpopular statement, but it’s also true – and an honest conversation about immigration will, at some point, have to include an admission of that basic fact. How do we go about admitting that migration is, in this sense, ‘out of control’ and unavoidably so?  And where should immigration policy go from that admission? I don’t have anything like a full answer. I suspect we probably have to talk about pull factors in the UK: has immigration been encouraged by low wages at the bottom of the labour market, making it more difficult for UK citizens to leave the benefit trap? Would a living wage mean more British citizens could fill jobs in Britain? Would greater regulation of the labour market – in at least some areas, some of the time – play a role? And should we be thinking about ways of harnessing some revenue from immigration (work permit fees, in particular) and investing it in training British citizens?  (The Liberal Democrats had a policy along these lines at one point.)

I’m sure there are other, better suggestions to be discussed: these are mainly starters for ten. But sometime, somehow, we need to have a more honest conversation about this – and stop pretending either that the establishment is conspiring to shut us all up or that we could buck the trend if we wanted to, if only we were a bit more competent. Because if we carry on arguing that migration levels can be reliably determined as a matter of political choice, and keep failing to deliver a given level of migration, more and more people will reasonably ask why we’re not delivering. And those answers could produce some ugly results.