The Arithmetic of Absence

Britain argues about illegal migration as if it knew the numbers. Israel learned the hard way what an uncounted workforce is worth.

In the spring of 2024 Reuters described a three storey building going up in Israel. Six months earlier, 125 workers had been on site. Now there were 25, two of them wrestling with an unfinished floor at the far end of the hall. The rest was empty space.

Israel knew exactly who was missing and why. Britain, arguing once more about illegal migration after 934 people crossed the Channel in two days this week, does not.

Ask the obvious economic question, whether people living here unlawfully add to national output or subtract from it, and the first discovery is that nobody knows how many of them there are. Neither the Home Office nor the Office for National Statistics publishes an estimate. The latest independent attempts are for 2017. The Greater London Authority put the population at about 674,000, or 809,000 including their British born children. The Pew Research Center offered 800,000 to 1.2 million. Oxford’s Migration Observatory warns that all such figures carry large margins of error. Any answer about gross domestic product begins with a population nobody has measured.

An outline is still possible, provided we are clear which question we are asking. There are three, and the public argument routinely blurs them.

The first is whether these people add to total output. Almost certainly they do. Most are of working age, and on the available evidence most did not come by boat; they arrived on visas and stayed after the visas lapsed. They clean offices, staff kitchens, carry bricks and ride delivery bikes. Work done off the books is still work.

Nobody has published a figure, so here is a crude one with every assumption on the table. Take 600,000 to 900,000 adults. Assume between 60 and 75 per cent of them are in work. Assume they earn £15,000 to £22,000 a year, less than a full time job on the minimum wage pays. That gives a wage bill of roughly £5 billion to £15 billion. Since employees take roughly half to three fifths of output across the economy, the output attached to this workforce comes to something like £9 billion to £30 billion a year. Against an economy of about three trillion pounds, that is between about 0.3 and 1 per cent.

That is a gross figure. It overstates what removal would subtract in the long run, since some work would pass to others or be mechanised, and understates it in the short run, since no sector can rehire overnight. The honest net contribution to total output is positive and smaller than the gross one: tenths of a per cent, not whole ones.

The second question is whether they raise output per head, which is what decides living standards. Here the sign is uncertain. Low paid work in a handful of sectors adds people to the population about as fast as it adds pounds to the total. Research on irregular labour suggests some downward pressure on wages at the bottom of particular trades, softened by the minimum wage for those able to claim it. Workers who cannot go to an employment tribunal appeal to bad employers for exactly that reason. The gains go to employers and customers; the losses to the lowest paid workers in the same jobs.

The third question, the one most voters mean, is fiscal, and here the paradox is sharpest. Ian Gordon of the London School of Economics estimated in 2010 that probably two in three illegal workers paid some tax and national insurance. Some do so through borrowed documents, while entitled to almost nothing in return. Yet the most visible bill in British migration politics belongs to a different group. The National Audit Office put the cost of supporting asylum seekers at about £4.9 billion in 2024 to 2025, £2.1 billion of it on hotels. Most of them are not, in law, illegal residents. They are awaiting a decision, and the rules stop them applying to work until they have waited a year, and then only for graduate level jobs. The state pays people to be idle, then books the idleness as the cost of migration.

Modelling of the obvious remedy points one way, though the studies were commissioned by, or produced with, bodies already sympathetic to it. Work by the LSE for City Hall in 2009 put the gain from regularising long resident irregular migrants at about £3 billion a year. Jonathan Portes of King’s College London later estimated roughly £2,000 in extra direct tax for each person regularised. Critics answer, not unreasonably, that every amnesty advertises the next.

Israel offers something rarer than a model: an accident that ran the experiment for real.

Before 7 October 2023, more than 150,000 Palestinians from the West Bank entered Israel each working day, and an estimated 40,000 more crossed without permits. In construction they made up about a third of the workforce. After the massacre the government shut the crossings almost overnight. By the following spring around 40 per cent of construction was still idle, and contractors said wages for the workers they could find had doubled. Adi Brender, head of research at the Bank of Israel, put the likely drag on the wider economy at two to three per cent. Replacements from India, Sri Lanka and Uzbekistan came slowly; by October 2024 about 13,000 had arrived.

Here the gross and net figures came apart in real time. In the first months the loss was close to the gross value of the missing work, because nothing could replace it. Later, foreign workers began to arrive and some Palestinians went on crossing without permits regardless. Demand for labour finds a way; the only questions are the price and the delay.

Israel made its decision for reasons of security, and few Israelis would reverse it lightly. But it taught something economists rarely see so cleanly: the value of a workforce the statistics barely register becomes visible only when it disappears. And most of Israel’s missing workers were at least counted. The state had issued their permits and knew whom it was turning away.

Britain cannot run even that experiment, because it does not know whom it would be removing. A government promising more deportations, and an opposition promising many more, are each making an implicit forecast about output, wages and tax receipts. A serious removals policy would need a serious estimate of what it removes: which sectors, how fast replacements could be found, and at what wage. Neither side has published one.

So can we gauge the net effect of illegal migration on British GDP? Not with any precision, and anyone quoting a confident number is selling something. What can honestly be said is narrower. The contribution to total output is positive and modest, measured in tenths of a per cent. The effect on income per head is probably close to zero, with the costs falling on low paid workers in the same trades. The heaviest fiscal burden comes less from people living here unlawfully than from a system that keeps lawful claimants waiting, idle and housed at hotel prices.

In that Israeli building, the empty floor turned out to be the most precise measurement anyone had taken. Britain has rooms of its own: kitchens, care homes, building sites. They are full for now. Nobody has counted who is in them.