F(airly) T(endentious) reports
Most people will by now know about last week’s report from the government’s Migration Advisory Committee on the impact of EEA migration. It’s a long read, and published together with a number of specially-commissioned papers on different aspects of the issue that provide an evidence base of sorts for much of the MAC’s recommendations. They’re all available here.
One of these aspects, and a widely reported one, is the net fiscal contribution made by EEA migrants — the difference between what they pay in taxes and the cost of providing them with benefits and public services — based on the research paper for the MAC by Oxford Economics. Here’s Chris Giles in the FT reporting on that, noting that substantially more is paid in than taken out by EEA migrants to the tune of £4.7bn in 2016/17. Sounds great. Especially when the average EEA migrant adult is putting in so much more than the average adult in the UK (‘citizen’ is a misquote). Even those low-paid Eastern Europeans are putting in a grand more than the average! And richer, younger migrants without any kids are putting in loadsamoney. Unfortunately, to my mind some of this is fairly tendentious stuff.
Let’s start with the easy one….
The research showed that young employees earning £50,000, for example, paid £20,000 more in taxes than they received in state spending.
Hmmm. The research simply modelled the net fiscal contribution of some different kinds of household, regardless of where its members came from. For a 20 year-old singleton earning £50,000 a year this was £20,000. But so what?The research didn’t show that there were such 20-year old single Europeans in the UK (or how many) merely that if there were then they’d each contribute £20,000 a year. Just as any 20 year-old earning that much would, wherever they were born.
To my eye, the article is written as if reporting a finding that of the groups who are paying more than the average, an identifiable component was rich young migrants with few children, and that some of these were employees earning £50,000 a year. Such a component might exist, and might even have been identified in the researchers’ spreadsheets somewhere, but it’s certainly not in any of the reports. In fact, from published HMRC data, the 90th percentile point for EEA taxpayer income in 2015/16 was £47,600 (lower than the £53,100 for all taxpayers). To be earning £50,000, your ‘richer’ European migrant would have be in around the top 6% of European earners. What proportion of these might be 20 year-olds is anyone’s guess, and my guess is …. a small one. The earnings data used by the researchers came from the ONS UK Labour Force Survey, and so just as a rough cross-check I’ve looked at the most recent quarterly dataset for Jan-Mar 2018. For the five-year 20–24 age band (so not just 20 year-olds) not a single EU-born person earning £50,000 appears in the dataset. The closest is someone with an annualised income of £35,000. Obviously this is a survey, and if someone had appeared they would have been ‘weighted’ up to 700 or 800 actual people but the point is that if there are high-earning 20 year-old singletons among the 2.4 million EEA workers in the UK there are likely to be but a handful of them.
This might seem just an irrelevant digression because the researchers didn’t make any assumption about how many such young thrusters were here in coming to their conclusions, they just crunched the data on migrants that there were and reported the overall results. But because that’s all they did, and they didn’t report any compositional or distributional breakdowns, I find it odd that their modelling which is nothing to do with the data has been reported by the FT as if it were something derived from the data and to do with EEA migrants.
So back to the data and results.
Migrants paid substantially more in taxes to the government than they took in benefits or public services in 2016–17. The report estimated that European migrants made a total contribution of £4.7bn to the public finances in 2016–17.
£4.7bn isn’t nothing and as a lump of money it’s quite a lot. However, as the MAC laconically noted at para 4.11 of its own report …
This sounds a very large number but averaged across the adult UK-born population it amounts to £1.70 per week.
So from that point of view it isn’t quite so substantial. On the other hand it’s a bit more substantial than that as weirdly the MAC seem to have made a mistake and used the whole adult population in the UK as their denominator (try it yourself: £4.7bn divided by £1.70 divided by 52 weeks in a year gives 53.16m people). The UK-born adult population at the time was around 45m so the actual amount averaged out across the adult UK-born population was more like £2 a week.
It’s also the difference between two quite large numbers, which on the data published appear to be something like £47.2bn revenues and £42.5bn expenditures. The research paper doesn’t report these amounts directly (again a bit weirdly but you can infer them) and that’s a decent enough difference to be beyond margins for error, but the MAC also note that of the £4.7bn, £4.4bn comes from people from ‘Original Member States’ and only £0.3bn from people from ‘New Member States’. This means the former paying a good 20% more than they’re costing, but the latter paying barely 1% more.
This isn’t easy to pick up directly from the underlying report as there is a stacked chart of revenues on one page and a stacked chart of expenditures on a different page. In neither is the total stated. So not only is a immediate visual comparison not possible, but you have to add up the components of a stacked column in one chart and then in the other chart before you know what’s what.
Anyway, if you do that and make your own picture, this is what you get
The black crosses mark the overall impact. On these data, people from the New Member States (NMS) are barely paying their way, with a tiny difference between revenues and expenditures of £160 in the year per average adult. Bearing in mind the broad assumptions needed to get to these figures, while the situation might be more positive, it could also be negative.
But but …. doesn’t the FT report say that people from the New Member States are contributing over a thousand pounds a head? Look, it says ‘pay more in taxes than they receive’ and for ‘EU members from 2004 and after’ it shows more than a thousand pounds….
The original, as reproduced by the MAC, was slightly different. Turning the thing on its side doesn’t matter, but I don’t think that relegating the original title to a sub-heading and replacing it with ‘Migrants pay more than they receive’ as the title is really on. The measure of that for New Member States was, as above, £160 a head and not £1,040. This brings us to a fundamental question about the comparison being made. Who are these ‘average adults’ we’re talking about, wherever they are from?
Revenues and expenditures for the different population groupings have been added up and then divided by the number of adults in each group (assigning the costs of dependent children to their parents) to give average tax and spend per adult. These amounts are then described as contributions or costs of an ‘average adult’. I think it’s unhelpful and obscurantist to use the term ‘average adult’ for this purpose, or ‘accountable adult’ which Oxford Economics uses interchangeably. For example, State Pension costs for the UK-born population were averaged over the UK-born adult population at £2,340 per head. But as the majority of UK-born adults are below state pension age and don’t receive any state pension at all, it seems unhelpful (and again rather weird) to say that the average UK-born adult receives £2,340 state pension a year. It is obscurantist because the context for the whole thing is what level and kind of immigration is desirable, and in particular who should be allowed or encouraged to come to come into the UK for work and who not. Comparing the average fiscal outcomes from the whole UK-born population with the average fiscal outcomes of whole born-abroad populations does not say anything about either the relative or the absolute value of admitting someone from abroad today or tomorrow. If asked who the ‘average UK-born adult’ is, and what they are doing, the common-sense and accurate answer is that he or she is of working age and in work. That’s what the majority of adults are doing even if you include retired people. The ‘average EEA migrant’ is also of working age and in work, albeit on both measures to a higher degree. I think that the more meaningful comparison in all the circumstances would be between such people rather than between ‘adults assigned whole population averages for revenue and expenditure’.
Why? Well at the moment the UK economy is still, slightly, in deficit. This means that on a whole economy/whole population basis the ‘average’ adult is making a negative fiscal contribution. If making a contribution better than that whole population average were the right criterion for allowing entry for work, that would mean admitting even people who were going to make a negative contribution, so long as it was less negative than the existing average. But that would be silly, because anyone admitted on that basis would increase the deficit and worsen public finances.
What I think is more relevant to the question of the desirability of migration for its fiscal impact is who’s making what contribution and how this compares to their peers, .
For example, just looking at income tax, in 2015/16 (the most recent year for which real data from HMRC is available) the average amount of income tax paid by a non-pensioner taxpayer was £6,000 while the average for OMS taxpayers was nearly double that and the average for NMS taxpayers not much more than a third of that.
Those non-pensioner taxpayers are not making any less of a contribution either individually or on average or in total because part of their tax is being used to pay pensions to other people in the same population grouping. But the implication of ‘average adult’ is that they are. Basically, on all the available data people from NMS pay far less tax than their UK-born peers (understood as working-age adults). On the other hand their peers from OMS pay much more. This really is anodyne and it should be unexceptionable that people who do poorly-paid work in the UK don’t make much of a contribution but people who do highly-paid work do. By and large people from NMS do poorly-paid work and people from OMS do highly-paid work. Of course some NMS workers will be in highly-paid jobs and some OMS workers will be in poorly-paid jobs, but this is exactly the point: it doesn’t matter where you come from but what you do when you’re here.
As the average contributions of NMS adults are only 1% more than expenditures on them, and with an average net contribution of £160, then assuming a normal distribution of net contributions, about half will be contributing and half not. For every 20-something earning £50k a year and contributing £20k, then there will be someone else costing £19,840. From a merely fiscal point of view it is clearly more rational to allow in the former and exclude the latter for a net gain of £20,000 than to allow in both and see the net gain reduced to only £160. Or if the distribution is skewed, for every 20-something earning £50k a year and contributing £20k, the same principle would apply to the exclusion of ten people costing nearly £2,000 each, or twenty costing nearly £1,000 each.
That I guess is why the MAC, taking fiscal contribution as something that matters, has recommended not letting in people for work unless, by and large, the job they’re coming to do will pay £30,000 or more. One might call this the ‘Ac-Cent-Tchu-Ate the Positive’ approach.
This approach can be criticised for a number of reasons, and Martin Sandbu has a couple in yesterday’s FT. While acknowledging that stopping low-paid workers ‘sounds commonsensical’ Martin points out first
But economic activity pays a return to capital, not just to labour, and the earnings on UK investments in sectors staffed with low-skilled migrant labour should at least figure in an economic assessment of immigration rules.
Clearly businesses may benefit in profit terms from finding plenty of migrants willing to work for low wages. But the assessment does take this into account (to some extent at least) because Oxford Economics treat corporate taxes that are paid by business essentially as incident on workers rather than on shareholders so their indirect contribution to the fiscal balance through any general pushing-up of company profitability is recognised. If they hadn’t done this then the net NMS fiscal contribution would have been decidedly negative, but that’s a matter for another day.
Martin might be alluding instead (or additionally) to migrant workers ‘growing the size of the economy’. George Osborne was very keen on assessing the economy by mere size, but history might judge George quite harshly for his focus on the size of the pie rather than the slices that individuals get, what he did with the fiscal contribution that all those new migrant workers made, austerity and all the rest of it.
Martin then says, more generally,
I think this a bit rough on the economists on the committee. Of course it needs a willing buyer, whether of labour or goods, for an import to take place. Not exactly of course, as someone could just send stuff to the UK in the hope of finding a buyer but if it ends up languishing in a warehouse they’re not going to send any more. Similarly if people come from another country looking for work and can’t find anything then they’re likely to go home and tell their friends not to bother either. So the point seems to be that just as it’s accepted that goods should be allowed into the UK if there’s a willing buyer for them without any ‘central decision’ on quantity, so labour should be allowed in if there is a willing hirer.
However, in reality there are constraints placed on trade in goods between foreign sellers and willing UK importers. I might want to buy stainless steel butt-welding fittings (seriously) at the bargain price offered by manufacturers in China or Taiwan but anti-dumping duties are imposed by central decision of the EU to dissuade me from doing so or to pay a price if I really insist on doing so. That’s because there are considerations taken into account beyond the narrow economic interest of the person who wants to buy the goods, with the potential impact of the bargain he’s getting on domestic producers and their workers being set in the balance against it. That this can be a problem with free trade is clear from the reams of text in every free trade agreement to make sure that ‘free’ remains ‘fair’. But the means by which this is done is a series of ‘qualitative’ decisions not decisions on quantity.
Nor are these decisions simply about international competitiveness. They might also be based on more amorphous social choices. I might want the most powerful vacuum cleaner that money can buy (as I’ve a lot of dog hairs to deal with) and not care about the extra electricity cost. There are foreign sellers able and willing to supply me with one and if they’re in South Korea or Canada there aren’t any tariff barriers cos free trade agreements with those countries. But I’m not allowed to import one because of the contribution my extra power consumption would make to the ultimately socialised cost of extra CO2 emissions. The ‘central decision’ on imported goods is again not quantitative, but qualitative.
Qualitative central decisions determine the possible mix of imports. These constraints set on mix can affect the level. They won’t do so necessarily, because manufacturers abroad can adjust their power outputs and UK importers might be prepared to pay anti-dumping duties, so the flow of vacuum cleaners and butt-welding parts continues at the same level. But the potentially adverse impacts of levels of inflows have been mitigated by changes in the mix determined by central decisions.
In contrast, where an employer’s narrow interest might be served by hiring migrant workers on minimum-waged or ‘self-employed’ contracts, this might impose socialised costs on others (taxpayer-funded in-work benefits, wage’n’housing squeezes etc). Changing the mix of migration through things like qualification and salary level requirements, which are what the MAC recommends, does not result from decisions on quantity. Leaving aside the general unattractiveness of comparing people with goods, the MAC seems to be recommending essentially qualitative restrictions on people of a kind analogous to those routinely and unexceptionably made in relation to trade (even free trade) in goods.
Whether this is desirable or not is a reasonable matter for debate, and coming down on one side doesn’t make for mere dilettantism.
None of this is intended as a criticism of either Chris or Martin, and they certainly aren’t dilettantes themselves!
