Migration Statistics

Charles Walker Excerpts

Thursday 26th June 2014 Westminster Hall

[Mr Jenkin]

I do share that concern, but if the right hon. Gentleman will forgive me, I will deal with that issue later.

I was talking about the 95% confidence interval in respect of smaller samples relating to individual countries. The ONS will publish estimates of immigrants by country only for the top 15 source countries, because for all the other countries the sample is too small to provide a meaningful estimate—in other words, the number of people from Iran or Afghanistan is actually smaller than the 95% confidence interval itself, so the number is meaningless.

We have vague estimates of the numbers coming in from China, India, Poland, the USA, Australia, Spain, Pakistan, France, Germany, Italy, Canada, Nigeria, New Zealand, Lithuania and Hong Kong. Those are the countries for which figures are published. For the other 180 or so countries, no figures are published, so we cannot tell from the data how many Russians, Iranians, South Africans or Romanians are coming to this country.

For the same reason, the ONS migration data cannot provide anything meaningful for local authorities that are trying to work out how migration flows affect their area or to plan for population changes. The UK Statistics Authority also stated:

“The IPS sample size is too small to enable the production of reliable international migration estimates at a local authority level, and cannot realistically be made sufficiently large to achieve robust local estimates.”

The census, which is designed to count every member of the population, provides the only reliable data on the number and characteristics of migrants at local level, but we get it only every 10 years, which is why it was so full of surprises.

In evidence to us, Westminster city council said that the current methodology for estimating migration was not robust enough to support accurate local-level estimates, so that

“the measurement of migration from the perspective of an LA user and as reliable information on our residents is failing”.

The leader of Westminster more or less told us that the only way it can find out the nationalities of the people in the borough is to go around and count them itself. That may be a responsibility that it should take on, but—[Interruption.]

[Mr Charles Walker (in the Chair)]

Order. I say to the official who just approached the Minister: please do not do that again. This is a Chamber.

[Mr Jenkin]

That, perhaps, is one of the shortcomings of Westminster Hall, Mr Walker.

The question is how this situation could be improved. We suggested, and I suggested to the Prime Minister when he came before the Liaison Committee, that we should expand the size of the international passenger survey and therefore increase the size of the migrant sample on which the estimates are based. We were advised that if we spent an extra £15 million on the IPS, that would quadruple its size. That would halve the size of the confidence interval, meaning that there would be a 95% chance that the data were within 17,500 of the estimate, rather than there being a total margin of error, on a 95% confidence interval, of 35,000. That brings the range down, but it is a lot of money for not much improved accuracy and it still helps us only with the headline figures. It does not help us with the quality of the data for smaller groups of migrants or for local areas.

The ONS could see what extra value it could derive from the IPS by, for example, asking respondents for various details, notably passport numbers but also national insurance or NHS numbers, which would allow responses to be linked to administrative data, but that would still not address the fundamental problem of the small sample size.

Alternatively, there could be a survey more specifically targeted at migration. A large-scale face-to-face survey of migrants in the UK has previously been considered, leading to a feasibility report published by the Home Office in January 2011. I ask my hon. Friend the Minister to give that further consideration. The Government concluded against funding such a migrant survey after it was estimated that it would cost a mere £2 million, based on the survey design envisaged. Unlike the Government, I think that that would be good value for money, and that option was recommended by the Office for National Statistics. I hope very much that the Minister will deal with that in his closing remarks.

A migrant survey could provide valuable information on the characteristics and distribution of migrants. That would increase the reliability of immigration estimates in relation to smaller geographical areas and be of some help to local authorities such as Westminster, which at the moment are reduced to doing surveys of their own.

In the longer term, as the Chairman of the Home Affairs Committee the right hon. Member for Leicester East (Keith Vaz) and my hon. Friend the Member for Ribble Valley (Mr Evans) said, the only sure way to improve migration data is to use the e-Borders information. That comes from the advance passenger information, or API, which airlines and other carriers provide to the authorities whenever there is an incoming aircraft or ship.

“Using e-Borders data in the production of long-term and short-term migration counts would be a ground-breaking improvement that would offer several advantages over the migration estimates produced solely from the International Passenger Survey”.

Those are not my words but those of the ONS.

The ONS and the Home Office should move as quickly as possible towards measuring immigration, emigration and net migration using e-Borders data, so that at least a significant proportion of people can be counted in and out of the country as they enter and leave. The e-Borders scheme has now been replaced with the border systems programme, but it should still be possible to use it to count people in and out of the country. Those administrative data would give information about cross-border movements different from that provided by the IPS, but they would still not be without faults. In many respects, the data would give a deeper understanding of the comings and goings from our country. In their response to our report, the Government said that the data gathered through the border systems programme

“does not hold the information to directly estimate net migration”

and that:

“The Border Systems Programme is not designed to provide direct statistical measurement of migration flows”.

My understanding is that that represents a significant downgrading of the Government’s original ambitions for the programme and a failure to deliver what was originally envisaged. The Government’s original business case for the e-Borders scheme said that it would provide

“the ability, for the first time, to comprehensively count all foreign national passengers in and out of the UK, improving public confidence in the integrity of the border and enabling a more accurate count of migrants for future planning and for informing the population count.”

Of course, not everyone entering or leaving the UK is migrating, but if people are on a visa, it should be possible to measure when they enter and when they leave the country. Passport checks are all about checking whether people have a valid visa and whether they are on a watch list. Currently, although 80% or 90% of visas are scanned on entry or exit, we are told that those data are not used for counting in and counting out visa nationals. Why not?

I think that most of the British travelling public would be astonished to find out that passports are scanned but not even people who are on a visa are recorded as they pass into or out of the country. The Home Office should move as rapidly as possible towards integrating visa information with border systems programme data, so that an accurate measurement can be made of immigration, emigration and net migration by people in different visa categories. That would also provide data on the number of people in different visa categories currently living in the UK, and it would enable the Home Office to gather detailed information on the characteristics of migrants who are subject to migration control.

As things stand, we simply do not know how many visa nationals are currently in the country; we do not know how many comply with the rules and how many overstay; and we do not know how many of the people migrating to and from the UK on a long-term basis entered the country in each visa category. That makes it hard to work out whether changes in visa policy are having the intended effect on migration flows and almost impossible to establish the scale of the problem of people who stay here illegally. There is no reason for the situation to persist now that the Government have committed to reintroducing exit checks, but in their response to the Public Administration Committee the Government made no commitment to track the entries and exits of visa holders once that becomes possible, even though it is fully within their power to do so. They say only that that

“may be feasible in future”.

We believe, however, that it should be done as a matter of urgency.

To be clear, we have not recommended that the Government should stop using the IPS by any means, but the Public Administration Committee recommended that the Government plan to end their reliance on that survey as the sole basis for estimating migration flows. The IPS was not designed for the important job that it now has. It was never intended to be used for the purpose of estimating international migration; it was designed to support the work of the then British Tourist Authority by providing economic data on travel and tourism.

The next five years will see much work in Government on developing new data sources that will eventually replace the decennial population census. It is vital that work on immigration be fully co-ordinated and that Departments share intelligence. That our official immigration and emigration estimates do not match our official net migration figure for a whole decade underlines the Committee’s main finding that the current system of relying solely on the IPS for migration statistics is not fit for purpose. Although the IPS provides useful information about the characteristics of migrants, it cannot be relied on to give us accurate numbers of those migrating into and out of the United Kingdom.

There is no reason why the Government cannot use border systems programme data dramatically to improve the accuracy of migration data. The Home Office told us:

“There will be some possibility to link e-Borders data in the future, in due course”,

but we have not yet received any clear commitment that that will happen, let alone a time scale. That is not adequate. The issue requires urgent action. Estimates based on a survey alone are no longer fit for purpose. Instead, we need to make proper use of the electronic data from the border systems programme. The public need and deserve to be given accurate information about migration to the UK, using the latest technology and methods available.

We are now in an election year, during which the issue of immigration will be hotly contested, but that debate is likely to do no more than produce despair in the minds of our voters. The politicians of the main parties are arguing about policies, the effects of which they cannot measure, in relation to numbers of migrants that they cannot determine. That can only undermine trust and confidence in political life, and it will provide an avenue for extremist parties to exploit at the expense of the proper government of this country. We owe it to our voters to deliver more accurate migration statistics as soon as possible.