Understanding migration statistics

Discrepancies between National Insurance Numbers and International Passenger Service Statistics

At a time of heightened rhetoric surrounding migration, any suggestion that numbers are higher than officially stated inevitably attracts attention.  So the apparent discrepancies between the number of people, particularly from Eastern Europe who have obtained a National Insurance Number and the numbers estimated in the quarterly migration statistics stimulated both headlines and hand-rubbing by opponents of migration.  The Office of National Statistics was potentially red-faced and the refusal last year by the Government to grant a freedom of information request on this matter added to the speculation.

The potential discrepancy is significantly large – on the one hand, in the year to September 2015, 861,936 NI numbers were issued, of which more than three-quarters went to EU nationals. The official immigration statistics, on the other hand, show that 617,000 people entered the UK in the same period, including 257,000 EU nationals.

This was potentially explosive and the chairman of the UK Statistics Authority, Sir Andrew Dilnot undertook to review and potentially strip the quality kitemark awarded to migration statistics. 

The outcome of the inquiry was announced on a day of busy news as David Cameron hosted his anti-corruption policy and the Queen was caught in a diplomatic faux pas.  This naturally fuelled expectations of a cover up. However, the much anticipated explanation turned out to be a damp squib as the two sets of data do not speak to each other so simplistically.  The professional statisticians were satisfied that short-term migration accounts for most of the differences. Further investigation of range of data sources continues. Nevertheless Migration Watch, on their website still maintain that there are potentially an additional 50,000 Eastern European migrants to those counted in the migration statistics.


Migration Matters’ Gill Green, a former senior researcher at the Audit Commission comments –

"NiNO numbers and migration statistics measure different things, so it isn’t surprising that they don’t give the same results.

"Long term migration measures people intending to stay or leave for over 12 months. That is the international definition that is used to allow comparisons between countries.

"People apply for NiNOs so that they can work or claim certain benefits.  Many EU workers are not are long-term migrants. There are many shorter-term migrants, e.g. workers in seasonal tourism or agriculture. The ONS collect short-term migration statistics for visits between 1 and 12 months. (And some visitors coming fro a shorter period may also apply for a NiNO.)

"However, you don’t need a NiNO to enter the country.  The NiNOs issued in any one year include some people already counted in migration statistics from previous years - often migrants who had not previously been working, such as spouses or children.  The ONS have looked at other data from tax and benefit records. They found that 17 to 25% of EU applicants for NiNOs between 2010 and 2013 had arrived in the country a year or more before applying for that NiNO.  So you cant assume that a new NiNO means a new migrant.

"This tax and benefit data confirms that many of the NiNO applicants are/were short term visitors; for one third of the EU migrants, their tax, benefit and national insurance records combined cover less than 52 weeks. 

"So what does this all mean for arguments over migrant numbers?

"The ONS don’t always get the numbers right – they had to revise earlier migration estimates upwards in 2015 following detailed work on census returns.  But it does look as if short-term migration is the most important factor in explaining the difference between long term migration figures and annual new National Insurance numbers."