Transforming population and migration statistics: NINo and NHS registration lags

Key messages
• EU nationals register more quickly for a National Insurance number (NINo) than non-EU nationals
• Non-EU nationals register more quickly with the NHS than EU nationals
Transforming Population and Migration Statistics
Case Study: NINo and NHS registration lags
Centre for International Migration
Published: 30th January 2019
Coverage: England and Wales (Personal Demographic Service), UK (Migrant Worker Scan)
Disclaimer: These Research Outputs refer to registration lags
produced using Migrant Worker Scan and Personal Demographic
Service data, and are not Official Statistics. These outputs must not
be interpreted as an indicator of registration lags, or migrant estimates
from the different administrative data sources.
What can linking Migrant Worker Scan and Personal Demographic Service data together tell us about
when and how new international migrants appear on different administrative data sources?
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Within 1
month
1-3
months
3-6
months
6-12
months
1-2 years 2-5 years >5 years
Millions Lag between arrival and NINo registration
EU Non-EU
0
10
20
30
40
50
60
Within 1
month
1-3
months
3-6
months
6-12
months
1-2 years 2-5 years > 5 years
Thousands Lag between arrival and NHS registration
EU Non-EU
Source: ONS analysis of DWP/HMRC Migrant Worker Scan data covering NINo
registrations between 2002-2017
Source: ONS analysis of NHS Digital Personal Demographic Service and
DWP/HMRC Migrant Worker Scan data covering type 4 PDS registrations
between 1st August 2016 – 7th August 2017

Background
Aim: This case study aims to build our understanding of when and how new
international migrants appear in different data sources, to feed into the work on
using these sources for international migration estimates
Data sources 1. Migrant Worker Scan (MWS), Department for Work and Pensions
(DWP) and HM Revenue & Customs (HMRC)
2. Personal Demographic Service (PDS), NHS Digital
Time period 2016-2017
Population coverage MWS: Overseas nationals who have registered for a NINo since 2002
PDS: People of all ages registered with the NHS during 2016-17
Further background information Previous research (Slide 11), Things you need to know (Slides 12-15)
PDS overview
The PDS is the master demographic database for the NHS
in England, Wales and the Isle of Man. It is the primary
source of information on a patient’s NHS number, name,
address and date of birth. Records are created for
newborns, or when a new patient makes contact with an
NHS service, primarily by registering with a GP or through
accessing A&E or attending hospital.
When someone from overseas makes contact with the
NHS for the first time, they will be given an NHS number. A
new record is created on the PDS and they should be given
a type 4 registration flag to indicate that it is a new
registration of someone from overseas.
MWS overview
The MWS contains information on all adult overseas nationals
who have registered for and been allocated a National Insurance
number (NINo). A NINo is generally required by any overseas
national looking to work or claim benefits / tax credits in the UK.
In order to apply for a NINo, in the first instance, a migrant
worker makes an enquiry to a Job Centre. They must attend an
"Evidence of Identity interview" at a local DWP Jobcentre Plus
office, where they must be able to prove that they are who they
say they are and that they satisfy the criteria for needing a NINo.
The MWS data contain useful information on arrival date and
nationality which are not available in other administrative
sources, such as the PDS.
Both data sources include both long-term and short-term migrants.

EU nationals register more quickly for a NINo than non-EU nationals
If administrative data sources are to be used to measure international migration, it is important to understand the lags between
migrants arriving in the UK and appearing in the administrative sources. The MWS contains arrival date and NINo registration
date, so it is possible to look at the time lag between arrival in the UK and registering for a NINo.
Out of all records in the MWS, 84% had registered for a
National Insurance number within a year. The median time lag
between arrival and registration for a NINo was 72 days for
EU nationals and 135 days for non-EU nationals.
The median lag has decreased over time, from 292 days for
individuals who registered in 2002, to 37 days for individuals
who registered in 2017.
For 2017 registrations, the median lags were 32 days for EU
nationals and 71 days for non-EU nationals.
One potential reason for quicker
registration among EU nationals is
that work is the main reason for their
migration, whereas non-EU nationals
are more likely to migrate for study or
family reasons.
Things to be aware of
0.7% of individuals had a registration
date before their arrival date. As
arrival date is self-reported, it is likely
that this arrival date is a date of
return following earlier residence in
the UK, rather than the date of first
arrival.
0
50
100
150
200
250
300
350
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Days
Year of NINo registration
Median lag between arrival and NINo registration
EU Non-EU
Source: ONS analysis of DWP/HMRC Migrant Worker Scan data
0
1
2
3
4
Within 1
month
1-3
months
3-6
months
6-12
months
1-2 years 2-5 years >5 years
Millions Lag between arrival and NINo registration

Non-EU nationals in MWS more likely to link to PDS than EU nationals
Of the 10 million records on the MWS², 62% were linked to the PDS and 38% were not. Some of the unlinked records will be
due to the fact that the PDS only covers England, Wales and the Isle of Man, whereas the MWS covers the UK.
Linkage rates were higher among non-EU nationals than EU nationals. Some potential explanations for this are:
• Most non-EEA nationals now have to pay a health surcharge before they move to the UK if they are planning to stay for
more than six months. They perhaps then have more incentive to register for NHS services. EEA nationals do not have to
pay the health surcharge, so may not register with the NHS until they need to use it.
• Short-term migration is more common among EU nationals than non-EU nationals. If an individual is here for only a few
months, they are less likely to need to use NHS services, but are likely to need a NINo to work.
• A large proportion of non-EU nationals are students, who are encouraged by their university to register with a GP.
This analysis shows the importance of using multiple data sources to measure international migration. If only the PDS was
used, a number of potential migrants could be missed.
¹More information is available on Slide 12, Things you need to know
²The ONS version of the MWS only contains individuals who registered for a NINo from 2002 onwards
Looking at the MWS in combination with the PDS can provide new insights into international migrant interactions with
administrative systems and the footprints that migrants leave in administrative data
Source: ONS analysis of DWP/HMRC Migrant Worker Scan and NHS Digital Personal Demographic
Service data
Methodology
The linkage between the MWS
and PDS was done in two steps:
1. Link the PDS to the DWP
Customer Information System
data using the ONS
matchkeys methodology¹.
2. Link to the MWS using the
National Insurance number
This two-step process was used
because the MWS is not updated
with address or name changes
and these variables are needed
for the linkage.
0
20
40
60
80
100
EU Non-EU
%
Linked to PDS Not linked to PDS
Linked to
PDS
Not linked
to PDS

Of those linked to a NINo using the Customer Information System data (CIS), but not recorded
in MWS, two-thirds were children, who we would not expect to find in MWS. The remaining may
be examples of incorrect type 4 flag allocations; for example, those given to returning UK
nationals. These potential incorrect flag 4 allocations represent around 3% of all type 4 flag
allocations in the target period.
Of those not linked to a NINo, 24% were under 15, so we would not expect them to be in MWS.
52% were aged 15-29, some of which may be students who are not working during their
studies.
Methodology
A cohort of individuals was selected who registered with the NHS between 1st August 2016 and 7th August 2017 and received a
type 4 flag upon registration. The type 4 registration indicates that they are a potential migrant.
Of the 710,000 individuals
in this cohort:
• 50% were linked to MWS
• 9% were linked to a NINo
(in CIS) but not to MWS
• 41% were not linked to a
NINo
This analysis
shows the
importance of
using PDS in
addition to
MWS to
measure
international
migration,
particularly
for those in
the younger
age groups.
Important to use PDS alongside MWS to measure migration
The PDS contains a variable that indicates the type of registration. If an individual has a type 4 flag, this indicates that they are a
new registration from overseas. By linking PDS to MWS, we can compare the use of type 4 flags in PDS with registration for a
NINo in MWS (which is another indicator that someone is a potential migrant).
Source: ONS analysis of NHS
Digital Personal Demographic
Service and DWP/HMRC Migrant
Worker Scan data
Linked to MWS
Linked to NINo but not MWS
Not linked to a NINo
0
20
40
60
80
100
120
140
160
0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65+
Thousands
Age group
Linked to MWS
Linked to NINo but not MWS
Not linked to NINo

Non-EU nationals register more quickly for the NHS than EU nationals
For those in the PDS cohort who have been linked to MWS, it is possible to look at the time lag between the arrival date stated
in MWS and the NHS registration date stated in PDS.
• This analysis excludes individuals who have not registered for a NINo. If an individual has not registered (yet) for a NINo,
we will not have their arrival date so do not know the lag. Therefore, the median lags are only valid for the individuals in the
linked cohort and do not represent median lags for the whole migrant population in the PDS.
• 0.5% of individuals in the cohort had an arrival date in the MWS that was after their NHS registration date in the PDS. This
suggests that the MWS arrival date is a date of return, rather than the date of first arrival.
Non-EU nationals tend to register more quickly with the NHS than EU nationals.
For those in the cohort, the median lags were:
• EU: 276 days
• Non-EU: 60 days
82% of non-EU nationals in the cohort had arrived within the last year whereas
57% of EU nationals had.
Nationality Median lag
(days)
Malaysia 20
Yemen 24
Congo 25
Malawi 26
Bangladesh 29
Nationality Median lag
(days)
Latvia 581
Lithuania 579
Poland 503
Slovakia 496
Hungary 392
As previously discussed, the shorter
registration lags for non-EU nationals may
be linked to the fact that most non-EEA
nationals have to pay a health surcharge
before they come to the UK, so perhaps
have more incentive to register with a GP
when they arrive. EEA nationals may just
register when they need to see a doctor.
Top 5 nationalities with the
longest lag
Top 5 nationalities with the
shortest lag
Source: ONS analysis of NHS Digital Personal Demographic Service and DWP/HMRC Migrant Worker Scan data
0
10
20
30
40
50
60
Within 1
month
1-3 months3-6 months 6-12
months
1-2 years 2-5 years > 5 years
Thousands Lag between arrival and NHS registration
EU Non-EU

Females register more quickly for the NHS than males
• This analysis excludes individuals who have not registered for a NINo. If an individual has not registered (yet) for a NINo, we
will not have their arrival date so do not know the lag. Therefore, the median lags are only valid for the individuals in the
linked cohort and do not represent median lags for the whole migrant population in the PDS.
• 0.5% of individuals in the cohort had an arrival date in the MWS that was after their NHS registration date in the PDS. This
suggests that the MWS arrival date is a date of return, rather than the date of first arrival.
• The chart does not include individuals aged 65+ due to low numbers in the cohort in this age group.
Females tend to register more quickly with the
NHS than males, with this trend seen in all except
the youngest age group.
For those in the cohort, the median lags were:
• Males: 291 days
• Females: 150 days
70% of females in the cohort had arrived within
the last year whereas 55% of males had.
Source: ONS analysis of NHS Digital Personal Demographic Service and
DWP/HMRC Migrant Worker Scan data
0
50
100
150
200
250
300
350
400
15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64
Days Median lag between arrival and NHS registration
Male Female
The NHS registration analysis can also be broken down by age and sex.
There are also differences across the age groups. This can be seen for males in particular where the median registration lag in
the cohort ranged from 61 days for 15-19 year olds, to just over a year (371 days) for 25-29 year olds. The short registration
lags for 15-19 year olds may reflect children being registered by their parents, or first year students being encouraged to
register by their university.
As 55% of non-EU nationals in the cohort are
female compared with 49% of EU nationals,
these differences by sex may partially explain the
shorter registration lags for non-EU nationals
shown on the previous slide.

PDS consistent with MWS for measuring international migration
Registration type 4s were also investigated by looking at whether individuals in the MWS obtained a type 4 flag upon first
registration with the NHS. For this analysis, a cohort of individuals was selected that had an arrival date in the MWS of between
1st August 2016 and 31st January 2017. There were 378,000 individuals in the MWS that had an arrival date in this period and
153,000 linked to the PDS. 138,000 of these had a registration date and were included in the analysis.
1) There will be individuals who arrived in the target period but have not yet
registered for a NINo. This means that we do not know that they have arrived
so they cannot be selected for the cohort. Some of these individuals may have
registered with the NHS but we do not know that they are in our cohort, as
arrival date is not available in the PDS.
2) Missing registration type information mainly relates to older PDS records
created prior to 2014. Pre-2014, the PDS system was maintained outside of
NHS Digital and the registration type variable was not used.
3) Of those with a registration type of other than 4, 38% had a PDS registration
date before their MWS arrival date. This suggests that their MWS arrival date
may not reflect their first arrival in the UK. These individuals may have been in
the country for some time before registering for a NINo and, therefore, we may
not be capturing their first registration with the NHS in this analysis. If this is
not their first registration, they will not receive a type 4 registration flag.
4) As the linkage was done using the 2017 CIS and PDS datasets, which have
extraction dates of July/August 2017, the individuals in the cohort have only
had 6-12 months to register with the NHS. This explains the lower linkage
rates for this cohort when compared with the MWS as a whole. The reasons
behind selection of this cohort can be found on slide 15.
Of all individuals in the cohort who linked to PDS, 91% received a type 4 flag. When
those with a missing registration type are excluded, this figure is 95%. This
suggests that the MWS and PDS are fairly consistent for measuring international
migration.
Type 4: 1st acceptance – from
overseas
Other: 1st acceptance not from
overseas or 2nd acceptance
Missing: No registration type for the
individual
Source: ONS analysis of NHS Digital Personal Demographic Service and DWP/HMRC Migrant Worker Scan data
PDS registration type
Type 4 Other Missing

It provides evidence for:
• Development of a Statistical Population Dataset (SPD) from
linked administrative data
Analysis shows that the inclusion of both Migrant Worker Scan
and Personal Demographic Service data is necessary to ensure
that the SPD covers all international migration and that linkage of
multiple datasets is required to develop a full understanding of the
population.
• Development of rules for stocks and flows approaches
Identifies the need to assess and take into account differences in
registration lags when applying rules to both develop the SPD
and in the application of methods to measure both stocks and
flows. This relates to both registration lag differences between
datasets and registration lag differences between groups with
different characteristics.
Next steps
Further work to use this linked dataset
to assess the quality of local level
address information.
Putting administrative data at the core of our evidence on international migration (UK)
and on population (England and Wales) in 2020
This analysis also helps us to understand how different types of
international migrants interact with services. For example, those
registering for a NINo aren’t accessing health services
immediately.
The findings from this case study provide important insights that
will be key to the successful development of a population and
migration statistics system based on administrative data sources.

Next Steps
Further research using MWS and PDS to help understand migrant destinations:
• Compare address information in the MWS and PDS for linked records to better understand the quality
of address information and how quickly the MWS address information becomes out of date (as
address is not changed after first registration).
• Investigate frequency and timing of address moves in PDS for new international migrants, which is
important for the internal migration component of population estimates.
Research from linking MWS/PDS to other datasets:
• Link MWS to Exit Checks data to better understand the accuracy of the arrival date variable in MWS,
for non-EU nationals.
• Link PDS to Exit Checks data to identify, for non-EU nationals:
• a first arrival date for individuals not in MWS, so that registration lags can be calculated.
• the quality of PDS data for estimating population stocks.
• For individuals in MWS not linked to PDS, link to benefits, PAYE and self-assessment data to look for
other signs of activity.
Further collaboration with DWP on using MWS data for measuring international migration.

Previous research and statistics
DWP statistics on National Insurance number allocations to adult overseas nationals entering the UK
The latest statistics showed that:
• 641,000 new NINos were registered in the year to June 2018, a decrease of 17% on the previous year
• 70% of these were from within the EU
DWP statistics on National Insurance numbers allocations to adult overseas nationals entering the UK – registrations
to March 2014
This release included a section focusing on EU2 nationals and included some analysis of registration lags:
• Restrictions on the employment of Romanian and Bulgarian migrants in the UK were lifted on 1st January 2014
• In 2013/14, there were 47,000 NINo registrations from Romanian nationals and 18,000 from Bulgarian nationals.
These were 163% and 71% higher respectively than in the previous year.
• Of those that registered between 1st January 2014 and 31st March 2014:
• Over 30% of EU2 nationals had arrived in the UK over a year prior to registration, compared with around 4%
for Polish and Spanish nationals
• 22% of EU2 nationals registered for a NINo within 3 months of arrival, compared with approximately 70% of
Polish and Spanish nationals.
Local Area Migration Indicators
The ONS local area migration indicators contain data on the number of records with a type 4 flag in each annual NHS
Patient Register stock dataset. More information on the Patient Register data is available here.
MAC report on international students in the UK
• International students can access public healthcare while studying in the UK; EEA students can use their
European Health Insurance Card and most international students from non-EEA countries must pay a health
surcharge of £150 per year as part of their visa application.
• Most universities have recommended GP practices near campus which they encourage students to register with.
Please note that our exploratory analysis is not directly comparable with these statistics.

Things you need to know
• Disclaimer: These Research Outputs refer to registration lags produced using Migrant Worker
Scan and Personal Demographic Service data, and are not Official Statistics.
• Rather, they are published as outputs from exploratory research to show users of international migration
statistics where we are on our transformation journey to put administrative data at the core of our evidence
on international migration (UK) and on population (England and Wales) in 2020.
• These research outputs must not be interpreted as an indicator of registration lags, or migrant
estimates from the different administrative data sources.
• These research outputs are based on experimental analysis of linked MWS and PDS data. We are still
developing methods; therefore, the numbers or proportions reported here may change in future reporting
on our journey to put administrative data at the core of population and migration statistics.
• It is important that the information and research presented here be read alongside the population and
migration statistics transformation update report to aid interpretation and avoid misunderstanding.
• These research outputs must not be reproduced without this disclaimer and warning note.

Averages
Average lags are based on the median, rather than the mean, as lags are not normally distributed. Lags are
skewed towards shorter lags, so if the mean was used, the smaller number of long lags would have a
disproportionate impact on the average. The median is therefore more representative of the data.
Nationality
Nationality has been classified using the International Passenger Survey (IPS) country groupings found in the
International Migration - Table of Contents (citizenship/nationality groupings tab).
The EU nationality group excludes British nationals.
Nationality is grouped into EU and non-EU instead of EEA and non-EEA (which is generally more related to
immigration control and policy, such as the immigration health surcharge) to be consistent with previous
research. For this reason the following three EEA countries will appear in the non-EU nationality group:
o Iceland
o Norway
o Liechtenstein
Additionally, Swiss nationals are included in the non-EU nationality group. Switzerland is neither part of the
EU nor the EEA but is part of the single market, so Swiss nationals have the same rights to live and work in
the UK as other EEA nationals.
Data linkage
The data linkage was conducted using the ONS matchkeys methodology. Matchkeys are created by putting
together pieces of information to create unique keys that can be hashed and used for automated matching.
For example, a matchkey might be constructed from an individual’s forename and surname, combined with
their date of birth, sex and postcode district. The resulting string is then ‘hashed’ and can be used to link
records between datasets in the anonymous data research environment.

Migrant Worker Scan (MWS)
The geographic coverage of the MWS is the UK.
The MWS has been used as the source of nationality information. It should be noted that nationality is based
upon nationality at the time the individual first registered for a NINo. There are no records of individuals
holding more than one nationality, nor when a change in nationality occurs after registration. Therefore,
naturalisation to UK citizenship is not recorded.
Arrival date in MWS has been taken as the individual’s first arrival in the UK. However, the arrival date is later
than the NINo registration date and/or PDS registration date in some cases, which suggests that the arrival
date may reflect the individual’s most recent, rather than first, arrival in the UK.
The MWS data extract includes any NINo allocations to overseas nationals since 2002. This means that
foreign nationals who registered for a NINo before 2002 are not captured in the data.
A Temporary Reference Number can be allocated while the NINo application is in progress. Temporary
Reference Numbers are not included in the MWS data.
For more information on NINo allocations to adult overseas nationals, see Background Information produced
by DWP.
Further information on the NINo application process is available here.
Customer Information System (CIS)
The CIS was used in the data linkage process. It is a dataset that contains a record for all individuals who
have registered and been issued with a National Insurance number in the UK.

Personal Demographic Service (PDS)
The geographic coverage of the PDS is England and Wales.
Records on the PDS are updated at GP surgeries, pharmacies, drop-in-clinics, A&E, hospitals and Primary
Care Trusts. The PDS relies on people interacting with the NHS to update their health records when entering
or leaving England and Wales, or when they change their address.
The analysis of the PDS data used a combination of the 2016 and 2017 PDS stock files and the 2016-17 PDS
movers files. The movers files contain updates to the PDS between stock extracts, such as new registrations
or address changes. As an individual could register and then move, all between stock extracts, the movers
files needed to be used so that their first registration could be captured. For example, if an individual
registered from overseas in November 2016 and then moved to a different part of the country in January
2017, they would be given a type 4 registration flag at that first registration and then get a type 3 flag after they
moved. They would not be in the 2016 extract as they registered after the data extract was taken, and in the
2017 extract it would have a registration date of January 2017 and a registration type of 3. Therefore, without
using the movers files, we would not know the date of first registration and would not know that they had come
from overseas. The movers files are only available from the 2016 stock onwards, which restricted the cohorts
that could be selected for analysis.
The registration date in PDS is the date of registration with a particular NHS system, rather than the date that
the individual first obtained an NHS number. The earliest available registration date has been taken but in
some cases this may not reflect the first registration.
Any missing values in the registration type variable may be due to the records being in the PDS prior to 2014.
Pre-2014, the PDS system was maintained outside of NHS Digital and the registration type variable was not
used.

Contact Us
We would welcome your feedback on the exploratory research
presented here. Please get in touch at:
Email: pop.info@ons.gov.uk
Tel: 01329 444661

Further Links
Go back to: An update on our population and migration statistics transformation journey: A research
engagement report
Explore our other case studies on SlideShare
Transforming population and migration statistics: Administrative data-based population stocks and
flows
Transforming population and migration statistics: Emigration patterns of non-EU students
Transforming population and migration statistics: International student employment activity
Transforming population and migration statistics: Benefits and income activity patterns
Transforming population and migration statistics: Patterns of circular movement into the UK

Transforming population and migration statistics: NINo and NHS registration lags

Recommended

More Related Content

More from Office for National Statistics (20)

Recently uploaded (20)

Transforming population and migration statistics: NINo and NHS registration lags