The Impact of the Recent Migration from Eastern Europe on the UK Economy

Discussion Paper
No.17
The Impact of the Recent Migration from Eastern Europe
on the UK Economy

by David G. Blanchflower, Jumana Saleheen and Chris Shadforth

---

External MPC Unit Discussion Paper No. 17*




The Impact of the Recent Migration from Eastern Europe on the UK
Economy

By David G. Blanchflower**, Jumana Saleheen† and Chris Shadforth‡

January 2007



ISSN: 1748 – 6203
Copyright Bank of England 2007




An earlier and shortened version of this paper was given as a speech at a lunch for
members of the Cambridgeshire Chamber of Commerce on Thursday, 4th January 2007.
We thank Kate Barker, Charlie Bean, Richard Dickens, Neal Hatch, Andrew Holder,
Ethan Lewis, Rachel Lomax, Lavan Mahadeva, David Metcalf, Andrew Sentance, Sally
Srinivasan, and Mark Stewart for helpful discussions.





(*) Disclaimer: These Discussion Papers report on research carried out by, or under
supervision of the External Members of the Monetary Policy Committee and their
dedicated economic staff. Papers are made available as soon as practicable in order to
share research and stimulate further discussion of key policy issues. However, the views
expressed are those of the authors and do not represent the views of the Bank of England
or necessarily the views of other External Members of the Monetary Policy Committee.

(**) Bruce V. Rauner '78 Professor of Economics, Dartmouth College, NBER and
Member of the Monetary Policy Committee, Bank of England.

(†) Research Advisor, External Monetary Policy Committee Unit, Bank of England

(‡) Economist, External Monetary Policy Committee Unit, Bank of England

---

Executive Summary

a) UK population growth over the last thirty-five years has been remarkably low in
comparison with other countries; the population grew by just 7% between 1971 and 2004,
less than all the other EU15 countries (except Germany), Australia, Canada, Japan, New
Zealand and the United States, plus China and India. The rate of UK population growth
did exceed that of six countries from the former Soviet bloc (Czech Republic; Croatia;
Estonia; Hungary; Latvia and Bulgaria), but was below the rates of growth of Poland,
Lithuania, Slovakia and Slovenia.

b) The UK population has grown at a faster pace since the turn of the millennium. This
recent growth has been driven primarily by changes in net migration. Both the inflow
and outflow rates have risen, but the inflow rate has risen more rapidly recently, with an
influx of migrants from eight East European (A8) countries. However, the increase in the
net migration flow predates the influx of A8 migrants, reflecting a steady rise in the
number of immigrants from Asia and the Middle East too.

c) The propensity to migrate to the UK is higher the lower is GDP per capita in each of
the A8 countries. The decision is also strongly correlated with life satisfaction scores and
unemployment rates, but is uncorrelated with employment rates or rates of inflation.

d) There appears to be consistent evidence from the Worker Registration Scheme and
National Insurance Number applications that approximately 500,000 migrants from the
A8 countries had come to work in the UK between May 2004 and late 2006. But other
sources suggest a significant proportion of these workers have likely returned to their
country of origin.

e) The empirical literature from around the world suggests little or no evidence that
immigrants have had a major impact on native labour market outcomes such as wages
and unemployment. Recent work by a number of other authors for the UK is also
consistent with this view.

(f) The impact of recent migration from the A8 countries on the UK economy will be
determined by the extent to which immigrants add to supply relative to demand, since it
is the balance between these two factors that determines prospects for inflation. We
argue that, at present, it appears that A8 immigration has tended to increase supply by
more than it has increased demand in the UK (in the short run), and thereby acted to
reduce inflationary pressures.

(g) There seems to be broad agreement that immigration is likely to have reduced the
natural rate of unemployment in the UK over the past few years. But there is some
uncertainty about what has happened to the natural rate in the very recent past and what
might happen to it in the near future. This is because immigration has not been the only
shock to affect the labour market very recently.

---

1) Introduction

The recent rise in migration to the UK from eight EU Accession countries (the Czech
Republic; Estonia; Hungary; Latvia; Lithuania; Poland; Slovakia; and Slovenia – the A8
countries) has generated a good deal of controversy. How many A8 immigrants are there
in the UK? Where did they come from and when? What impact has their influx had on
the UK economy and what likely impacts will they have in the future? Most importantly
for the Monetary Policy Committee (MPC), what macroeconomic effects have they had?

We attempt to address these questions here. First, we examine the change in population
in the UK over the last thirty-five years and note that growth is very low by international
standards. The UK population has, however, grown at a faster pace since the turn of the
millennium, driven most recently by migration from the A8 nations. It appears that the
propensity to migrate to the UK from these countries is higher the lower is GDP per
capita. Second, we examine the various sources of data that are available on the numbers
of A8 immigrants that have arrived in the UK in recent years. There is broad agreement
from the various data sources on the numbers involved – half a million workers is likely
to be an upper bound for the stock of A8 migrants who are in the UK in late 2006. Many
of the new 'migrants' may have stayed for only a short time and then returned home, to
possibly return again at a later date. Third, we examine the characteristics of the recent
flow of individuals from the A8 countries that have arrived in the UK since accession,
and find that they are relatively young, male, have low unemployment rates, lower wages,
and high self-employment rates and are especially likely to be in temporary jobs. Finally,
we turn to the macroeconomic implications of A8 migration to the UK, and argue that
this immigration has made the labour market more flexible and likely lowered the natural
rate of unemployment and the NAIRU.

1) Population changes

According to official estimates published by the Office for National Statistics, the UK
population grew by just 8.2% between 1971 and 2006, from 55.9 million to 60.5 million.
In contrast, the United States population grew by 44.6% over the same period, from 207.7
million in 1971 to 300.3 million in 2006.2 Indeed population growth across many
countries has been greater than in the UK over the past three decades (Chart 1). Over the
period 1971-2004, population growth in the UK ranks 31st out of 38 European and other
large nations for which data are available (listed in Table 1, excluding Romania), with
only Germany (East and West) and six East European countries having had slower
population growth (Czech Republic; Croatia; Estonia; Hungary; Latvia and Bulgaria).
All the other major industrialised nations have had faster rates of population growth.
Growth was particularly rapid in the US (+42%) as noted above, but also in Australia
(+54%); Canada (+45%); Spain (+25%); Japan (+21%); and France (+18%). The Indian
population roughly doubled over the same period (96%), while the Chinese population
grew by 52%. It is clear that UK population growth has been extremely low by
international standards – it is even below that of Italy (+7.6%).

2 Source: Statistical Abstract of the United States, 2001, 2006 and www.census.gov.

1

---

Chart 1: Population growth, 1971-2004
Chart 2: Annual UK population growth,
based on official estimates of migration
Percentage change in population,
Percentage change on a year earlier
1971-2004
100
Births-Deaths
0.7
A8 net migration
0.6
80
Non-A8 net migration
0.5
Total population growth
60
0.4
40
0.3
20
0.2
0.1
0
0.0
-20
-0.1




l
i
a



da

r
ia
-0.2
i
a

n
ce
a
d

s
t
r
a
na
A
i
n
S
a
K
lg
I
n
a
p
pa
r
an
u
971
976
981
986
991
996
001
006
Au
C
U
S
J
a
F
U
B

1
1
1
1
1
1
2
2

Source: Eurostat, US Statistical Abstract 2006 and Health Source: ONS
Statistics Quarterly, 32, Winter 2006


1. Official disaggregated data are not yet available for 2006.
1. The chart shows the ordering of population growth between
1971 and 2004 of the 38 countries listed in Table 2.

Long-run trends, however, mask some significant short-run changes in population growth
(Chart 2 & Table 2a). The UK population grew by 2.8 million (4.9%) between 1971 and
1999, but the population has subsequently risen by a further 1.8 million (3.2%). The
main cause of this increase has been a rise in net inward migration, driven by an increase
in the inflow rate to the UK; the outflow rate has remained little changed over the years,
although there has been a pickup since 1998. The ratio of births to deaths has seen less
variation. In 2004-5 net migration accounted for two-thirds of the change in population
(248/375). To place these numbers in some degree of context, net (legal) migration in the
United States accounted for approximately one third of net population growth in 2004-
2005.3 Table 2b makes it clear that the scale of net inward migration to the UK has been
much lower than in most other EU countries until recently, and even now remains below
the levels of both Italy and Spain.
In summary, UK population growth appears to have been extremely low by international
standards over the past three decades. As such, it would seem likely that the UK could
absorb a relatively small inflow of immigrants, by international standards. As we note
below, it appears that it has already done so to a considerable degree; the entry of recent
A8 migrants appears to have improved the workings of the labour market, reduced wage
and inflationary pressures and lowered the natural rate of unemployment.




3 Population in the United States on July 1st 2004 was 293,657,000 increasing to 296,410,000 on July 1st
2005 a net increase of 2,754,000 or 0.94%. This increase was made up of 4,129,000 births, 2,425,000
deaths and net legal migration of 1,050,000. Source: US Census Bureau -
http://www.census.gov/compendia/statab/tables/07s0004.xls
2

---

2) Why has immigration to the UK increased since the turn of the millennium?

The increase in net migration coincides with changes in UK immigration policy and the
relative attractiveness of the UK’s economic position over the past decade. Most
recently, the increase in the inflow rate of migrants appears in large part to be attributable
to immigration policy that accompanied the accession of the A8 countries on May 1st
2004;4 citizens from the A8 nations obtained free movement and the right to work in the
UK, Ireland and Sweden from this date.5 But there has continued to be a steady growth
in the inflow of migrants to the UK from non-A8 countries too, especially from the New
Commonwealth.

The literature on immigration focuses on the economic factors that determine migration.
Very simply, the literature says that individuals will compare the income benefits from
migration with the economic and social costs of moving. If the benefits outweigh the
costs, they may choose to migrate. The gain from moving will be calculated as the
expected income differential between the destination country and the country of origin,
which will in turn be determined by the relative probability of getting a job – captured by
differences in the unemployment or employment rates.

Naskoteen and Zimmer (1980) find for the US that a 10 percentage point increase in the
wage differential between the countries of destination and origin increases the probability
of migration by 7 percentage points. Borjas (2005) finds that a 10 percentage point
increase in the rate of employment growth in the state of origin reduces the probability of
migration by approximately 2 percent. There is also evidence that migration is most
common among younger and more educated workers (Borjas 2005). Moreover, workers
who have just migrated are extremely likely to move back to their original locations. The
probability of a migrant returning to the state of origin within a year is about 13 per cent
and the probability of moving to another location is 15 per cent (Devanzo, 1983 and
Dustmann, 2003). Zaiceva (2006) summarises the empirical literature on the potential
migration flows, which he shows to be consistent with between 2 and 4 per cent of the
residents of Central and East European countries (CEECS) moving West, in the long run,
which constitutes around 1 per cent of the EU15 population. Zaiceva also presents
evidence from simulations suggesting that the majority of migrants are predicted to come
from Romania, Poland and Bulgaria, which is consistent with other estimates from the
literature.

Gilpin et al (2006) examine whether A8 citizens have come to the UK because it offers a
higher standard of living (GDP per capita) or a higher probability of getting a job
(measured by the inverse of the unemployment rate), or both. They consider data from
the Worker Registration Scheme (see Appendix A for more details on this data source).

4 In addition Malta and (South) Cyprus also joined the EU at that date. Bulgaria and Romania joined the
EU on January 1st 2007.

5 Finland, Greece, Portugal and Spain opened their labour markets to these workers on May 1st 2006, while
Italy followed in late July 2006. Five other countries (Belgium, Denmark, France, the Netherlands and
Luxembourg) alleviated restrictions in 2006 (Zaiceva, 2006).

3

---

They compute the number of WRS registrations as a percentage of the home country
population and show it is correlated with GDP and unemployment. We update their
analysis in Table A. It is apparent that a larger fraction of people of Lithuania (1.60%);
Latvia (1.25%); Slovakia (0.92%) and Poland (0.79%) have come to the UK compared to
Estonia (0.42%), the Czech Republic (0.24), Hungary (0.14) and Slovenia (0.02) (Table
A). Gilpin et al. find that countries with the lowest GDP per head, such as Lithuania
(2500 Euros) are more likely to be registered on the UK WRS than those from countries
with higher GDP, such as Slovenia (11,400 Euros).6 Workers in the WRS data are also
more likely to come from countries with the highest unemployment rates, such as Poland
(19.0%). Pedersen et al (2004) found similar effects for GDP per capita and the
unemployment rate in both source and destination countries in their study of migration
flows into OECD countries in the 1990s.

Table A: WRS applications May 2004 – September 2006 as a proportion of home
country populations

WRS registrations
GDP per head
as a percentage of
WRS
Population Unemployment Employment (2005) (Euros per
home country
registrations
(2004)
rate
rate
head at 1995
population
(000s)
(millions)
(2004)
(2004)
exchanges rates and
Czech Republic
0.24
24.7
10.2
8.3
64.2
5,200 €
Estonia
0.42
5.5
1.3
9.7
63.0
4,000 €
Hungary
0.14
14.6
10.1
6.1
56.8
5,000 €
Latvia
1.25
28.9
2.3
10.4
62.3
3,100 €
Lithuania
1.60
54.6
3.4
11.4
61.2
2,500 €
Poland
0.79
307.6
38.6
19.0
51.7
4,200 €
Slovakia
0.92
50.2
5.4
18.2
57.0
4,200 €
Slovenia
0.02
0.5
2.0
6.3
65.3
11,400 €
Average/Total
0.66
486.6
73.3
UK
59.5
4.7
71.6
EU-25
458.9
9.1
63.3
Correlation
0.508
-0.198
-0.709


Source: Gilpin Table 4.3 updated. Human Development Report, 2006.

The correlation coefficient is clearly highest with 2005 GDP per head, as noted by Gilpin
et al., and even higher when GDP is in logs (r=0.837). The correlation is slightly weaker
with the unemployment rate, but especially low with the employment rate.7
Interestingly, Schiopu and Siegfried (2006) found that the difference in GDP between the
host and home countries increases the size of remittances.

Data are also available on the country’s rank on the 2005 Human Development Index
taken from the Human Development Report of the UN, and their average life satisfaction

6 Expressed as Euros per inhabitant at 1995 exchange rates and prices.

7 The correlation with the CPI is only -0.166, also drawn from the HDR 2006, Table 14.

4

---

score for 2002 taken from the Eurobarometer Surveys.8 A lower rank on HDI is better, a
higher life satisfaction score is better. The propensity to migrate is even more highly
correlated with these two measures than it is with GDP per capita (Table B).9

Table B: Human development and life satisfaction correlations with WRS
applications

Life
Gross
Life
HDI
expectancy enrolment GDP per capita
satisfaction
Rank
rate
rate
(PPP US 2004)
score (2002)
Czech Republic
30
75.7
81
$19,408
2.82
Estonia
40
71.6
92
$14,555
2.55
Hungary
35
73.0
87
$16,814
2.60
Latvia
45
71.8
90
$11,653
2.46
Lithuania
41
72.5
92
$13,107
2.44
Poland
37
74.6
86
$12,974
2.69
Slovakia
42
74.3
77
$14,623
2.56
Slovenia
27
76.6
95
$20,939
3.05
Romania 60
71.5
75
$8,480
2.39
Bulgaria
54
72.4
81
$8,078
2.02
UK
18
78.5
93
$30,821
3.17
Correlation
0.797
-0.526
-0.005
-0.823
-0.751


Source: Human Development Report 2006 & Eurobarometer

We explored the correlations with each of the components of the HDI, which includes the
life expectancy rate (r=-0.526); the combined gross enrolment ratio for primary,
secondary and tertiary schools (r=-0.005); and GDP per capita (PPP US$) for 2004,
which has an even better correlation (r=-0.823) than the GDP data used by Gilpin et al,
and in logs the correlation is -0.836. It is already well-known that East Europeans are
more likely to report that they are unhappy (Blanchflower, 2000 and Blanchflower and
Freeman 1999). In a recent Candidate Eurobarometer collected between September and
October, 2002 (ICPSR #4062) respondents were asked the following question.

Q. On the whole, are you very satisfied, fairly satisfied, not very satisfied or not at all
satisfied with the life you lead? A. Very satisfied=4; Fairly satisfied=3; Not very
satisfied=2; Not at all satisfied=1


8 The HDI is published annually by the United Nations and is a score that amalgamates three indicators:
lifespan; educational attainment and adjusted real income (Blanchflower and Oswald, 2005).

9 The correlation is similar with the 2001 life satisfaction score (-0.72), but lower in 2003 (-0.55), 2004 (-
0.57) and 2005 (-0.58)
5

---

The means of the life satisfaction score variable correlate reasonably well with the
propensity to migrate (r=-0.751) and considerably better than the unemployment or
employment rates.

In summary, the favourable macroeconomic climate (low unemployment) and high
standard of living in the UK (GDP per capita) are reasons why immigrants from the A8
countries may have been attracted to the UK, and why immigrants from Bulgaria and
Romania may be attracted in the future. The OECD has recently (2006b) projected that
GDP will grow particularly quickly in both Poland and Slovakia over the next couple of
years. The OECD projects a growth rate of around 8% in both 2006 and 2007 in
Slovakia, and 5% in Poland, where, it suggests remittances from migrants will sustain
consumption. The other member of the OECD is Hungary, which is projected to grow at
4% in 2006, but only at 2% in 2007. Rapid GDP growth in some A8 countries and
improvements in their unemployment rates might suggest a reduction in the flows of both
permanent migrants and especially temporary workers to the UK from the A8 countries in
the near future. Pedersen et al (2004) studied migration flows into 27 OECD countries
from 1990-2000 and found that network effects, measured as the coefficient of the stock
of immigrants of own national background already resident in a country, had a large
positive impact on immigration flows. This suggests that rather than dissipate, the flows
may well continue well into the future although the scale of the flow remains uncertain.
Further the Bank’s regional Agents have undertaken a survey of business contacts
concerning their use of migrant labour that suggested that a majority of firms were
expecting to make greater use of such labour into the future.

3) Size of migration flows from the A8

It is difficult to get an exact estimate of the size of the flows of individuals from the A8
countries to the UK since accession. Some estimates suggest that around 500,000 A8
workers have come to the UK, but other sources suggest many fewer. It is also unclear
what proportion of such workers are long-term migrants and what proportion are here for
a short time and have subsequently returned home, perhaps to return again in the future.
The scale and nature of this flow is an important question for policy makers because it
affects the labour market and the wider economy. It is therefore important to try and
understand and reconcile the differences between different data sources as far as possible.
Doing this, we find that 500,000 workers is an upper bound for the stock of post-
accession A8 workers in the UK in late 2006.

There are four main sources of data on the flow of A8 individuals. Further details of the
sources are provided in Appendix A. We examine estimates of the numbers using each
data series in turn.

1) Worker Registration Scheme
Table 3 reports the number of workers, by A8 country, who registered, and were
approved for work, on the WRS – a registration scheme for employees from the A8
countries. The self-employed do not need to register under the WRS. The largest
number in every year has come from Poland. Between 2004Q2 and 2006Q3 over
300,000 Poles had registered, which constitutes 63% of the total. The next largest group
(11%) is from Lithuania, which had only 3.4 million residents in 2004, compared with
6

---

38.2 million in Poland.10 The flows by quarter are very similar. The WRS suggests that
486,660 workers had registered to work in the UK since accession – a further 3,895 were
refused, 1,035 were exempt and 14,950 applications were withdrawn making a grand
total of 510,340 applicants. The three main countries from which migrants have come
are Poland (63%); Lithuania (11%); and Slovakia (10%).

Ireland has been the other major recipient of migrant flows from the A8. According to
Ireland’s National Training and Employment Authority (FÁS), over the first twelve
months to April 2005, 83,000 Personal Public Service Numbers (PPSN) were issued to
EU10 nationals, which is equivalent to almost 4% of Ireland’s labour force. Some of
these people were, however, already in Ireland prior to May 2004. Another 66,000 PPSNs
were issued in the six months from May to October 2005, an increase of 46% on the same
period for the previous year. The OECD (2006a) suggests that the country composition
of A8 workers to Ireland is similar to that of the UK - Poland (54%), Lithuania (19%),
Latvia (9%), and Slovakia (8%).11

2) Total International Migration, predominantly from the International Passenger Survey
Table 4 provides official estimates of migration based on the International Passenger
Survey (IPS), showing changes in the inflow and outflow rates of 'long-term international
migrants' – a migrant is someone who changes his or her country of usual residence for at
least 12 months.12 It is apparent that both have risen steadily since 2002, although the
inflow rate has experienced the more rapid rise. Table 4 suggests that there was a net
positive migration flow of 223,000 in 2004 and a further 185,000 in 2005, making a total
of 408,000. As part of these, the IPS suggests that 132,000 individuals have come to the
UK from the A8 countries, which is a considerably lower number than from the WRS,
why?

Some concerns have been expressed about the accuracy of the data from this source as it
covers only the principal air and sea routes, and the Channel Tunnel. There is evidence
of an increase in the number of passenger journeys terminating at airports not routinely
covered by the IPS – ie to terminals outside of Gatwick, Heathrow and Manchester –
particularly to and from Poland (having increased from 516,000 in 2003 to 1,845,000 in
2005) – so some migrants may have been missed.13,14 Furthermore, many of these extra
journeys appear to have resulted from an increase in short-term visits to the UK – those
of less than three months. Table C shows that there has been a steady increase in the

10 Source: Eurostat.

11 OECD (2006, Table 1.22) suggests that the other major recipient countries other than Ireland and the UK
have been Italy (84,000); Norway (36,000) and the USA (21,000).

12 This is the UN-recommended definition of a migrant. See ONS (2006a).

13 See www.caa.co.uk for airline statistics on passenger journeys in and out of the UK.

14 There have also been notable increases in the number of passenger journeys to and from Slovakia (from
29,000 to 284,000) and to and from Latvia (from 61,000 to 309,000), again with the vast majority of the
increases being to airports other than Heathrow, Gatwick and Manchester, not routinely covered by the IPS.

7

---