Economic contribution of A8 migrants

19 October, 2006


1. Further investigation into the earnings of East European workers reveals that 95% earn less than 8 per hour. At this level their contribution to GDP is probably slightly negative. Their tax and National Insurance contribution is just over half that of the UK employed population. However, they have very few dependants. For so long as this lasts and for so long as they are not taking jobs from British workers, they make a small positive contribution to the Exchequer. However both these could change. Conclusions and recommendations are at paragraphs 19-25.


2. When the 10 countries from Eastern and Southern Europe acceded to the EU on 1st May 2004 only citizens of Malta and Cyprus enjoyed total freedom to work throughout the EU. Restrictions were placed on the employment of the citizens of the other 8 countries (the A8 countries) by most of the existing 15 EU states with the exception of the UK, Ireland and Sweden. These restrictions did not apply to the self-employed. The UK imposed restrictions on benefits to A8 nationals until they had worked in the UK for a full year. As a result employers had to register A8 nationals and the government keeps statistics, published each quarter (Accession Monitoring Report), on their employment details. There is no data on those who work on a self-employed basis and so the following analysis only considers employed registered workers.

3. The latest Accession Monitoring Report published in August 2006 [1], shows that 427,000 A8 workers registered to work in the UK between May 2004 and June 2006 inclusive. There is no data on how many have stayed in the UK but we have assumed that the data in the Accession Monitoring report is representative of the workers who have remained in the UK.

4. The Accession Monitoring Report shows that the A8 migrants were accompanied by 19,270 dependent children and 16,965 dependent adults. We assume that dependants will mainly accompany migrant workers who intend to stay in the UK for some time and that any dependants who have left are balanced by additional dependants who joined migrant workers after they became established in the UK. For the purpose of this analysis we have assumed that half of the 427,000 migrants have remained in the UK and that all their dependants remain in the UK.

5. This would give following A8 populations:

Total 249.7 thousand
Working 213.5 thousand
Dependants per 100 workers 17

6. The existing UK populations in summer 2005 were:

Total [2] 60.2 million
Working [3] 28.75 million
Dependants per 100 workers[4] 110

7. The analysis which follows looks both at the economic contribution of registered workers from the A8 countries measured by their contribution to GDP and at their fiscal contribution (taxes paid versus benefits received)

GDP per head contribution

8. Over the years the government has published various statistics on the impact of immigration on GDP or GDP per head.

9. The most widely quoted is that migrants make up 8% of the population but contribute 10% to GDP . This, of course, implies that they make a positive contribution to GDP[5] per head.

10. The governments use of this measure was unsatisfactory as it did not include UK-born children of migrants in the migrant population. When these children were taken into account the contribution of migrants to GDP per head was roughly the same as the UK-born population. [6]

11. We have now examined the GDP contribution of migrants from the A8 countries (the countries which acceded to the EU on 1st May 2004) based on this measure.

12. The Accession Monitoring report indicates that 78% of registered workers earned between 4.50 and 5.99 per hour. We have subsequently obtained a breakdown of the remaining 22% from the Home Office. The following table gives the results:

Home Office provided data

Hourly pay range


< 4.50




















13. The Accession Monitoring report shows that the vast majority of registered workers (97%) were in full-time employment (more than 16 hours a week). We have assumed that the average time worked is 39 hours per week (97% at an average of 40 hours and 3% at an average of 16 hours) for the full 52 weeks in a year.

14. This gives average annual earnings of 11,800 [7]. By comparison the average earnings of the employed working population overall was 22,000 in 2005 [8]. Thus earnings of A8 migrant workers was just over half those of the UK employed population as a whole.

15. Measured by earnings per worker the productivity of the A8 workers is therefore extremely low. However, they support just 17 dependants for every 100 workers (see table in paragraph 4 above) compared to 110 dependants for every 100 workers for the UK population as a whole (see table in paragraph 5 above). Average earnings per head are therefore about 10,000 for A8 registered workers and their dependants compared to about 10,500 for UK employees and their dependants. This implies that their contribution to GDP per head for the economy as a whole is probably slightly negative.

Fiscal contribution

16. Based on this earnings data the average A8 worker and his/her employer would contribute approximately 2,900[9] in employment taxes (tax and employees and employers national insurance contributions). This compares with about 5,500 for the UK employed population [10]. Because most A8 workers are young and have few dependants the benefits and services they receive from the public purse are relatively small. (They are not entitled to unemployment benefits until they have worked in the UK for at least a year). The main attributable costs are health, policing, transport, child benefits, tax credits and education for their children. We have estimated the cost of these benefits and services as follows (see Appendix for detailed calculations):


Average cost per AC8 taxpayer



Public order and safety




Child benefits


Tax credits


Children’s education




17. This would indicate that, based on personal direct taxes, the fiscal contribution of A8 migrant workers is roughly neutral. When indirect taxes and their contribution to corporate taxes are taken into account the position will clearly be more positive. However, no allowance has been made in these calculations for other public services (such as defence or public sector debt interest) on the basis that these costs will not vary directly, and in the short term, with an increased population. Nor has account been taken of any displacement effect on native workers; if migrant workers have replaced British workers their tax contribution does not add to the total tax take but it does add to costs of providing benefits and services.

18. Also as A8 migrants gain more dependants and as they acquire full access to benefits any small fiscal benefit would quickly disappear if they continue to work in low paid occupations.


19. A8 migrants contribution to the economy as a whole is probably slightly negative. Their fiscal contribution has been slightly positive while the number of their dependants has remained very low.

20. However, their output per worker is poor as they are engaged mainly in low productivity/low value added activities.

21. If their productivity stays constant and the number of dependants increases their contribution to overall GDP per head would become more negative. Their fiscal contribution would also become negative.

22. Their GDP and fiscal contributions would be reduced further if migrant workers had filled positions which resulted in UK workers remaining unemployed. The latest figures from the ONS show that the number of people unemployed has increased by 280,000[11] over the last year so it seems reasonable to assume that some of this increase results from A8 immigration.

23. We conclude that, although immigration of people to fill low-paid positions may have limited economic impact in the short-term, it is a highly negative strategy for the longer-term as it lowers productivity, will probably lower GDP per head and will lead to a fiscal deficit as the workers are joined by more dependants. It will also make it more difficult to move the unemployed and people on disability benefits back to work. In addition it will add to the strains on the countrys environment, infrastructure and public services.

24. At a recent conference addressing the impact of migrant workers on the Scottish economy the deputy director of the CBI Scotland said [12] We cannot fall into the trap of thinking that immigration is the sole solution to the skills problems we face. It is not an alternative to up-skilling our indigenous workforce, nor should immigration be seen as an alternative to labour market policies that target those on incapacity and unemployment benefits and help them back to work. Developing the skills and abilities of the existing and future working population must be a priority. Dave Moxham, deputy director of the Scottish TUC, supported this position saying . instead of migrant workers being targeted in low skill occupations, their qualifications should be properly accredited, their language skills improved and they should be encouraged to seek out better skilled occupations higher up the jobs ladder where there is a great deal of demand. Our concern is that employers use migrant workers because they pay the minimum wage, with poor terms and conditions. Growth that is based on low pay is at odds with Smart Successful Scotland. We are looking to business to show social responsibility by employing indigenous workers in the lower skilled jobs and we are encouraged that the CBI is aware of that issue.

Policy suggestions

25. We broadly support the suggestions made by the CBI and TUC quoted above. In particular:

a) A8 migrants should be encouraged to fill more highly skilled positions where their qualifications and skills justify it. To help them to do this we recommend that we stop long-term work-related immigration (immigration leading to settlement) from outside the European Union apart from the very highly-skilled.

b) Work migration from Romania and Bulgaria should be restricted to skilled migrants who are filling vacancies which cannot be filled by workers from the EU25 countries. We should not open our labour market to them until other major EU countries do the same.

Appendix Calculation of costs

1. Health costs

Total estimated health costs 2005-6 = 89.435 billion[13]
Distribution of costs by age (DoH annual report Hospital and Community Health Services by age):
Children under 16 (including birth costs) 14.1% (Population 11.6 million [14]) cost per child 1,090
Age 16-44 24.2% (Population 24.2 million) cost per person = 890
Age 45-64 18.4% (Population 14.7 million) cost per person = 1,120
Age 65+ - 43.1% (Population 9.7 million) cost per person = 4,000

Distribution of migrant workers 93% - age 16-44 and 7% aged 45-64.
Assuming adult dependants have the same profile then the overall distribution of AC10 migrants by age is:

86% aged 16-44 (population = 214.3k), 6.5% aged 45-64 ( 16.1k) and 7.5% aged below 16 (19.3k).

The overall cost is therefore about 230 million and the cost per worker is 1,080.

2. Public order and safety
The average cost per person is assumed to be the same for the AC8 migrant population and the total population. Total cost is 30.1 billion[15] or 500 per head of population. This equates to c 580 per AC8 worker who bear the costs of their dependants.

3. Transport
Assumptions as in 2. Total cost 18.4 billion[16] . Cost per head of population 305. Cost per AC8 worker (bearing dependants costs) = 360.

4. Child benefits
Current benefit is 17.45 a week for the eldest child and 11.70 a week for each additional child. Assume an average of 1.3 children per claim (The Accession Monitoring Report indicates that 9,446 children were involved in the 7121 claims allowed in the Q2 2006) the average benefit would be about 21 a week or 1,092 p.a.. The Accession Monitoring Report shows that 27,280 applications for child benefit have been approved. This equates to a claim for every 7.8 workers giving a cost per worker of 140 p.a.

5. Tax credits
According to the Accession Monitoring Report 14,000 applications have been approved for tax credits. The Inland Revenue (in its leaflet WTC2 Child tax Credit and Working Tax Credit A guide) gives the total benefit for someone receiving child tax credits and working tax credits as 4,160 for someone earning 10,500 p.a with one child or 5,855 with two children. For a person earning 15,000 a year the corresponding payments are 2,495 and 4,190. Given an average earnings of 11,800 and average number of children of 1.3 (see 4 above) a typical payment may therefore be in the region of 4,000 p.a. The cost per worker based on this would be about 260.

6. Childrens education
Total education costs for the under fives, primary schools and secondary schools were estimated to be 42.8 billion in 2005-6. The cost per child (under 16) was therefore in the region of 3,700 per child. The cost per worker is therefore about 330 p.a. ( 213,000 workers supporting 19,000 children under 16).


[1] rt May 2004 to June 2006 published jointly by the Home Office, DWP, HMRC and DCLG
[2] ONS Mid-year 2005 Population estimate
[3] No. of people in employment in July 2005 – ONS.
[4] The working population shown includes self-employed and it is implicitly assumed that the self-employed and employed support a similar number of dependants.
[5] Claim originally made by Ruth Kelly, then Financial Secretary to the Treasury, on 7th May 2002 – Hansard column 33W. See also MWUK Briefing paper 1.5
[6] Assumes earnings at the mid-point of the range shown in the table, £4.50 for those earning less than £4.50 an hour and £25 per hour average for those earning above £20 an hour.
[7] The percentages in each earnings band are assumed to be 0.7%, 0.6 %, 0.4%, 0.2% and 0.1% for those earning £12.00- £13.99, £14.00- £15.99, £16.00-£17.99, £18.00 – £19.99 and £20.00 respectively.
[8] Annual Survey of hours and earnings (ASHE) for all employee jobs 2005.
[9] Based on 2005-6 tax bands and rates and national insurance scales.
[10] Based on data from ASHE (see 8) across all employee jobs 2005.
[11] ONS News Release 13 September 2006.
[12] As reported in The Scotsman 30 Sept. 2006
[13] HM Treasury: Public Expenditure Statistical Analysis 2006
[14] Populations by age – GAD 2004-based principal populations for mid 2005
[15] As 18
[16] Ibid