An Amnesty for Illegal Immigrants?
Migrationwatch Response to the GLA Paper

4 November, 2009

Summary

1 The following are the main points:

• The key weakness of the amnesty proposal is that it will encourage further illegal immigration. Rather than solve the problems it will simply transfer them to a successor group of illegals. This is the very clear lesson of various amnesty schemes elsewhere, notably Spain, which the report largely omits.
• The number of illegals is seriously underestimated, given that 2 million visas are issued every year and there are no checks on departure. The number of illegals is more likely to be 1 million than 0.5 million. Subsequent dependants could make this 2 million.
• The “calculation” of extra output is no more than a speculative example.
• The impact on social housing is seriously underplayed. All will become eligible yet the report assumes that only 40% will acquire social housing. Even on that optimistic assumption the cost of public sector subsidy would be £6.2 billion.
• The report itself acknowledges that: "It is highly unlikely that the extra dwellings required would actually be built. The much more probable outcome is that there would be increased competition for the existing social stock. This in turn might adversely affect local relations and social cohesion, as the evidence shows that competition for housing is an important element in increasing tensions at the local level." (p 75).
• The assumption that those legalised will earn, on average, more than double the minimum wage is highly optimistic. We have taken minimum wage plus 50%.
• The costings contain understated amounts for health and education, apparently on the grounds that illegal immigrants already have almost full access to them. The report admits (p 104) that the longer term costs of benefits could be £2.0 billion per year. Even this does not include post retirement costs. Our estimate of the lifetime cost would be of the order of £134 billion.
• The administrative problems, notably the risk of fraud, would be severe. The report notes that a U.S attempt to regularise Mexicans recorded a 75% rate of fraud.

Numbers

2 The estimate of the number of illegals in Britain is clearly critical. The Report recognises (page 36) that the consensus among informed observers has been that those who stay beyond the validity of their visas are likely to be a much larger group than those who entered illegally. However, it then estimates that the total number of both categories from 2001-7 at only 50,000 (page 42). This is an absurdly low estimate. In that period approximately 12 million visas had been issued to applicants who, by definition, pose a risk of overstaying (that is why they need a visa). The GLA assume that less than 1 in 200 will overstay despite the widely known fact that there is no check on departures. If, in fact, 1 in 20 overstayed this would approximately double their central estimate to over 1 million. All these people once granted settlement would be able to bring over their families or their future spouses. In due course, that could well amount to some 2 million people.

3 A further weakness of the Report is the exclusion of any consideration of those who are now here legally but are working illegally. These are precisely the people who are most likely to stay on if they see the prospect of an amnesty.

Effectiveness

4 The central question is whether an amnesty would resolve the problem of illegal workers or would simply encourage more illegal immigration. The Report lists six amnesties in Italy since 1986 which have attracted 2 million applications. On almost every occasion the numbers applying were greater than on the previous one. The Report's section on Spain is incomplete. It stopped at 1996. Had they included the amnesties in 2000, 2001 and 2005 they would have found that the numbers had doubled on each occasion, reaching 700,000 at the last amnesty. Although the circumstances differ in each country, these precedents illustrate beyond doubt that amnesties attract further illegal immigrants. This may explain the footnote in the report to the effect that they do not discuss international experience in detail (page 55). They claim that relevant points are made throughout the report where appropriate but there is no proper examination of the experience of Spain or Italy. There is certainly no mention of the fact that the unemployment rate in Spain is now 17% and could reach 20%.

5 The Report goes on to claim that

"the likelihood of large scale additional irregular immigration is far lower in the UK, where most irregulars come from further afield - and could only occur if border controls were ineffective."

The distance is irrelevant, given the relatively low cost of transportation. With 2 million visas issued every year, it is surely obvious that the scope for overstaying would be huge, especially given that many visas are issued in countries where wage rates are between one fifth and one twenty-fifth of those in Britain.

Effect on Output

6 The Report is astonishingly thin on this point - just two paragraphs in 125 pages. This passage (page 73) starts "Neither the literature nor currently available UK data provides a quantitative basis from which one could at all reliably start to estimate the scale of effects on UK output which would be likely to follow from regularisation". It goes on to take as what is described as "an example". The full text is at Annex A. This proposes that if employment rates were to rise by 6% and if average weekly earnings were to rise by 25% and if other things were equal and if other categories of income grew in parallel, the expected increase in GDP would be some £3 billion. This, clearly, is nothing but a guess. Most of the money would go to immigrants in the form of higher income. The impact on the national budget is tackled separately (see below).

Housing

7 The section on housing seeks to make a distinction between regularising illegal immigrants and granting indefinite leave to remain (ILR) yet the whole purpose of this scheme is eventually to grant ILR at which point illegal immigrants become entitled to social housing and full access to the welfare state. They are also entitled to bring over their dependants which will move them up the priority list. For no very obvious reason, they assume that only 40% of those who are regularised would obtain social housing. Even on that optimistic assumption, the public sector subsidy would cost £4.4 billion for London and £6.2 billion for the UK.

8 The Report acknowledges that it is

"highly unlikely that the extra dwellings required would actually be built. The much more probable outcome is that there would be increased competition for the existing social stock. This in turn might adversely effect local relations and social cohesion, as the evidence shows that competition for housing is an important element in increasing tensions at the local level." (Page 75)

The impact on the national budget

9 The report admits (p113) that "Many of the numbers that have been generated for this report have required heroic assumptions". Indeed so. The first such assumption is that, once legalised, these workers would earn on average more than twice the minimum wage - £13.29 rather than £5.73 an hour. Given that a major reason for their present employment is their willingness to accept a low wage, we regard this assumption as unrealistic. Accordingly we have assumed an average of the minimum wage plus 50 %.

10 The report estimates the total costs into the longer term at just £2.4 billion per annum. It achieves this low figure partly by taking insufficient amounts for health and education on the grounds that illegals already have access to them. However, the costing is done at the point of amnesty. It takes no account of the lifetime costs as people grow older and draw pensions, although it admits that the longer-term cost just of benefits could be £2 billion a year. Annex B contains an estimate of the lifetime costs of an amnesty on stated assumptions. The outcome £134 billion is a staggering amount but, even so, is probably an underestimate for the reasons given in paragraph 6 of that Annex. Annex C contains a comparison of the two costings.

Annex A
Extract from GLA Economics Paper

"Economic Impact on the London and the UK Economy of an Earned Regularisation of Irregular Migrants to the UK" (Page 73)

Neither the literature nor currently available UK data provides a quantitative basis from which one could at all reliably start to estimate the scale of effects on UK output which would be likely to follow from regularisation. They are, however, supportive of a qualitative argument that says that irregularity must operate as a significant constraint depressing the level of jobs to which migrants can get access, relative to what would be expected from their skills and other personal characteristics. It may also very well allow some employers to sustain relatively unproductive activities or methods of production by exploiting the weak position of irregular workers to keep wage costs down. Legal recognition of a substantial proportion of the UK's current irregular resident population should thus allow for increases both in productivity and in the level of employment (without further migration).

The potential scale of the positive effects that could follow is certainly large. To take an example, if it were the case that - over some time interval - regularisation allowed employment rates among currently irregular adult residents to rise by six per cent (from a base of 50 per cent) and average weekly earnings among those employed to increase by 25 per cent, this would represent a 40 per cent addition to the earnings of this population. Other things being equal, we would expect this to be accompanied by an equivalent percentage increase in production from their work. In numeric terms, taking a baseline estimate of £372 per week earnings for workers from this group, generating an annual income of £9,672 per adult head, the increase would be £3,869 per eligible adult. On our central estimate of numbers, this would total some £1.6 billion, increasing total UK income from employment by about 0.2 per cent. If other categories of income (including profits) grew in parallel, the expected increase in GDP would be some £3 billion.

Annex B
Financial impact of an amnesty for illegal immigrants

1 Annual Income tax receipts

The GLA calculates that an amnesty could yield an additional £846 million a year of Income Tax and National Insurance receipts.

However they add:

“It has to be emphasised that these are ballpark estimates at best”

The most important factor that defines the tax that an individual pays is their hourly rate of pay.

For their ballpark estimate the GLA have used an average weekly wage of £465 a week (£13.29 an hour or £24,180 a year). This is far too high.

The GLA report itself states that the earnings of 28% of working illegal immigrants are in the lowest 20% of weekly earnings and 12% are paid less than the minimum wage. Their report states that once regularised, 18% of working immigrants will still be in the lowest 18% of weekly earnings and 10% will still earn less than the minimum wage.

Taking the above statistics into account, the £465 a week (£13.29 an hour) ‘ballpark estimate’ is clearly excessive; their average pay is bound to be closer to the £5.73 an hour minimum wage than the £13.29 an hour they have used in their calculations.

Taking account of tax credits, single people do not pay tax until their pay reaches £6.02 an hour (£210 a week). A sole-earner family with two children does not pay tax until their income reaches £12.42 an hour (£435 a week). This helps to explain why our estimate of income tax receipts is a negative £427 million a year, comprising £130 million tax paid by single workers and £131 million paid by two income families but more than offset by £688 million of taxes received by sole earner working families, their tax credits received being more than the taxes they pay.

2 The annual cost of an amnesty

The GLA report’s best estimate of the extra costs is £2.4 billion a year[1] They arrived at this figure by looking at what illegal immigrants currently cost and then what the extra costs would be once they were regularised.

Whilst they acknowledge that future costs will be incurred they have taken insufficient account of them, for example:

- the costs of paying pensions in retirement are omitted. When presenting their report[2] the GLA acknowledged that there would be pension costs but had chosen to ignore them because they were so far in the future.

- many of those legalised will exercise their right to bring their families to the UK or will marry and raise a family in the UK.

- health costs will increase as people age. The GLA Report acknowledges that future Social Care costs have been deliberately excluded from their costings[3].

- The cost of supporting unemployed immigrants. The GLA report states that only 50% of working age illegal immigrants are in work and this will only increase to 56% once the workers are legalised. The report has taken insufficient account of supporting the remaining 44%. This is an important omission.

3 Lifetime net costs (Costs minus Tax and National Insurance Revenues

For this paper Migrationwatch has calculated the lifetime costs for the following five family groups:

A working single person

A working family with two children assuming a sole wage earner. The Government’s own figures on which the Migrationwatch figures are based also assume a sole wage earner family [4].

A working family with two children and both spouses working full time

An unemployed single person

An unemployed family with two children

To take account of the fact that the wage rates of some immigrants will be higher than the £5.73 an hour (£200 a week) minimum wage, a rate of £8.60 an hour (£300 a week) has been used, 50% above the minimum wage.

Separate costings have been done for immigrants resident in London because Housing Benefit payments are higher. The GLA estimate that 70% of illegal immigrants live in London and the Lifetime costs per family unit resident in London are given below,

A working single person or family
Using official Income Tax, Tax Credit, Benefits, Pensions and Housing Benefit rates, Appendix 1 to this paper shows that the lifetime costs of a single person working for £8.60 an hour is £217,000, for a sole earner two child family they are £556,000 and £172,000 for a two child family where both spouses are working full time.

An unemployed single person or family
Appendix 2 shows that the lifetime cost of an unemployed single person is £666,000 and for an unemployed two child family is £1,029,000.

Please refer to Appendices 1 and 2 for summary statements of the content of the above lifetime costs. Please refer to Annex D for spreadsheets showing the detailed numbers.

4 The total net lifetime cost of an amnesty

This section applies the above lifetime net costs per family to the total number of family units in order to calculate the total lifetime costs of an amnesty.

Migrationwatch has calculated the costs using the following optimistic bases:

- The GLA central estimate of 618,000 illegal immigrants, of which 124,000 are children, 12,000 are over 65 (the costs in retirement for these people have not been included in this paper) and 482,000 are of working age [5].

- The GLA estimate that of the 482,000 working age illegal immigrants, 67% (323,000) are male and 33% (159,000) are female [6].

- The assumption that all of the 159,000 female illegal immigrants are already in a relationship with another illegal immigrant and so will be supported by their partner.

The results of the above assumptions are that the family make ups of the 482,000 working age immigrants would, at the point of an amnesty, be:

- 318,000 would be in couple relationships, comprising 159,000 males and 159,000 females

- the remaining 164,000 illegal immigrants would be single

- Of the 164,000 single immigrants, 50% (82,000) would bring families to the UK or marry in the UK and raise a family. The remaining 50% (82,000) would remain single.

- 7.5% of economically active immigrants would be unemployed, in line with the UK national average

- Those in work would earn an average of £8.60 an hour, 50% above the minimum wage

- 80% of working families have one spouse working full time and 20% have both spouses working.

- The GLA report assumption is that, once legalised, the employment rate amongst previously illegal immigrants would increase from 50% to 56%. Given that unemployment is rising rapidly, this 6% growth in immigrant employment would result in the equivalent number of existing workers being displaced. This level of displacement may be an underestimate.

- 90% of working families live in private rented accommodation and 10% are owner occupiers.

- 20% of two income families are in a pension scheme to the extent that in retirement they will not receive any Pension Credit, Housing Benefit or Council Tax Benefit. Families not saving for a pension receive full Pension Credit and have all their rent and Council Tax paid in benefits.

5 Total Net Lifetime Costs Pre and Post Retirement

On these assumptions the net lifetime cost of an amnesty would be £134 billion, comprising £77billion of costs up to retirement and £57 billion after retirement. The GLA report excludes the retirement costs.

6 Can the cost really be so high?

The calculations are probably understated for the reasons given below:

Average earnings
The rate chosen of £8.60 an hour, 50% above the minimum wage, could well be too high. The costs would be higher if earnings were only at the minimum wage, because less tax would be paid and more Tax Credits and Housing Benefit would be received. The GLA report itself states that ‘many households will be lower-income families’.[7]

A £1 reduction in the hourly rate would increase the lifetime costs by £17 billion.

Unemployment rates
The national average rate of 7.5% has been used. In reality, immigrants from developing countries are much more likely to be unemployed.

Each 1% increase in the unemployment rate increase the lifetime costs by £2 billion.

Displaced existing workers
It has been assumed that just 6% of jobs taken by legalised immigrants would result in an existing worker, or a person looking for work, being displaced. Given that UK unemployment is rising rapidly, a 6% displacement rate may be an underestimate at least until recovery starts.

Each 1% increase in the displacement rate increases the lifetime costs by £2 billion.

Marital status
We have assumed that only 50% of single immigrants would bring a wife to the UK or would marry in the UK. The net cost of a family is much higher than for a single person and so a higher marriage rate would increase the cost estimate significantly.

Each 10% increase (i.e. to 60%) of single immigrants marrying increases the costs by £4 billion.

Family size
A family size of two children has been used. In reality, average family sizes for immigrants from developing countries are more than two. They therefore receive more Tax Credits, more Child Benefit and more Housing Benefit.

An average family size of three children would increase the lifetime costs by £14 billion.

Ages of immigrants and their children
The costing has assumed an average age of 35 for an illegal immigrant and an average age of seven for their children, inclusive of those born in the UK.

Each one year reduction in these assumptions would increase the lifetime costs by £5 billion.

Resources
The costings only include the consumption of Health Service and Education resources. In reality there are numerous other cost headings that have to be financed out of taxes. This will include such costs as Social Services and translation.

Capital costs
The capital costs, such as new schools and hospitals, that an increasing population requires, have not been taken into account.

Notes:

1. Page 104 Chart 5.5 comprises £0.4million excluding Housing and Welfare Benefits. Page 104 final paragraph then states Housing and Welfare costs of £2million a year
2. Seminar 16th June 2009, Bridget Rosewell, Chief Economic Advisor, GLA Economics
3. Page 90 ‘In the longer term, of course, the increase in migrants in recent years will add to demands for social care for the elderly’
4. Department for Work and Pensions Tax Model Tables
5. GLA Report Table 2.8 referenced to Table 2.9
6. GLA Report page 47
7. Page 12 section 42

Annex C
Comparison with the GLA costing

Costs from date of amnesty to retirement
The GLA report estimates that the net annual costs would eventually be £1.6 billion a year, comprising £2.4 billion of costs, partially offset by £0.8 billion of Income Tax and National Insurance receipts. Assuming an average age for illegal immigrants of 35 and a retirement age of 68, gives a remaining working life of 33 years, the total net costs up to retirement on the GLA basis would be £52 billion.

The Migrationwatch figures have been calculated using official Tax, Tax Credits and Benefits figures, giving a cost estimate of £77 billion, £25 billion higher than the GLA estimate. The items making up the difference are summarised below.

Tax receipts
The Migrationwatch income tax calculations include net tax receipts of £5.2 billion and £11.6 billion of Employee National Insurance contributions, giving a total of £16.8 billion. The GLA report estimates £27.9 billion and so the Migrationwatch estimates are £11.1 billion lower than the GLA estimate. The Migrationwatch figures have been calculated using official Income Tax, Tax Credits and National Insurance rates.

Employer National Insurance receipts
Migrationwatch has included Employer National Insurance contributions of £13.5 billion; The GLA report has not included this income in their figures and so in this respect the Migrationwatch figures reduce the cost of an amnesty.

Education
The Migrationwatch estimate is £17.3 billion compared with the GLA estimate of £5.2 billion, an increase of £12.1 billion. The Migrationwatch costing estimates the total cost; The GLA costing only includes the costs over and above those already being incurred for the education of illegal immigrant children.

In addition, the GLA report estimates there are currently 124,000 illegal immigrants under 16 years of age. The Migrationwatch estimate, explained in section 5 of this paper, is that existing partnerships, plus those resulting from immigrants bringing a spouse to the UK once legalised, plus marriages in the UK would bring the total number of married couples to 241,000. Migrationwatch has then made the conservative assumption that the average family size would be two children, thus resulting in an eventual total of 482,000 children.

Health and Social Care
The Migrationwatch estimate is £27.7 billion against the GLA estimate of £8.3 billion, an increase of £19.4 billion. Again, the Migrationwatch costs reflect the total costs, not just the costs over and above the Health costs already being spent on illegal immigrants.

The Migrationwatch figures also take into account that health costs increase as people age. The Migrationwatch estimates have used an annual Health cost per year person of £1,185 for an under 16 year old, £1,092 for a working age person, and £4,350 for an over 65 year old [8] The total lifetime cost per person equates to an average of £1,722 a year. This is well within the Government’s cost per person included in their 2009/10 Budget, which budgets Health spending of £119 billion for a population of 60 million, giving an average of £1,983 per person.

Income Support
The GLA report takes little account of the costs of supporting unemployed immigrants. The Migrationwatch estimates include £6.7 billion for Income Support costs alone.

Housing and Welfare Benefits
The GLA have included £66 billion (£2 billion a year for 33 years) for Housing and Welfare benefits. Migrationwatch estimates the figure to be £56.0 billion, a reduction of £10 billion. The Migrationwatch estimates have been based on official Benefits Tables and rents [9] and comprise £52.2 billion for Housing and Council Tax Benefit plus £3.8 billion for Child Benefit and Free School Meals. The calculations have been reduced by £6 billion on the assumption that 20% of families will be owner occupiers.

The Migrationwatch figures have taken no account of the subsidy costs of Social Housing.

Costs in retirement

The GLA have chosen to ignore the costs in retirement even though they have acknowledged the costs will be incurred [10].

Assuming just 7 years of retirement, from age 68 to 75, then the total costs in retirement would be £57 billion, comprising Pension Credit (£21 billion), Housing Benefit (£21 billion), and Health costs (£15 billion).

Notes:
8. These estimates have been taken from Migrationwatch Briefing paper 1.12 October 2006, uplifted for cost inflation.
9. London rents have been extracted from the Valuation Office Agency Local Housing Allowance figures. Rents outside London have been taken from the Department for Work and Pensions Tax Model Tables.
10.Seminar 16th June 2009, Bridget Rosewell, Chief Economic Advisor, GLA Economics

See Annex D
- Spreadsheet showing the details of the lifetime costs

Appendix 1
Lifetime costs for a working family unit

Appendix 2
Lifetime costs for an unemployed family unit

Appendix 3
Total lifetime cost of an amnesty

---