House of Lords - Economic Affairs

Select Committee on Economic Affairs Minutes of Evidence

Supplementary memorandum by Professor Robert Rowthorn, Cambridge University

EXECUTIVE SUMMARY

The government submission contains a great deal of valuable material, and some parts of it contain a balanced treatment of their subject. Other parts are very one-sided. The following is a summary of my comments. They are mostly critical, because I focus exclusively on weaknesses in the government submission.

1. The government submission fails to acknowledge of the scale of what is now happening. Population is projected to rise by 25 million from 60.6 million in 2006 to 85.4 million in 2081. This is equivalent to1 million people, or the population of Birmingham, every three years. Over the long-term, the projected increase is entirely the result of migration. Indeed, in the complete absence of inward or outward migration, projected population would be lower in 2081 than it is today. Quite apart from its cultural and political implications, the influx of such a large number of diverse people may impose costs on the local population which are ignored in the government submission. These include the cost of integrating the new-comers and regulating an increasingly diverse society, plus the impact on housing and open space, urban sprawl and congestion.

2. The government cites estimates of the fiscal surplus generated by migrants as a whole. These estimates are extremely small in relation to the national economy—around 0.2% of GDP, 0.4% of personal consumption and 0.6% of government expenditure. They also exclude much of the extra public expenditure specifically associated with immigration. If such extra expenditure were included, the net fiscal contribution of migrants as a whole might become negative, although not by very much.

3. Estimates of the overall fiscal contribution of migrants are not much use for policy-making. What matters is the contribution of particular types of migrant. Some generate a surplus for the exchequer, whereas others are a net cost. The submission provides no estimates regarding the likely fiscal contribution of different types of migrant, in particular of unskilled migrants. There is a reference to a 10-year old American study on this subject which finds that unskilled migrants pay almost as much tax as they receive in government expenditure. The submission does not cite later research by a leading expert in the field which contradicts this finding. Besides, American experience is of limited relevance to the UK because the USA has a much smaller welfare state than ours and net public expenditure on people with low income is much less than here. The submission does not refer to the various European studies on the fiscal impact of immigration.

4. The submission claims that the UK faces a serious ageing problem. Its observations on this issue have been partly overtaken by events. New projections by the Government Actuary's Department (GAD) indicate that the population is growing faster and ageing more slowly than was previously thought. It was previously projected that, in the complete absence of migration, the dependency ratio would rise to 82.4% in 2056. Under the new projection the figure is 73.6%. The birth rate has been rising in recent years and if this trend continues it will ameliorate the ageing problem still further.

5. The submission argues that large-scale immigration is required to stave off rapid ageing of the population. My own calculations, which are based on the latest official projections, indicate that this claim is exaggerated. I consider a scenario in which long-run migration is balanced. By this I mean that the number of people entering the country is equal to the number leaving. This rejuvenates the country because those who enter are on average younger than those who leave. It also boosts population a bit. With balanced migration there would be a modest rise in population and the ageing process would be greatly slowed. The dependency ratio would still increase, but the rise would be comparatively small—from 61% today to 68% in 2056.

6. Higher rates of immigration can lower the dependency ratio still further, but they lead to a large increase in population. For example, to reduce the dependency ratio by 1 extra percentage point through migration adds an extra 80,000 to the population every year. By 2081 this amounts to almost six million people. The government submission does not pay much attention to the trade-off between population growth and rejuvenation.

7. The submission cites estimates of the contribution of migrants to GDP. However, it does not compare these with the impact of immigration on population. The National Institute estimates that people who arrived in the UK in 1998 and after added 3.1% to GDP, 3.8% to total population and 5% to the population of working age. If these figures are to be believed, immigration over this period led to a fall in GDP per capita. The government submission correctly argues that the output figure is probably an underestimate, and that the productivity of migrants will increase over time. Even so, it should have presented the crude figures for per capita GDP growth.

8. The government submission is unduly upbeat about the impact of immigration on the labour market. It argues that the economy will eventually adapt to the arrival of immigrants, leaving local workers no worse off "in the long-run" than before. This may be true, but it leaves open the question of adjustment speed. A given burst of immigration may have no permanent effect, but it may take a number of years before the economy adapts fully to the new influx. During the transitional period certain types of local worker may suffer. The submission has little to say on this question, except for the occasional remark praising the flexibility of the UK economy.

9. The submission is also very upbeat in its assessment of the literature on labour market outcomes. It claims that "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". This is misleading. It is true that most research in this area does find that the labour market impact is small, but there are a number of articles by respected authors which find just the opposite. For example, a recent OECD study on the advanced economies finds that a one-off increase in the share of immigrants in the labour force may have a substantial impact on unemployment, lasting up to five or ten years. The share of immigrants in the UK is likely to rise for some decades. If the findings of the OECD study apply to this country, future immigration will have a prolonged impact on unemployment. The OECD findings may be wrong. Or perhaps they do not apply fully to this country because our economy is more flexible and better able to absorb immigrants than most other economies. Even so, it is cavalier to dismiss concerns about unemployment with bland assurances that everything will be fine in the long-run. Such assurances may ring hollow if economic growth in this country falters.

10. The government submission gives the impression that research specifically on the UK is unanimous in finding that the immigration has only a minor impact on the local labour market. This is an exaggeration. There are only a small number of studies on the UK, and the majority of them conclude that the labour market impact of immigration has been small. However, there are some exceptions which find quite large effects. There is also a DWP study which is cited at length by the government. In its more sophisticated econometric formulations, this study finds that recent immigration from Eastern Europe has had a large, but not statistically significant impact on local unemployment. These estimates are unreliable and cannot be taken as firm evidence that the impact of immigration has been large. But neither do they support the authors' claim that the impact is "small" or "not discernable".

Final Word

After reading the government submission, my general views are unchanged. Britain needs immigration. Our country would stagnate without it and immigrants have made an important economic (and cultural) contribution that is hard to quantify. However, from our own national point of view, the projected rate of immigration is excessive. It will give rise to population growth on a scale that most people consider to be undesirable, and it is of little or no economic benefit to the local population as a whole. Certain types of migration may also be harmful to more vulnerable sections of the local workforce, especially if economic growth falters. Most research on the UK suggests that this effect will be small, but I believe there is still cause for concern.

However, this is not the end of the story. There are plenty of arguments in favour of immigration which have nothing to do with its impact on the native population. One of them is that immigration, whether or not it improves our lives, does usually improve the lives of immigrants. It may also benefit the countries from which they come. There is a strong case for believing that we have a duty to share the economic benefits of our society with people from poorer countries, who sometimes risk their lives to get here.

Commentary

The section headings in this commentary refer to section headings in the cross-departmental submission.

SECTION 2: PUBLIC FINANCE AND NET FISCAL IMPACTS

This section of the government submission looks at the impact of migration on public finances. It relies heavily on a Home Office Study on this topic and an update by the IPPR.[37] These studies estimate that immigration over the past few decades has generated a net fiscal surplus. This claim is hard to evaluate. The answer depends largely on what types of government expenditure are assigned to the migrant population and what adjustments are made to correct for the impact of the economic cycle on government tax revenue. In my own submission, I reworked the IPPR estimates a bit and came up with a net fiscal surplus for migrants of £2.6 billion a year, which is slightly larger than the original Home Office estimate. However, there are other adjustments that would make the picture less favourable to migrants. For example, in his submission, Professor Coleman provides a list of costs associated with immigration which between them involve a public expenditure of almost £9 billion a year. If only a third of these costs were included, the fiscal surplus apparently generated by migrants would become negative.

The government submission correctly points out that there are no published estimates of the long-run fiscal contribution of immigrants to the UK. It then refers to a 1997 study for the United States which estimates that a skilled immigrant typically makes a large net fiscal contribution, whereas the typical unskilled immigrant makes a small negative contribution.[38] The submission does not refer to a paper by a leading expert on this subject, Storesletten, who obtains an estimate of $7,400 for the net present value of the average immigrant into the United States. This average conceals a wide variation across different kinds of immigrants. The net present values for representative high-, medium-, and low-skilled legal immigrants are found to be $96,000, -$2,000, and -$36,000.[39] The large deficit for low-skilled immigrants contrasts with the very small figure cited in the government submission. None of this is of great relevance to the UK, since the United States has a much smaller welfare state than we do, and net public expenditure on low-income people is much less than here. There is no reference in the submission to academic work on the fiscal impact of immigration in European countries. In fact, there are quite a lot of studies on this topic, some of which find that the net fiscal contribution of immigrants as a group is negative, whereas others find it to be positive.[40] Either way, the impact is typically small in relation to GDP or total government expenditure. The failure of the government submission to take proper account of evidence from other countries is a serious failing. It is especially serious given the importance which the government assigns to fiscal benefits in its case for large-scale net migration.

The government submission makes the following claim,

"In the long run, it is likely that the net fiscal contribution of an immigrant will be greater than that of a non-immigrant. For migrants of working age who enter the country this is relatively clear; the UK is receiving the fiscal contribution of their work without paying for the education and training that enables them to work." (para. 2.2.6)

This statement ignores the fact that immigration may give rise to a variety of extra public expenditures over and above what is spent on the average native. It also ignores the fact that the employment rate of immigrants is, on average, lower than that of natives in the same educational category. When these factors are taken into account, it is by no means clear that the average future immigrant will generate a fiscal surplus. More fundamentally, as far as policy is concerned, it is not the fiscal contribution of the average immigrant that matters. No-one is calling for a complete end to immigration. What matters for policy is the fiscal contribution of different kinds of immigrant. If mass immigration means the permanent settlement of a large number of unskilled workers or people who do not work, this is likely to impose a net fiscal cost on the local population. It is true that the UK does not have to pay for the education of such people, but this is not the issue. What matters is the net cost that the exchequer will incur during their time in this country. In the case of permanent immigrants of the unskilled or non-working variety, the total tax they pay during their life here is likely to be less than what they receive in the form of public expenditure. The situation is quite the opposite for highly skilled, employed immigrants who will normally make a large net fiscal contribution.

The government submission takes particular exception to the MigrationWatch claim that 50% of public expenditure on children of mixed parentage (one immigrant and one native) should be ascribed to the migrant population. The government's argument is that

"Children born in the UK are UK citizens and it is inconsistent to view them as `part migrant' before the age of 16, but UK nationals after this age"(para 2.2.7)

This is a very weak argument. A similar point could be made about children who are UK citizens and have two immigrant parents. Such children are regarded under the Home Office methodology as `100% migrant' until they are 16 and UK nationals after this age. Public expenditure on such children is ascribed 100% to the migrant population. The same is true for children with a lone migrant parent.

There is a deeper problem with the Home Office methodology. It was designed for a specific purpose which is not of great relevance to the task of formulating migration policy for the future. The original purpose of the Home Office study was to combat the notion that immigrants as a group are a serious drain on the taxpayer. This notion is false and the Home Office and the IPPR have played a valuable role in refuting it. However, their estimates of the overall fiscal impact of immigration are not very helpful as a guide to future policy. In particular, they do not support blanket claims of the "immigration is good for us" variety. As these studies point out, some types of immigrant make a positive fiscal contribution, whilst for others the net contribution is negative. If the sole aim of policy were to maximise the contribution of migrants to the public purse, the logical step would be to encourage the immigration of people who are likely to be net contributors and exclude those who are not. Many would consider this to be an unethical approach, but it makes sense from a fiscal point of view.

SECTION 3: OUTPUT AND GDP PER CAPITA

This section of the government submission is devoted to the impact of migration on key macroeconomic indicators. It lists the following key points:

1. Migration affects trend growth principally through changes in the working age population;

2. Work by the National Institute of Economic and Social Research suggests that around 17% cent of economic growth in 2004 and 2005 is attributable to immigration;

3. The Treasury estimates that between Q3 2001 and mid-2006 migration added 0.5% per annum to the working age population and therefore supported growth in economic output. On this basis, migration contributed around £6 billion to output growth in 2006;

4. There is no quantitative evidence available on the impact of immigration on GDP per head. Wage data suggest migrants may have a positive impact directly through their own output and indirectly through raising the productivity of others.

Comment

I have no quarrel with points (1) and (2). Points (3) and (4) are questionable.

Point (3) apparently rests on the assumption that the contribution of new immigrants to GDP is strictly proportional to their share of the working-age population. Thus, if new immigrants add 5% to the working-age population they are assumed to add 5% to total GDP. This is hard to square with evidence presented in the National Institute article cited by the government in point (2) above. This article estimates that immigrants arriving in the country in 1998 or after added 5% to the working-age population but only 3.1% to GDP.[41]

Point (4) asserts that there is no quantitative evidence available on the impact of immigration on GDP per head. In fact, the National Institute article cited in the government submission provides evidence on precisely this point. It estimates that that people who arrived in the country in 1998 and after added 3.1% to GDP and 3.8% to total population. The government also claims that immigration in 2004-05 was responsible for a 0.9% increase in GDP. Official statistics indicate that the net inflow of non-British migrants during this period was equal to 1.1% of the population. If these figures are to be believed, the effect of immigration was to reduce GDP per capita by 0.7% over the longer period from 1998 onwards and 0.2% over the shorter period 2004-05. These are quantitative estimates which the government submission could have easily provided.

The government submission argues, with justification, that the contribution of immigrants to output is larger than the above estimates imply. Immigrant workers complement the skills of the existing workforce and make the latter more productive. They also help to increase profits. In the course of time, immigrants may also become more productive as they become more integrated into the economy and firms invest in extra capital to make full use of them. As a result, the present and future impacts of immigration on national output are likely to be greater than estimates based simply on their current wages. The question is how much greater, and what implications does this have for GDP per capita? The answer to the second question depends on what kind of immigrant arrives in the future and how effectively immigrants are integrated into the labour market. An influx of young, highly educated immigrants who quickly find useful employment will normally raise GDP per capita. An influx of unskilled workers or of people who do not get a job will normally reduce GDP per capita.

SECTION 4: THE LABOUR MARKET

This section of the government submission is devoted to the impact of immigration on the labour market. It contains a number of questionable assertions.

Theory

Paragraph 4.2.2 begins with the following statement:

"There is no theoretical reason why immigration need depress either native wages or increase native unemployment"

This statement is misleading. It is true that there are conditions under which immigration should theoretically have no effect on native wages and unemployment. However, there are also conditions under which just the opposite is true. There is also the question of adjustment speed. A given burst of immigration may have no long run effect, but it may take a number of years before the economy fully adapts to the new influx of workers. During the transitional period some native workers may suffer a significant loss. The submission gives several quotations to the effect that in the long-run immigration has no effect on labour market outcomes. Even if this were always true, which is doubtful, the question still remains as to how long is the long run? As Keynes once said, "In the long run we are all dead". The submission has little to say on this question, except for the occasional remark praising the flexibility of the UK economy. Such optimism does not square with the rather gloomy findings of a recent OECD working paper with regard to the scale and duration of the unemployment resulting from immigration. The authors' exact words are as follows:

At the aggregate level, however, the transitory impact may be substantial; its magnitude and duration largely depends on the persistence of unemployment shocks, and it may last between five and ten years

Five to 10 years is a long time and it refers to a once and for all rise in the share of immigrants in the national labour force. The share of immigrants in this country is currently rising and is likely to rise for some decades. If the UK were a typical OECD economy, the estimates of Jean and Jiménez would imply that there will be a prolonged and substantial rise in native unemployment because of immigration. In fact, the UK has a relatively flexible labour market, so the effects supposedly identified by these authors should be smaller and less durable than those of the typical OECD economy. Even so, it is cavalier to dismiss concerns about unemployment with bland assurances that everything will be fine in the long-run.

Evidence

The government submission summarises the evidence regarding immigration and native workers by quoting the following passage:

et al

The first part of this passage does not accord with my reading of the literature. It is true that most studies find that the impact of immigration on native workers is small, but there are also studies by highly respected researchers who find that its impact is large. In addition to the paper by Jean and Jiménez cited above, there has been a major paper in the Economic Journal by Angrist and Kugler on unemployment in the European Union. These authors find, in some of their formulations, that immigration has a large and statistically significant effect on male employment. For each 100 male immigrants, they estimate that between 35 and 83 male native jobs will be lost.[44] On the subject of wages, a succession of articles by Borjas, one of the world's leading migration economists, find that the immigration of unskilled workers into the USA, especially from Mexico, has seriously harmed native unskilled workers, especially blacks.[45]

In the case of the UK, there is not a great deal of evidence, but even here the conclusions are somewhat mixed. The government submission makes the following claim with regard to research on the UK:

"A number of papers have looked at the employment impacts of migration; none appears to have found a statistically significant impact" (para. 4.3.1)

This is false. In a study for the Home Office, using Census data for the UK and the "difference in differences" method, Dustmann and his colleagues estimate that 23-60 native jobs are lost for each 100 immigrants.[46] These are big numbers and the coefficients on which they are based are statistically significant. Using LFS data, the same study finds a smaller and less statistically significant effect. This is also the case in a later paper of these authors based on the LFS.[47] It is interesting to note that Hatton and Tani, in their study of internal migration within the UK, warn that their results imply that the method used by Dustmann et al may underestimate the impact of immigration on employment.[48]

The government submission cites an article in the National Institute Review to support its claim that around 17% of economic growth in 2004 and 2005 was due to immigration. It fails to mention that the same article estimates that immigration over this period also led to an increase of 0.2-0.3 percentage points in the unemployment rate.[49]

There have been several studies that have sought to quantify the effect of immigration on wages in UK. They find that, taken as a whole, immigration has had a minor effect on wages. The government submission cites a recent paper by Dustmann and colleagues which estimates that immigration has led to a small increase in the average wage and a small reduction in the wage of the bottom quarter of the labour force.[50] I am not surprised by this result since it is concerned with the combined effect of all types of immigration on wages. Economic theory suggests that different types of immigration affect different types of worker in different ways. Many of the immigrants into the UK have been highly skilled and their entry has helped to create jobs and higher wages for local unskilled workers. There have also been many immigrants seeking unskilled jobs. Such people may compete with local unskilled workers, thereby reducing employment opportunities and wages for the latter. Thus, local unskilled workers have gained from some types of immigration and lost from others. It is not surprising that the overall impact of immigration on local unskilled workers has been small.

Of particular interest to public policy is the impact of unskilled immigrants on the wages and employment of their local counterparts. This was examined in a paper by Manacorda et al, which is not cited in the government submission.[51] They found that unskilled immigration does harm the local unskilled workforce, but its effects are confined mainly to previous immigrants. This is because, in the unskilled part of the labour market, recent immigrants frequently enter occupations where there are already a lot of immigrant workers. Many of these previous immigrants belong to ethnic minorities and many of them are now naturalised British citizens.

The government submission lays great weight on a recent DWP study by Gilpin and colleagues.[52] This study gets statistically insignificant, but sometimes very large estimates for the impact of recent immigration from Central and Eastern Europe on unemployment in the UK. Most of their "long-run" coefficients are equal to at least 0.6, which implies that in the "long-run" 60 or more local workers will become unemployed for each 100 immigrants that enter the region. It must be stressed that these coefficients are not statistically significant, but this does not mean that they are "small" as the authors claim. On the contrary, many of them are large. Nor does it mean that there is no "discernable" evidence that recent immigration has caused unemployment, as the authors also claim in a passage cited in the government submission. It simply means that there is too much noise in the system or too many confounding factors to permit reliable estimation.

The papers cited above do not prove conclusively that the effects of immigration on employment and wages are either large or long-lasting. However, they are sufficient to undermine the claim that there is a virtual consensus amongst experts that immigration has only minor or very short-lived impacts on the labour market. My own views on this subject are as follows. I do not believe that the impact of immigration is anything like as big as some of the above estimates imply. Nor are most of its effects permanent. The economy will eventually adapt to absorb most of the new inflow of workers without significant harm to the local workforce. However, I also believe that this adaptation may sometimes be rather slow, and in the meantime immigration on the scale now envisaged may be harmful to sections of the national population, mostly those at the lower end of the economic spectrum. I am not sure about the potential scale of this effect, but the possibility should be taken seriously. It could become important if there is a downturn in the UK economy. Fears on this score are virtually brushed aside in the government submission.

Complements or Competitors?

The government submssion mentions the possibility that immigrants may complement the labour of native workers, thereby increasing the productivity of the latter. There is also a recent literature suggesting that migrant labour has a complementary rather than competitive effect on the labour market.[53] This implies that there is less competition between migrants and natives in the same skill category than would at first appear. Instead of engaging in head to head competition, they tend to sort themselves out into separate groups performing somewhat different kinds of work. This reduces competition between them, which in turn reduces the impact of immigration on the wages and employment of natives in the same skill category. On re-reading the literature while preparing this commentary, I now think that the complementarity effect may be more important than I had previously thought. This does not mean that unskilled workers in the UK have nothing to fear from a large influx of immigrants seeking unskilled jobs. But it does suggest that they may have less to lose than would at first sight appear.

SECTION 6: DEMOGRAPHIC CONSEQUENCES OF IMMIGRATION

The government submission has quite a lot to say about the impact of immigration on the age structure of the UK population. It says very little about the implications of immigration for population growth. In fact, apart from a couple of brief asides, it says nothing at all on the subject. This is unfortunate, since age structure and population growth are closely linked.

As the government points out, immigration helps to rejuvenate the population. This is because immigrants are on average younger than the local population. They are also on average younger than the people who emigrate. Immigration also adds to the population. It does so directly because of the inflow of people and also indirectly because many immigrants will have children after they arrive in the UK and their children will have children, and so on. Projections by the Government Actuary's Department (GAD) imply that for each 1 million immigrants who arrive, there will be an eventual addition to population equal to approximately 1.5 million.

Projections by GAD indicate that without migration the UK population would rise by a modest amount for the next few decades and then start to fall. By 2056, the population would be virtually the same as it is now and by 2081, it would be somewhat lower. With migration at the rate assumed by GAD in its Principal Projection, there will be rapid and sustained growth in population. Thus, over the longer term, projected population growth is entirely the result of migration. To the extent that rapid population growth is seen as undesirable, the resulting costs must be weighed against the presumed benefits of rejuvenation due to from migration.

The age structure of the population is often measured by the dependency ratio. This is the total number of children plus persons of state pension age divided by the population of working age. The dependency ratio is expressed as a percentage in the government submission and information is provided on the behaviour of this ratio under a variety of scenarios. This is based on old 2004-based GAD projections. Table 1 presents information based on the more recent 2006-based projections.

Before discussing table 1, a point on terminology is in order. In the government submission the terms "zero migration" and "zero net migration" are used interchangeably. This is not in accordance with normal practice. Conventionally, the term "zero migration" is used to denote a situation in which there is no inward or outward migration at all. Thus, all changes in the size and age structure of the population are the result of births and deaths amongst the initial population and their descendents. This is also known as "natural change". The term "zero net migration" is normally used to denote a situation in which the number of inward migrants is equal to the number of outward migrants. To avoid confusion, I shall avoid the terms "zero migration" and "zero net migration". Instead, I shall use the term "no migration" to indicate that there is no migration at all into or out of the UK, and the term "balanced migration" to indicate that the number of people entering the UK is equal to the number who are leaving.

Table 1 contains four GAD projections together with an additional projection which refers to the case of balanced migration. I have constructed this projection myself by extrapolating from published projections of GAD.[54] Under this projection, it is assumed that long-run net migration is zero.

The main points to note are as follows:

— The projected rise in the dependency ratio is much smaller under the new 2006-based projections than under the old 2004-based projections used in the government submission. For example, in the complete absence of migration, it was previously projected that the dependency ratio would rise to 82.4% in 2056. Under the new projection the figure is 73.6%. This big reduction is explained by the higher birth rate assumed in the new projection. The ageing problem has not disappeared, but it is much less serious than was previously thought. If the birth rate continues to increase, the importance of the ageing problem will decline still further.

— Balanced migration has a modest impact on population, but a large impact on age structure. In the complete absence of migration, population would eventually fall to 57.3 million by 2081 and the dependency ratio would rise to 77.3%. The corresponding figures under the Balanced Migration projection are 64.3 million and 71.3%. Thus, moving from no migration at all to balanced migration reduces the dependency ratio by 6.0 percentage points and adds 7.0 million to the national population in 2081. This works out at an extra 1.2 million people for each one percentage point reduction in the dependency ratio. The big impact on age structure is due to the fact that those entering the country are on average younger than those leaving. Information for other time periods is shown in table 2.

— Higher rates of immigration can lower the dependency rate still further, but they have a large impact on population. Under the Balanced Migration projection, the dependency ratio rises to 71.3% and population to 64.3 million by 2081. Under the Principal projection, the corresponding figures are 67.7% and 85.3 million. Comparing these two scenarios, the extra immigration required to reduce the dependency ratio by 3.6 percentage points in 2081 adds an extra 21 million to national population by the end of the period. This works out at 5.7 million extra persons for each one percentage point reduction in the dependency ratio. It is equivalent to an extra 80,000 people every year for the next 75 years. Information for other time periods is shown in table 2. Note that the longer the time period, the more is the eventual addition to population associated with a given reduction in the dependency ratio.

Once net migration becomes positive, further increases in immigration generate relatively small changes in the dependency ratio. Other things being equal, these changes are beneficial to government finances because there are proportionately fewer people to be supported from taxes levied on the working age population. However, other things may not be equal. Some immigrants may not be able to find employment or they may remain at home because of domestic responsibilities. If they do get a job, their wages may be low and during the course of their life-time in the UK the government may spend more on them in the form of welfare benefits and public services than they pay in taxes. Other types of immigrant may be large net contributors to the exchequer. The overall contribution of immigrants will depend on the exact mixture of these various types. From a fiscal point of view, there is no benefit in admitting a large number of unskilled or non-working immigrants simply because they are young.

A policy of restricting unskilled immigration whilst encouraging skilled immigration is what a number of countries, such as Australia and Canada, already do, and what the EU is now proposing under its Blue Card scheme. It is also the intention of the new UK points-based system. From a purely selfish point of view, this may be a desirable policy, but it is ethically dubious since it may involve depriving poorer countries of their most talented and skilled people. Indeed, this is what is currently happening on a small scale in some of the new EU member states.

Table 1

PROJECTED CHANGES IN UK POPULATION AND AGE STRUCTURE


	Population (Millions)
Projection	Net migration	2006	2031	2056	2081	Change 2006-81

No Migration (Natural Change)	0	60.6	63.8	61.5	57.3	-3.3
Balanced Migration	0	60.6	65.1	65.2	64.3	3.7
Low Migration	130,000	60.6	69.2	74.3	78.6	18.1
Principal Projection	180,000	60.6	71.1	78.6	85.3	24.7
High Migration	250,000	60.6	73.0	82.8	91.9	31.3

	Dependency Ratio (Percent)
Projection	Net migration	2006	2031	2056	2081	Change 2006-81

No Migration (Natural Change)	0	60.7	68.2	73.6	77.3	16.7
Balanced Migration	0	60.7	66.5	68.2	71.3	10.7
Low Migration	130,000	60.7	64.6	65.1	68.6	7.9
Principal Projection	180,000	60.7	63.9	63.9	67.7	7.0
High Migration	250,000	60.7	63.1	62.9	66.9	6.2

Source: GAD except for the Balanced Migration projection, which is estimated by extrapolation from the published GAD projections. The No Migration projection assumes that there is no migration at all. The Balanced Migration projection assumes that the number of migrants entering the UK is equal to the number of migrants leaving the UK. The dependency ratio is the number of children plus pensioners per person of working age expressed as a percentage. The figures shown assume that the pension age increases in accordance with changes already announced.

Table 2

REDUCING THE DEPENDENCY RATIO THROUGH MIGRATION THE MARGINAL IMPACT ON POPULATION


	2006-31	2006-56	2006-81

Zero Migration to Balanced Migration	0.8 million	0.7 million	1.2 million
Balanced Migration to Principal projection	2.2 million	3.1 million	5.7 million

This table shows much population will increase for each one percentage point reduction in the dependency ratio. It is derived by from pair-wise comparisons of the projections shown.

1 November 2007

37 Ceri Gott and Karl Johnston ( 2003), The Migrant Population in the UK: Fiscal Effects, Home Office, RDS Occasional Paper No 77. Back

38 Smith, J and B Edmonton (eds) (1997), The New Americans: Economic, Demographic and Fiscal Effects of Immigration, Washington DC: National Research Council; National Academy Press. Back

39 Storesletten, K 2000. "Sustaining fiscal policy through immigration," Journal of Political Economy 108(2): 300-324. Back

40 For a survey of this literature see: David Coleman and Robert Rowthorn, "The Economic Effects of Immigration into the United Kingdom", Population and Development Review, 30(4): 579-624 December 2004). Back

41 Riley, R and M Weale 2006, "Commentary: Immigration and Its Effects", National Institute Economic Review No 198, October, pp 4-9. Back

42 Jean, S. and M Jiménez (2007), "The unemployment impact of immigration in OECD countries", Economics Department Working Paper #563, OECD, Paris. Back

43 Blanchflower, D, Saleheen, J and C Shadforth (2007), "The Impact of Recent Migration from Eastern Europe on the UK Economy", Bank of Enlgand. Available at http://www.bankofengland.co.uk/publications/speeches/2007/speech297.pdf Back

44 Angrist, J D and A D Kugler (2003), "Protective or Counter-Productive? Labour Market Institutions and the Effect of Immigration on EU Natives", Economic Journal, vol 113, no 488, June, pp F318, F322. Back

45 See for example, Borjas, G J (2003), "The Labour Demand Curve Is Downward Sloping: Re-examining the Impact of Immigration on the Labor Market", Quarterly Journal of Economics, November, pp 1335-1374. Back

46 These numbers are derived from the coefficients given in Table 4.1 of Dustmann, C, F Fabbri, I Preston and J Wadsworth (2003), "The local labour market effects of immigration in the UK", Home Office Online Report 06/03. Back

47 Dustmann, C, F Fabbri and I Preston (2005), "The Impact of Immigration on the British Labour Market", Economic Journal, vol 115, November. Back