Net migration and the macroeconomy: inflation and output effects.

Kirby, Simon ; Riley, Rebecca

Since the last NIESR forecast in October the Government Actuary's Department (GAD) has released new projections of the UK population. (1) In comparison to previous projections these show substantial upward revisions to growth in the population of working age and to total population growth from 2005 to 2007. From 2008 and onwards there have been smaller upward revisions. Figure 1 illustrates the profile for growth in the population of working age implied by both the GAD 2003-based and the most recent GAD 2004-based mid-year projections.

[FIGURE 1 OMITTED]

The revisions to the GAD projections are almost exclusively the result of revised assumptions about net migration to the UK. According to Total International Migration estimates produced by the Office for National Statistics, net migration to the UK was exceptionally strong in 2004, rising to 223 thousand from a stable average of 160 thousand per annum in the five years before. This was a result of an increase in the inflow of people to the UK, most notably from the European Union which expanded with ten new member states on 1 May 2004. Net inward migration in 2004 was also particularly strong from the Commonwealth, with a relatively sharp offset from a reduction in net inward migration of British citizens.

The upward revision to population growth in all years of the projection follows persistence of historically high net inward migration. The sharp increase in population growth estimated for last year takes into account provisional migration data for the second half of 2004. Otherwise the relatively strong upward revision for 2005-7 is a direct result of an allowance for additional net inward migration from the accession countries to the European Union.

In comparison to previous estimates, the population of working age is expected to be 1 per cent greater than previously thought by 2012 and 2 per cent greater than previously thought by 2026. The total population rises by a little less, since migrants are typically young and of working age, such that there is a small reduction in the dependency ratio (the number of non-working age people per 100 people of working age) of 1/4 percentage points.

A population shock like the one described above affects the economy directly in two ways. First, by raising the labour force it increases the long-run supply capacity of the economy. Second, by reducing the dependency ratio it reduces the share of government spending in GDE Since the revisions to the GAD projections do not change the dependency ratio by much, this second effect has only minor consequences for the macroeconomy. The first effect is by far the most important.

Net migration may also change the equilibrium rate of unemployment. For example, if the skills of net migrants are such that they complement the skills of the existing population and relieve bottlenecks in the labour market, wage pressure will be reduced, all else being equal. This would tend to reduce the equilibrium rate of unemployment. Dustman et al. (2005) suggest that, at the national level, the skill mix of migrants to Britain is fairly similar to the skill mix of native born workers. Frijters et al. (2005) find that on average the duration of unemployment spells is greater for migrants to the UK than for people born in the UK. On the basis of the numbers reported there, one might expect an increase in the population of the nature discussed here to lead to a very minor increase in the equilibrium rate of unemployment. The level of employment would still rise.

Based on simulations using NiGEM (2) it is possible to illustrate how this additional population growth, arising from stronger net migration to the UK, may be expected to affect the macroeconomy. The NIESR forecast is based on the assumption that the population of working age and the population as a whole grow in line with GAD projections. Hence, this exercise also illustrates the approximate effects on the forecast for the UK economy that result from changes to the population data.

Table 1 shows the effects on the unemployment rate, consumer price inflation and GDP growth of a shock to the population of the magnitude implied by the revisions to the GAD projections. In these simulations net migrants are assumed to display similar labour market characteristics to those of the existing population. Thus, labour market participation is assumed to be unaffected by the population shock and the labour force rises in line with the population of working age. In addition, net migrants looking for work are assumed to exert the same pressure on wages as the existing unemployed, leaving the equilibrium rate of unemployment unaffected in these simulations.

The first set of results in table 1 are generated assuming that wages are set with reference to expected price inflation as well as past inflation and assuming that equity prices, long-term interest rates, and exchange rates are forward looking variables and expectations are rational. As shown in the table, the additional labour supply that results from the increase in the population of working age is not absorbed immediately and as a result the unemployment rate rises for a few years. This puts downward pressure on wages and two years after the initial shock to the population reduces inflation by approximately 0.2 percentage points per annum. This helps to raise demand. Real wage growth is also dampened, increasing the demand for labour. In response to weaker anticipated inflation, monetary policy is relaxed. As a result the exchange rate depreciates immediately, boosting export demand and helping to explain why inflation initially rises, albeit marginally. In the longer term GDP growth is 0.1 percentage points greater than it would have been in absence of the population shock, due to the shift in trend growth of the labour input. The increase in net inward migration also improves the medium-term budgetary position by approximately 0.2 percentage points of GDP. This outcome is a result of stronger real growth. The improvement comes about, amongst other factors, via an increase in taxes on profits, which rise as a share of the economy as real wage growth is restrained, and a reduction in government interest payments, which fall as monetary policy is relaxed.

In the second set of results shown in table 1, wages are assumed to be entirely backward looking and the nominal exchange rate is fixed. These results are shown to illustrate the importance of the exchange rate effect and the forward looking element in wage setting in raising aggregate demand and labour demand initially and in helping the economy absorb the additional labour supply quickly. In the backward looking scenario GDP growth is almost 0.1 percentage points weaker for the first two years of the shock and the rate of unemployment rises by an additional 0.1 percentage points for several years. In the longer term GDP growth is stronger as the economy eventually responds to the shock. In the very long term the level of GDP changes by the same amount in both cases.

In the simulations underlying the results in table 1 net migrants were assumed to be identical to the rest of the population in terms of their labour market characteristics. If, instead, activity rates or the skill composition differed, the effect of the population shock on the economy would be marginally different from the results shown in table 1. For example, if on average the labour market participation rate for net migrants is 10 percentage points greater than for the existing population, the initial rise in the rate of unemployment would be approximately 0.04 percentage points greater than shown in table 1. GDP growth would be on average less than 0.02 percentage points above the figures in table 1.

It is not obvious that activity rates should differ between net migrants and the existing population. Speculatively, if one of the main reasons for migration to the UK is to participate in education, one might expect the increase in net migration to reduce the aggregate activity rate. If, on the other hand, one of the main reasons for migration is economic, one might expect the average activity rate to increase. We know that migrants are relatively young. According to the Labour Force Survey, the labour market participation rate for 18-24 years olds is approximately 4 percentage points below the average for the population of working age as a whole. The labour market participation rate for 25-34 years olds, who may also be counted as relatively young, is approximately 5 percentage points above the average for the population of working age as a whole. Hence the impact on the overall participation rate will only become clear once the new flows are well established.

Another source of uncertainty in ascertaining the effect of increased net migration to the UK on the aggregate economy relates to the data itself. As discussed in the Bank of England's (2005) August Inflation Report, there are reasons to believe that the official migration data may not accurately reflect true migration to the UK. In particular, the official data do not capture the flow of migrants who come to work in the UK only temporarily.

NOTES

(1) These are available at http://www.gad.gov.uk

(2) National Institute Global Econometric Model V4.05, October 2005.

REFERENCES

Bank of England (2005), Inflation Report, August.

Dustmann, C., Fabbri, F. and Preston, I. (2005), 'The impact of immigration on the British labour market', Economic Journal, 115, F324-41.

Frijters, P., Shields, M.A. and Price, S.W. (2005), 'Job search methods and their success: a comparison of immigrants and natives in the UK', Economic Journal, 115, F359-76.

Simon Kirby and Rebecca Riley, Thanks to Ray Barrell and Martin Weale for helpful suggestions.

Table 1. Economic effects of revisions to population projections

Assuming: Forward looking wages and financial markets

 Unemployment rate Inflation GDP growth

 (absolute difference from base)

2005 0.20 0.02 0.09
2006 0.40 -0.07 0.13
2007 0.39 -0.20 0.10
2008 0.33 -0.23 0.12
2009 0.24 -0.23 0.14
2010-2014 0.06 -0.10 0.11

Assuming: Backward looking wages and fixed nominal exchange rates

 Unemployment rate Inflation GDP growth

 (absolute difference from base)

2005 0.20 0.00 0.02
2006 0.48 -0.09 0.04
2007 0.51 -0.22 0.07
2008 0.47 -0.28 0.12
2009 0.37 -0.28 0.17
2010-2014 0.08 -0.13 0.17