IMMIGRATION AND INTER-REGIONAL

MOBILITY IN THE UK, 1982–2000

Labor Market Analysis Presentation

Presenter: TAEWAN NOH

Authors: Timothy J. Hatton and Massimiliano Tani
Journal: The Economic Journal, 115 (November), F342–F358
Background: Immigration In the U.K.
Interest & Issue Increase
• Until 1992, the
UK Immigration policy
reform: Easier to immigrate scale and effect
of immigrants are
small
No Significant Effect by immigration

• After the policy

reform, the
interest about
adjustment
mechanism have
been increasing
Main Approach
• There should be adjustment mechanisms account for small
labour effects

• Inter-Regional migration make adjustment

• Spatial correlations: local or regional labour markets, within a

given country
1. Immigration Effects and Non-Immigrant Labour Supply

• Immigration -> wage and Employment: US(No significant),

Europe(negative effect)
• => Various Evidences

• Immigrant-induced labor supply shocks with labor demand

curves slope down: inconsistent

• If the supply of other factors to a locality is elastic -> the

immigration only effect the wage rates of particular groups
insofar as it alters the skill mix
1. Immigration Effects and Non-Immigrant Labour Supply

• However, recent studies for the US suggest that even immigrant-

induced changes to the skill composition have little effect on
relative wages
(Explanation)
1) Since local economies are very open, the output mix adjusts and
any skill-specific wage effects are muted by the changing
specialisation in inter-regional trade
2) Adjustment mechanism is through the labour market itself:
immigration to one locality causes inter-regional migration that
spreads the wage effects of immigration throughout the
economy
=> key issue is whether (and if so, how) the effects of immigration
on wages and/or unemployment may be masked by inter-regional
labour mobility
1. Immigration Effects and Non-Immigrant Labour Supply

• Setting out a simple perfectly competitive model of labour supply and

demand in a local labour market that includes inter-regional migration
• Regional labour market i, where immigrant and non-immigrant labour
are perfect substitutes(i.e. Indidividual characteristics are not considered)

• Growth of immigrant labour supply

• Non-immigrant labour supply

1. Immigration Effects and Non-Immigrant Labour Supply

• Using (1) and (3) we can solve for the regional wage level

• Special case where internal migration is unresponsive to the relative

wage or to the direct displacement effect of immigration, 𝛾1 = 𝛾2 =0
1. Immigration Effects and Non-Immigrant Labour Supply

• As a Instrument 𝑚 approach

1. Well-documented tendency for new immigration( join immigrants

community): 𝑚 𝑧

2. Random regional component instead of nation-wide shock(time

dummy: 𝜆𝑡 ): 𝑚[ 𝑣 − 𝑒 (𝑤𝑖,𝑡 )], but corr(𝑥, 𝑚) ≠ 0->omitting
𝑥 is cand be biased
1. Immigration Effects and Non-Immigrant Labour Supply

• As a Instrument 𝑚 approach
3. Coefficient of m: inverse of elasticity of labor demand is lager than 1
(∵ 𝐷𝑖𝑟𝑒𝑐𝑡𝑙𝑦 𝑒𝑠𝑡𝑖𝑚𝑎𝑡𝑒𝑑 𝛼 < 1)
=> In the regional level, labor demand can be highly elastic
(∵ 𝐺𝑜𝑜𝑑𝑠 𝑚𝑎𝑟𝑘𝑒𝑡 𝑖𝑠 ℎ𝑖𝑔ℎ𝑙𝑦 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑒𝑑, 𝑜𝑡ℎ𝑒𝑟 𝑓𝑎𝑐𝑡𝑜𝑟𝑠 𝑒. 𝑔. 𝐶𝑎𝑝𝑖𝑡𝑎𝑙 𝑎𝑟𝑒 𝑚𝑜𝑏𝑖𝑙𝑒)
=> Immigration: Increase of labor supply and demand
(Cobb-clark(1999): demand effects of immigration will overweigh the labor supply effects
on the job prospects of residents)
-> coefficient of m should be more negative when labor demand is added in
the equation (∵the level of coefficient of △ 𝑥 = coefficient of 𝑚)
1. Immigration Effects and Non-Immigrant Labour Supply

• As a Instrument 𝑚 approach
4. If equation (4) is the correct specification: coefficient of 𝑚 can be downward
biased estimate on 𝛼
5. Wage effect of immigration depends on both direct(𝛾2 ) and indirect effect:
If 𝛾2 → 1, 𝑊𝑎𝑔𝑒 𝑒𝑓𝑓𝑒𝑐𝑡 → 0
(when the crowding effet is high, wage effect close to 0 even in inelastic labor demand
1. Immigration Effects and Non-Immigrant Labour Supply

• Using (1) and (3) for the change in non-immigrant labour supply

If 𝛾1 𝑎𝑛𝑑 / 𝑜𝑟 𝛾2 > 0, coefficient of 𝑚 < 0(otherwise 0): n and 𝑚 is are not

independent
(if 𝛼 is large then the effect of m close to 0 even 𝛾1 is substantially larger than 0)
1. Immigration Effects and Non-Immigrant Labour Supply

• Two other cosiderations

1. For simplicity, in the framework set out above, the market clears and there
is no unemployment. However a broadly analogous system could be set
out where wages do not adjust and where immigration and internal
migration are determined by inter-regional differences in unemployment
rates(But if immigrants compete less effectively in the search for jobs as
suggested by Frijters et al. then the effect on the mobility of non-
immigrants would be muted)
2. finding negligible effects of immigration on internal migration, not
controlling for employment shocks, could be consistent with the results of
studies that find interregional migration is large in response to shifts in
labour demand
2. Immigration and Internal Migration in Britain

• The most important areas are the New Commonwealth, particularly the
Indian subcontinent, Old Commonwealth countries (including South
Africa), and the US

• Studies of these immigration flows show that they are determined by

economic conditions at home and abroad, operating through the filter
of immigration policies. The pattern of flows to and from different parts
of the world is also conditioned by the source country composition of
the existing stock of foreign-born residents in the UK (Hatton, 2005)

• While gross inflows are sometimes sizeable, especially if cumulated over

a number of years, the annual average net inflow is typically fairly small
2. Immigration and Internal Migration in Britain

Averaged
nealy 1.7 per
thousand
2. Immigration and Internal Migration in Britain

• With a few exceptions, those from the EU and Other Foreigncountries have regional concentrations
similar to the foreign-born as a whole
 2. Immigration and Internal Migration in Britain

• measure of the relative skills of British and foreignemployees by region, averaged over the
years 1992–7

> Few foreigners Area(Table 2)

<

• Overall, it appears that compositional differences between natives and foreigners are not large enough to
have very decisive effects at the regional level
• While it would be interesting to compare the skill mix of the immigrant inflow with that of existing
residents, unfortunately it is not possible to observe relative skills in the immigration flow data by region
2. Immigration and Internal Migration in Britain
• Net flow of migrants from abroad compare with the flow of migrants within Britain
• These include all individuals regardless of nationality, place of birth or labour force status, and they
represent movements within the UK including Northern Ireland

• suggest that net immigration to London from abroad has been associated with a displacement
of population from London to other regions(a strong inverse correlation between immigration
and regional in-migration across the other regions)
3. Explaining Inter-Regional Migration Flows

𝑛𝑖𝑗,𝑡 : Net inter-regional migration rate from region j to region i at time t

𝑥𝑖,𝑡 − 𝑥𝑗,𝑡 : a vector of variables determining migration from j to i
𝑚𝑖,𝑡−1 − 𝑚𝑗,𝑡−1 : the inflow rate of foreign immigrants to region i – rate of region j

- period for which all these series could be assembled is the period from 1981 to 2000, which
(allowing for one lag), gives a panel of 1,045 observations (19 years 55 region pairs)

- In initial estimation we instrumented the net immigration rates, using their lagged values(When
instrumental variables are used the negative coefficients on immigration were always larger than
in the OLS regressions in Tables 5–7, although the standard errors were also larger)
• Potential concerns about the
endogeneity of the other
explanatory variables and we
experimented with lags of these
too
3. Explaining Inter-Regional Migration Flows

• labour market and housing

variables are added in
column (2), with the effect
of increasing the size and
significance of the
excludes all
immigration variable.

<
other variables
except the
region pair and • This suggests that it is
year dummies
important to control for
these additional variables
and that in their absence
the coefficient on
immigration is biased
downwards(Q)
3. Explaining Inter-Regional Migration Flows

• gross flows of international

migration tend to be larger
relative to population in the
southern regions, particularly
London – a fact that is also
reflected in population stocks.
• We might expect that where
immigration flows are larger
their effects would be more
clearly discernible

< • immigration effects are larger in

the Southeast, but it is only
significant when the other
variables are included, as
column (4) shows
• While the coefficients are
expected to be smaller for gross
as compared to net inter-
regional flows, there also seems
to be greater heterogeneity
3. Explaining Inter-Regional Migration Flows

• Coefficient suggests that an

increase of 100 in net
immigration to a region from
abroad generates a net out-
migration to other regions of
about 35.
• Thus, the displacement effect
could be substantial
• This is consistent with the
findings for the bilateral flows in
Table 5 and it suggests that,
where immigration is sufficiently
large, the effects on inter-
regional migration can be more
clearly discerned.
4. Conclusion

• To understanding why the local labour market effects of

immigration seem to be relatively benign
• The evidence indicates consistently negative correlations
between immigration to one region from abroad and in-
migration from other regions. But they are only significant for
the southern regions where immigration of foreign citizens is
most concentrated
• Even though the effects of immigration on unemployment
and earnings in one particular region may be small, this may
be a poor guide to their effects on the whole economy
Appendix