Immigration

1
Immigration
2
General Questions

• The type of questions that economists have been
  interested in are
• what is the impact of immigration on the
  receiving/sending country ?
• how does the impact vary across groups?
• What determines attitudes to immigration?
• What should public policy on immigration be?
• More interest in it currently because of rise of
  immigration in many OECD countries

3
Fraction Foreign-Born
4
Fraction Foreign-Born continued
5
Comment

• Perhaps not what you expected
• Reason is probably fast growth in immigrant share
  in some countries
• Note that skill mix varies from country to
  country so hard to generalise.

6
Models of Impact of Immigration

• Simplest model homogeneous labour, immigration
  raises supply of labour in the economy YF(K,N)
• Basic idea is immigration surplus natives gain
  from immigration
• Size is small
• Probably even smaller in LR with CRS
• Distributional effects may be more important
  workers lose, capitalists gain

7
The Immigration Surplus
8
Heterogeneous Labour

• Skilled and unskilled labour
• Can write production function as
• Wages will be

9
Implications

• Wages only depend on relative supplies
• Implies immigration only affects natives if
  alters skill mix
• Will always be gains to natives if immigration
  affects skill mix
• Largest if immigrants very different

10
The Immigration Surplus- Heterogeneous Labour
11
Other Issues

• external effects
• taxes and the welfare state this affects net
  benefits of immigrants even such an ardent free
  marketer as Milton Friedman says that You cannot
  simultaneously have free immigration and a
  welfare state
• social consequences of immigration perhaps
  these are more important than the economic
  effects.
• Assimilation immigrants often start at the
  bottom of the pile but many groups then rise up.
• Does immigration grease the wheels of the
  labour market?

12
Empirical evidence of impact of immigration

• The experiment one would like to do is to drop
  some immigrants at random into certain labour
  markets and then observe the outcomes.
• That is difficult if not impossible to do -
  though some studies have tried to use the
  dispersal policies applied to asylum-seekers by
  some countries.

13
Common Empirical Specification

• A typical regression using non-experimental data
  would try to run a regression of the form
• where I is some measure of the impact of
  immigrants on the local market (e.g. the share of
  immigrants).
• Problems
• Endogeneity of immigrant flow
• Responses of natives
• What is the right level of aggregation

14
Card, ILRR 1990Mariel Boatlift

• April 20 1980 Castro allows Cubans to leave for
  the US
• Between May and September 125000 did
• Most went to Miami
• 7 increase in labour force
• 20 increase in number of Cubans in Miami
• Clear exogenous shock to the Miami labour force
• Compares labour market performance before and
  after with comparison cities

15
Effect on Earnings
16
Effect on Unemployment Rate
17
Cards Conclusions

• virtually no effect on the wage or unemployment
  rates of less-skilled non-Cubans.
• How is this possible?
• off-setting flows of other immigrants or natives
  not very important
• industrial structure made it relatively easy to
  absorb large numbers of low-skilled immigrants.
• Other studies have used a similar methodology
  studying those who returned to Portugal after the
  end of its colonies, to France after Algerian
  independence and to Israel following the collapse
  of the Soviet Union.
• The pattern found by Card seems fairly common.

18
Borjas, QJE 2003

• Argued that cities are not distinct labour
  markets
• Labour mobility between them
• Trade between them
• Divides US labour market as a whole into segments
  by age and education similar to Card-Lemiuex
• Looks to see whether education-age cells with big
  changes in immigrant shares are correlated with
  wage changes of natives

19
Typical Borjas Result
20
Borjas Conclusions

• Borjas concludes immigrants do depress wages of
  natives with whom they compete
• But treats immigrants and natives as perfect
  substitutes within age-education cells
• This is relaxed by Ottaviano and Peri they find
  natives and immigrants are imperfect substitutes

21
Implications of Ottaviano-Peri

• Effect of new immigrants is primarily on wages of
  existing immigrants
• Effect on wages of natives does exist but is
  small and positive
• Conclusion is controversial

22
Manacorda, Manning, Wadsworth

• UK study
• Other UK studies found little impact of
  immigration on wages of natives
• Puzzle to reconcile this with Card-Lemieux who
  find that relative supplies do matter
• Estimated Card-Lemieux model but with third
  level in which immigrants and natives are
  imperfect substitutes

23
Rising Immigration into the UK
24
Manacorda, Manning, WadsworthConclusions
25
Implications for Effect on Wages
26
The Final Word?

• Perhaps not really
• Card quite critical of Borjas results
• Borjas has 3 education groups college, HS,
  drop-out (Card-Lemiuex had 2)
• Card argues it is important to distinguish
  between college/high school and high
  school/drop-put.
• Card argues the former wage ratio is sensitive to
  relative supplies but the latter is not
• It is the latter that, in the US in recent years,
  has been most affected by immigration.

27
Card evidence EJ 2005
28
Whats the big deal about immigration?

• Economists routinely fail to find large effects
  of immigration on natives
• The effects they do find are often positive
• Perhaps this misses the point because people get
  very upset about immigration

29
A Recent NBER Working PaperImmigration, Wages,
and Compositional Amenities Card, Dustmann and
Preston

• Economic theory says effects of immigration like
  effects of trade
• But people much more hostile to immigration than
  free trade
• Perhaps because immigration also alters
  communities as well as economies
• These effects seem at least as important

30
European Social Survey Data
31
And some more