Transcript Slides - University of Nottingham

The Slump and Immigration
Policy
Tim Hatton
University of Essex
and
Australian National University
University of Nottingham, 12th March 2014
Slumps and immigration policy
• In the aftermath of the Global Financial Crisis many expected
a sharp fall in support for immigration and a policy backlash.
• After the GFC some assessments invoked historical parallels;
others looked for reasons why this time is different.
• For European countries the backlash has been milder than
some expected. The question is why. Here I examine:
• 1) Immigration and the labour market: do migrants bear a
large part of the adjustment burden in a recession?
• 2) Public opinion: have attitudes become more negative in the
recession, and if so, where and why?
• 3) What are the political forces driving policy and what has
influenced them?
Slumps and policy in the past
• US: massive tightening of policy in the early 1920s as
unemployment increased from 5 to 11 percent.
• The Great Depression: sharp turn to restrictive policies in
other immigration countries: Australia, Canada, Brazil,
Malaya.
• The oil crisis 1973-4: Sudden stop of guestworker policies in
Europe, increasingly restrictive policies in UK and elsewhere.
• 1997/8 Asian financial crisis led to restrictions and
deportations, but the crisis was short-lived.
• The sharp backlash typically occurred in response to a ‘sudden
shock’ but after a long period of rising immigration and slowly
mounting public opposition. Is this time different?
Immigration in recessions.
• Does immigration cushion the labour market? If so then that
might mute the policy response.
• The great transatlantic migrations were very responsive to
business cycles. In the 1890s US and in the Great Depression
(Canada, Australia) immigration fell steeply.
• We see the same response in US net migration (including
illegals) in the 1980s and 1890s.
• In these cases immigration per 1000 fell by about the same as
unemployment (per 100) increased.
• This is the 10 % rule. Every 100 jobs lost results in 10 fewer
(net) immigrants. What about now?
Immigration and unemployment
OECD, International migration policies and data, Stocks and flows of immigrants, 2001-2011 and OECD
StatExtracts: Short-term Labour Market Statistics.
Immigration and unemployment
OECD, International migration policies and data, Stocks and flows of immigrants, 2001-2011 and OECD
StatExtracts: Short-term Labour Market Statistics.
Immigration and slumps in the past
• We can see this in Europe recently, in Ireland and Spain.
• But also for the UK and Sweden , where the slump was less
severe (and against a rising immigration trend).
• But it requires some freedom of movement for this to work,
e.g. pre-1914, elastic illegal immigration, free EU movement.
• So, net immigration bears a bit of the burden of recession.
• But it provides rather modest relief to the labour market. For
non-immigrants the 10 percent rule is a 90 percent rule.
• What about immigrants who stay?
Unemployment Rates in Western Europe, 2008-12
Non-immigrant men
2008
2012 Increase
Immigrant men
2008
2012 Increase
Austria
2.9
3.5
0.6
7.3
8.7
1.4
Belgium
5.3
5.8
0.6
14.3
17.6
3.2
Germany
6.6
5.2
-1.4
12.3
8.9
-3.3
Denmark
2.7
7.1
4.3
6.4
13.5
7.1
Spain
8.8
22.4
13.6
16.4
36.5
20.1
Finland
5.9
8.1
2.2
12.4
14.5
2.1
France
6.3
9.0
2.7
11.4
15.5
4.1
Great Britain
6.1
8.5
2.4
6.8
8.2
1.5
Greece
5.2
20.1
14.9
5.0
34.5
29.5
Ireland
7.0
17.8
10.8
8.2
19.4
11.2
Italy
5.6
9.7
4.1
5.9
12.4
6.5
Netherlands
2.1
4.6
2.5
5.3
10.5
5.2
Norway
2.4
3.1
0.7
6.0
7.2
1.2
Portugal
6.8
16.2
9.4
7.8
20.0
12.2
Sweden
5.1
6.7
1.6
11.5
16.9
5.4
Country average
5.3
9.9
4.6
9.1
16.3
7.2
The burden of unemployment
• Across 15 Western European countries from 2008 to 2012
unemployment rates among native men increased from 5.3 to
9.9 percent. For foreign men from 9.1 to 16.3 percent.
• This is the ratio rule. In recessions the ratio of unemployment
rates stays roughly the same but the absolute gap increases.
On average it was 1.86 in 2008 and 1.84 in 2012.
• Recent studies show that the responsiveness is much greater
for non-western immigrants.
• So: (1) high unemployment reduces immigration and (2) more
of it falls on the foreign born. These are cushioning effects
• But: (1) increased demand by immigrants on the welfare state
and (2) unemployment has fallen disproportionately on youth.
Unemployment incidence, 2008-12
Source: OECD, Labour market outcomes of immigrants, 2008-2012 (cf. Annex 2.A1)
Analysis of public opinion
• Recent analysis by economists of public opinion towards
immigration stresses three themes.
• 1) Labour market: fears that immigrants ‘take jobs’ or depress
wages.
• 2) Welfare state : concerns that immigrants increase pressure
on welfare services and/or increase the tax burden.
• 3) Prejudice: against non-white ethnic groups (possibly
exacerbated by the recession).
• But this is across individuals very little analysis of changes.
• So what happened in Europe? Let us look at the European
Social Survey in 2006 and 2010.
ESS Immigration attitudes
ESS opinion (average of 18 countries)
2006
2010 change
Immigrants enrich the culture
5.75
5.71
-0.03
Immigration good for the economy
5.04
4.92
-0.12
Immigrants make country a better place to live
4.95
5.02
0.06
Allow many of same race/ethnicity as home
country
Allow many of different race/ethnicity
5.60
5.65
0.05
5.02
5.11
0.09
Allow many from poor countries outside
Europe
4.93
4.91
-0.01
Scale: 0 – 10
Scale: 2-4-6-8
Countries: Belgium, Switzerland, Germany, Denmark, Estonia, Spain, Finland, France, United Kingdom,
Hungary, Ireland, Netherlands, Norway, Poland, Portugal, Sweden, Slovenia, Slovakia.
What drives the change in attitudes?
• Across 18 countries public opinion to immigration is not very
positive but it changed little between 2006 and 2010.
• The main decrease is in whether immigration is good for the
economy.
• Cross-country correlations of changes in average opinion(next
slide) show that:
• The rise in unemployment only matters for the question
‘Immigration is good for the economy’.
• The rise in social benefits to GDP ratio matters much more,
and across for all facets of attitudes.
• The budget deficit, symbolic of the financial crisis, also
exhibits strong correlations.
Cross-country correlations
Change in ESS Opinion (18 countries)
Correlation with change in:
Unempt
rate (%)
Immigrants enrich the culture
Social
benefits
(% GDP)
Budget
deficit
(% GDP)
0.18
-0.27
-0.60
Immigrants good for the economy
-0.57
-0.83
-0.89
Immigrants make country a better place to
live
Allow many of same race/ethnicity
-0.14
-0.54
-0.66
-0.04
-0.49
-0.71
Allow many of different race/ethnicity
-0.15
-0.39
-0.63
Allow many from poor countries outside
Europe
-0.27
-0.54
-0.62
Countries: Belgium, Switzerland, Germany, Denmark, Estonia, Spain, Finland, France, United Kingdom,
Hungary, Ireland, Netherlands, Norway, Poland, Portugal, Sweden, Slovenia, Slovakia.
Unemployment and opinion
Social benefits and opinion
Why the modest change in opinion?
• 1) There is probably an underlying trend improvement in
attitudes. (Because they are related to education).
• 2) Changes since 2006 are related mainly to concerns about
the public budget, but probably reflect broader malaise.
• 3) Economic performance has varied widely across European
countries, so negatives have been balanced by positives.
• 4) The salience of immigration policy has fallen. While
economic concerns have spilled over into attitudes, they have
also pushed immigration down the agenda. This has been
helped by the slowing down of immigration.
The Economist on salience
Salience in Europe
Source: Dyrbekk (2011)
Policy: Why waste a good crisis?
• Politicians can create salience; it might be good in time of
crisis to try and divert attention away from areas (the
economy) where they are perceived to have failed. Also,
restraints operating in normal times might be weaker.
• Policy is influenced by populist right-wing parties, which have
grown since the 1980s (10 percent in early 2000s to 15
percent in 2011). And they seem to be resurgent now.
• In North Europe, these have managed to conflate or combine
anti-immigration with anti-EU and anti-bailout rhetoric. (UK:
UKIP; Finland: True Finns; Netherlands: Freedom Party;
Norway: Progress Party, Austria: Freedom Party).
Policy and parties
• In the South, anti-immigration and anti-austerity have been
linked, but only one major right-wing surge, in Greece: Golden
Dawn.
• The main influence of right-wing populist parties is to shift the
agendas of mainstream parties—mainly on the right but also
on the left.
• So policy may become more restrictive despite relatively
benign (average) public opinion.
• But there are also constraints. One is the structure of
immigration itself.
Composition of entry channels
Austria
Belgium
Denmark
Finland
France
Germany
Ireland
Italy
Netherlands
Norway
Portugal
Spain
Sweden
UK
Employment Family
1.7
11.7
15.6
5.6
11.4
9.0
7.7
33.4
10.4
5.8
19.7
38.9
6.6
35.5
Humanitarian Free
21.3
11.0
33.4
3.7
13.9
5.4
38.4
11.0
39.9
5.1
18.6
3.8
48.9
0.4
27.9
2.3
65.5
51.2
56.9
41.3
33.7
67.9
43.0
34.9
0.4
0.0
8.1
3.7
10.0
0.7
0.0
1.6
21.2
21.4
38.8
16.4
37.7
26.2
58.3
63.8
32.7
42.6
38.0
22.6
0.0
0.0
8.6
1.9
0.0
11.6
10.1
8.9
0.2
0.3
17.6
4.0
Source: OECD, International Migration Outlook 2013, Country Notes.
Other
Policy channels
• Free (EU) migrants: No control over these, except for
accession states, but issues about rights to welfare.
• Employment stream: Most of the research on migrant
selectivity and the skill content of migration focuses on
migration for employment. But this is less important in
magnitude (and in opinion) than is often believed.
• Policy increasingly selective (Blue Card, points systems)—
predates the GFC but recently introduced in some countries.
Job search visas introduced in some countries e.g. Germany.
• For most countries with pro-skills policies, the scope for
further restriction is fairly limited.
Policy channels (contd)
• Family reunification: a larger component of immigration that
has been a focus of recent restrictive policies
• These include narrowing of eligibility by: (a) relationship, (b)
income/resources, (c) language test. E.g. Denmark, France,
Netherlands, UK. This process initiated pre-GFC by the family
reunification directive (2003).
• The evidence suggests that countries with more inclusive
policies (MIPEX) have a larger share of family migrants, so
there is some scope for restriction. But the MIPEX index of
reunification policies shows an overall improvement from
2007 to 2010 : EU 15, 60->62; EU25, 53->54.
Policy channels (contd)
• Humanitarian stream: A controversial issue.
• Two severe policy crackdowns on asylum applicants in mid1990s and mid-2000s. Tougher determination procedures,
reduction in support, tighter border control.
• Increasing focus on illegal immigration. CEAS initiatives on
border control pre-date the crisis (Frontex), but further
tightening in some countries.
• These policies helped reduce the numbers (which have also
fallen for other reasons).
• The Refugee Convention is opaque, but probably not much
scope left for tightening; the CEAS now easing policy (a bit).
The EU dimension
• According to the IoM (2010) early post-crisis migration policy
changes in EU countries ‘were not solely restrictive’.
• In asylum policy, evidence that the EU has put a floor under
tough national policies.
• In family reunification policy is constrained by the ECHR (Art
8). The ECJ has challenged some national policies on income
requirements and pre-entry tests e.g. the Netherlands.
• Evidence that the EU provides a coordinating mechanism that
limits the ‘race to the bottom’.
• Countries that are not part of the common immigration
system have shifted a bit more to restrictive policies, e.g. UK.
Denmark.
Conclusions
• History suggests that severe recession is followed by reduced
net immigration and a sharp policy backlash.
• In the slump the 10% rule seems to have operated. But it
provided modest respite from labour market pressures.
• Public opinion overall has not shifted very much; partly as the
severity of the slump has varied widely across countries.
• Opinion has responded more strongly to the rise in social
spending and the budgetary crisis than to the labour market.
• Immigration has diminished in salience, but in some countries
it has been conflated with anti-EU by minority parties.
• Policy makers may have wanted more draconian policies but
the policy options are somewhat constrained.
Regressions on country/year fixed effects from ESS;
114 obs, with country f/e and year dummies
Foreignborn (%)
Scale: 0 – 10
Immigration good for the economy
Immigrants enrich the culture
Immigrants make country a better
place to live
Scale: 2-4-6-8
Allow many of same race/ethnicity
as home country
Allow many of different
race/ethnicity
Allow many from poor countries
outside Europe
Unempt
rate (%)
Social Exp
R2
% of GDP (between)
-0.038
(2.04)
-0.047
(2.88)
-0.024
(1.56)
-0.022
(2.03)
0.013
(1.35)
-0.001
(0.12)
-0.089
(4.87)
-0.023
(1.46)
-0.043
(2.86)
0.495
-0.073
(4.98)
-0.041
(2.95)
-0.038
(2.70)
0.008
(0.99)
0.001
(0.17)
-0.000
(0.03)
-0.036
(2.52)
-0.037
(2.69)
-0.045
(3.25)
0.321
0.258
0.382
0.294
0.308
Country Average Opinion Scores
Country (no of rounds)
Belgium (6)
Switzerland (6)
Czech Republic (5)
Germany (6)
Denmark (6)
Estonia (5)
Spain (6)
Finland (6)
France (5)
Great Britain (6)
Greece (4)
Hungary (6)
Ireland (6)
Netherlands (6)
Norway (6)
Poland (6)
Portugal (6)
Sweden (6)
Slovenia (6)
Slovakia(5)
More/less
same
ethnic grp
5.63
6.07
4.93
6.03
6.10
5.65
5.25
5.44
5.42
5.30
4.76
5.34
5.68
5.43
6.01
5.94
4.51
6.49
5.57
5.55
More/less
different
ethnic grp
5.01
5.37
4.46
5.30
5.19
4.62
5.10
4.81
5.05
4.91
3.79
3.90
5.30
5.20
5.45
5.56
4.32
6.32
5.22
5.05
More/less
from poor
countries
5.02
5.32
4.47
5.19
4.94
4.07
5.10
4.62
4.87
4.76
3.73
3.66
5.22
5.01
5.42
5.60
4.26
6.28
4.99
5.02
Immigrt
good for
economy
4.59
5.96
4.20
5.15
5.11
4.60
5.35
5.33
4.85
4.56
3.53
3.84
5.20
5.08
5.58
5.17
4.67
5.49
4.34
4.22
Immigrt
enrich
culture
5.74
6.11
4.42
5.99
6.01
5.16
5.93
7.13
5.27
5.00
3.47
5.21
5.65
6.10
5.89
6.43
5.18
7.03
5.18
5.04
Immigrt
better
place
4.63
5.34
4.25
4.99
5.72
4.33
5.06
5.46
4.60
4.62
3.21
4.09
5.46
5.06
5.18
5.73
4.03
6.23
4.60
4.40