Transcript Copenhagen Economics Institute Short Course March 5

Copenhagen Economics Institute
Short Course
March 5-7 2007
Globalization in the Very Long Run
March 7
Political Economy:
Protection and Immigration
Part 1. The political economy of
protection
Motivation
What determines tariff policy?
It can’t be conventional economics, since every mainstream
economist since Smith (1776) agrees that free trade is a
good thing for national income.
Yet, the politics of free trade have been surrounded by
controversy ever since Alexander Hamilton tried forcing his
protectionist policies on a new United States congress after
1789, and since Robert Peel ruined his political career
forcing free on the British Parliament in 1846.
Political leaders have never been solely, or even largely,
interested in maximizing national income, let alone
maximizing world income. Their main goal has always been
to get a larger slice of the pie for their supporters.
Protection and free trade have always been for sale in the
political market place (Grossman and Helpman 1994), but
having said so doesn’t make the question -- what
determines tariff policy? – much easier to answer.
A word of warning
Europeans, please note! The belief that
pre-1914 trade policies were liberal is a Big
Myth created by scholarly obsession with
western Europe. Blame your Euro-centric
Teachers!
Road Map
1. Four big world tariff facts
2. Do immense tariffs always mean
globalization backlash?
3. Endogenous tariffs in the world economy
1870-1938
4. Endogenous tariffs even earlier:
Zollverein 1818-34; US after 1789; Latin
America after 1823
Four big tariff facts
First: Inflations and deflations played a powerful
role at key points in the past. Import duties were
typically specific until modern times, quoted as
pesos per bale, dollars per yard, or yen per ton.
Under a regime of specific duties, abrupt changes
in price levels changed import values in the
denominator, but not the legislated duty in the
numerator, thus producing big percentage point
changes in equivalent ad valorem tariff rates.
Second: The well-known surge to world protection
in the interwar is almost matched by a less well
known protectionist drift worldwide between 1865
and about 1900. And what looks like a modest pre-World
War I anti-globalization backlash – led by a European
retreat from the liberal pro-global trade positions in midcentury -- is far more dramatic when the world averages
are disaggregated. Indeed, there is a very pronounced rise
in tariffs across Latin America, across the non-Latin
European offshoots and across the European periphery.
This steep rise up to the 1890s in the periphery’s tariff rates
far exceeds that of the European core, a notable fact given
that almost nothing has been written on this anti-global
tariff trend in the periphery.
Third: There was enormous variance in levels of
protection between regions. The richer new world
European offshoots had levels of protection almost
three times that of the European core around 1900.
To take a second example, in 1925 the European
periphery had tariffs about two and a half times
higher than those in the European part of the
industrial core. To take a third example, in 1885 the
poor but independent parts of Latin America had
tariffs almost five times higher than those in the poor
and dependent parts of Asia, while the poor but
independent parts of Asia had tariff rates about the
same as the poor but dependent parts of Asia.
Fourth: There was great variance within regions. In
1905, tariffs in Uruguay were two and a half times
those in Canada, while tariffs in Brazil and Colombia
were ten times those in China and India. The same
high-low range appeared within the industrial core (US
five times the UK) and the European periphery
(Russia six times Austria-Hungary). From 1919 to
1938, the tariff variance between countries was about
the same as tariff variance over time, but from 1865 to
1914, the tariff variance between countries was more
than twice that of the tariff variance over time. Thus,
explaining differences in tariff policy between
countries may be even more challenging then
explaining tariff policy changes over time.
Addendum
In the new economic order (Lewis), labor-abundant
and land-scarce Europe exported manufactures and
imported primary products. The periphery did the
opposite. Thus, European tariffs were mainly imposed
on import-competing primary products (and higher
than their average) while periphery tariffs were mainly
imposed on import-competing manufactures (and
higher than their average).
In others words, future work needs to look at the
structure of protection.
Restrictive trade policy in the First Global Century
The 35-country sample
(85% of 1914 world population, 95% of 1914 world GDP)
6 Core industrial leaders: AH, Fr, Ger, It, UK, USA
8 European Periphery: Den, Grc, Nor, Port, Serb, Sp, Swe, Rus
8 Latin American Periphery: Arg, Brz, Col, Ch, Cuba, Mex,
Per, Ur
10 Asia-MidEast: Bur, Cey, Egy, Ind, Indo, Jap, Phil, Siam, Turk
3 English-speaking European Offshoots: Aus, Can, NZ
Figure 1: Unweighted World Average Own Tariff, 35 Countries, %
30
25
20
15
10
5
0
1860
1880
1900
1920
1940
1960
1980
2000
Figure 2 Unweighted Average of Regional Tariffs Before World War II
45
Unweighted Average Tariff (%)
40
35
30
25
20
15
10
5
0
1865
1870
1875
1880
1885
Asia
1890
1895
Core
1900
1905
Euro Perip
1910
1915
Lat Am
1920
1925
Offshoot
1930
1935
US
1940
1945
1950
Endogenous Tariffs in Table 1
Right-hand side variables (all but dummies in logs):
• Export Share: X/GDP, a measure of a
country’s export boom or slump. Revenue
tales;
A Note on Tariffs for Revenue
R = tpM
where R is revenue, t is the average ad valorem tariff rate, p is the
average import price and M is import volume.
Totally differentiating with respect to t, and assuming that the typical
19th century country in the periphery was a price taker for
manufacturing imports, yields
dR/dt = pM + (tp)dM/dt.
The revenue-maximizing tariff rate, t*, is found by setting
dR/dt = 0
in which case
t* = -1/(1 + e)
where e is the price elasticity of demand for imports. Irwin (1998)
estimates e to have been about -2.6 for the US between 1869 and
1913, so assuming e = -3 can’t be too far off the mark. In which case,
the revenue-maximizing tariff in the periphery would have been very
high indeed, about 50 percent.
Suppose some government in the periphery had in mind some
target revenue share in GDP
R/Y = r
and could not rely on foreign capital inflows to balance the
current account (so pM = X), then
r = tpM/Y = tX/Y
If foreign exchange earnings from imports (and thus imports)
were booming (an event which could be caused by a terms of
trade boom, denoted here by a fall in the relative import price,
p, or by a supply-side expansion of export quantities, X), then
the target revenue share could have been achieved at lower
tariff rate, t.
Moral? The bigger the export boom, the higher the export
share, the bigger the import share, and the lower the
necessary tariff rate.
Endogenous Tariffs in Table 1
Right-hand side variables used (all but dummies in logs):
Export Share: X/GDP, a measure of a country’s export
boom or slump. Revenue tales;
GDP/capita, and Schooling, the latter the primary school
enrollment rate. Skill endowment proxies. StolperSamuelson stuff;
Population. Easier for local firms to find spatial niches in
bigger domestic markets. Alternatively, big and dense
populations imply efficient internal tax administration and
less need for tariff revenue. Balkanization matters;
Lagged Partner Tariffs. Weighted average of the tariff
rates in trading countries’ markets, the weight being trade
volumes. Strategic tariff stuff;
Effective Distance. Distance from each country to the US
or the UK, adjusted by seaborne freight rates specific to
that route. Transport-tariff trade-off;
Railway Mileage. Railway mileage added in the country.
More transport-tariff trade-off;
Urbanization: Percent of population in cities and towns
greater than 20,000. More Stolper-Samuelson stuff;
Federal. Dummy variable = 1if a federal system, = 0 if
centralized. Lacking other tax instruments, federal
governments need more tariff revenue;
Colony: Dummy = 1. For 1870-1938, includes Burma,
Ceylon, India, Indonesia, the Philippines. Also, Cuba
(1870-1901), Egypt (1882-1938) and Serbia (1870-1920).
Inflation and inflation-squared. Rates in home markets.
The much-touted specific duty effect;
Lagged Px/Pm: Country’s terms of trade as observed in
world markets. De-industrialization fears versus revenue
needs. Note: now confirmed with more complete data.
Capital inflows: Country’s absorption of British capital
exports. Easing short term liquidity constraint and tariff
revenue needs. Caution: only pre-1914 data.
Gold Standard: Dummy = 1 when on gold standard or
pegged to a core currency. Losing control over the real
exchange rate (an alternative means of protection). Note:
assessed elsewhere. Note: exploring availability of other
tax instruments.
Table 1. Tariff Rate Determinants the World Around 1870-1938
Dependent variable: ln Own Tariff
Specification
TS, country dummies
CS, year dummies
Years
1870-1938 1870-1938 1870-1938 1870-1938 1870-1938 1870-1938 1870-1938 1870-1938 1870-1938 1870-1938
Revenue Motive
ln Export Share
-0.0285
-0.0832
-0.0609
-0.0463
-0.0924
-0.0397
-0.0645
-0.0601
-0.0539
-0.0753
(-1.36)
(-3.02)
(-2.30)
(-2.07)
(-3.32)
(-1.37)
(-1.67)
(-1.60)
(-1.80)
(-2.02)
Strategic Tariff Motive
ln Partner
0.2490
0.2507
0.2992
0.2246
0.2526
-0.0440
-0.0983
-0.0338
-0.0648
-0.0953
Tariffs
(9.06)
(6.64)
(8.45)
(7.54)
(6.67)
(-1.22)
(-1.82)
(-0.60)
(-1.76)
(-1.73)
Stolper-Samuelson Scarce Factor Compensation Motive
ln Terms of
0.0798
0.1219
0.1037
0.1371
Trade Index
(2.22)
(2.68)
(2.55)
(2.66)
ln GDP per
-0.1412
-0.2227
-0.1745
-0.1810
-0.2260
-0.1025
-0.1445
-0.1228
-0.1439
-0.1435
capita
(-2.40)
(-2.86)
(-2.28)
(-2.95)
(-2.90)
(-1.48)
(-1.44)
(-1.24)
(-2.00)
(-1.45)
ln Schooling
0.1640
-0.0560
-0.0573
0.1719
-0.0416
0.0672
-0.3046
-0.2993
0.0548
-0.3053
(4.02)
(-0.82)
(-0.84)
(4.30)
(-0.61)
(1.49)
(-2.96)
(-3.01)
(1.22)
(-2.99)
ln Effective
-0.0735
-0.1072
-0.1267
-0.0584
-0.1086
-0.0169
-0.0644
-0.0514
-0.0309
-0.0616
Distance
(-4.86)
(-4.95)
(-5.97)
(-3.76)
(-5.02)
(-0.74)
(-1.53)
(-1.28)
(-1.29)
(-1.48)
ln Railway
0.0354
0.0639
0.0579
0.0347
0.0590
0.0055
0.0212
0.0190
0.0042
0.0219
Mileage
(3.38)
(2.25)
(1.98)
(3.41)
(2.08)
(0.80)
(0.93)
(0.84)
(0.56)
(0.94)
ln Urbanization
0.0478
0.0198
0.0013
0.0462
0.0235
0.0242
-0.0890
-0.0989
0.0211
-0.0787
(2.13)
(0.30)
(0.02)
(2.10)
(0.36)
(0.99)
(-1.58)
(-1.66)
(0.79)
(-1.41)
Controls
ln Population
-0.1084
-0.1716
-0.1441
-0.1172
-0.1721
-0.1224
-0.0433
-0.0545
-0.1302
-0.0504
(-2.50)
(-3.35)
(-2.81)
(-2.58)
(-3.38)
(-2.85)
(-0.84)
(-1.12)
(-3.00)
(-1.00)
Federal
0.0100
0.0524
0.0585
0.0071
0.0509
(0.35)
(1.45)
(1.55)
(0.25)
(1.35)
Colony
-0.0033
-0.1649
-0.2797
-0.0695
-0.1515
(-0.05)
(-0.83)
(-1.58)
(-1.50)
(-0.79)
Inflation
-0.0004
-0.0005
-0.0004
-0.0003
(-1.45)
(-1.46)
(-0.90)
(-0.69)
Inflation Squared
0.0000
0.0000
0.0000
0.0000
(2.45)
(1.77)
(0.44)
(0.52)
Constant
N
R-squared
2.7797
(4.75)
5.8022
(7.80)
5.4237
(7.45)
2.6333
(4.28)
5.1674
(6.68)
2,138
0.224
1,169
0.271
1,300
0.25
1,951
0.251
1,169
0.266
2,067
0.144
1,116
0.203
1,238
0.195
1,889
0.149
1,116
0.211
Endogenous Tariffs Even Earlier
United States 1789-1830:
Hamilton’s famous 1791 Report on infant
industry was irrelevant. Trade policy was
about revenue and strategic tariffs.
Zollverein 1818-1834:
List and Schmoller’s rhetoric on infant
industry and British-induced deindustrialization fears was irrelevant. The
customs union was all about revenue.
Latin America 1823-1867
In place of a colonial customs union, they
got independence, balkanization and
bloodshed. Tariffs were all about revenue.
High tariffs + balkanization + bloodshed =
five “ lost decades.”
Tariff rates rise in the US
Figure 3 Average Tariff Rates in the United States 1790-1836
70
60
Tariff Rate
50
40
30
20
10
Average Net Tariff
Average Net Tariff on Dutiable Imports
0
1790
1795
1800
1805
1810
1815
Year
1820
1825
1830
1835
Was it all about List’s infant
industry or was it about revenues?
Table 2
Size and Administration Efficiency:
The Zollverein 1818-1836
Custom Administration
Costs/Customs Revenues
Zollverein (1836)
Prussia (1818)
Bavaria (1826)
Wurttemberg (n.d.)
Kurhessen (n.d.)
Hessen-Darmstadt (n.d.)
0.147
0.175
0.250
0.430
1.000
1.000
Kuehne's
Ratio
0.130
0.200
0.240
0.470
1.130
1.060
Area
(meilen)
9569
5104
1337
354
166
153
Figure 4 Fiscal Scale Economies in the Zollverein
Predicted Cost Ratio
10
1
0
5000
10000
15000
0.1
Area in English Square Miles
20000
25000
30000
Six Morals
 Stolper-Samuelson mattered, but
it was never a key player
 Infant industry and ISI ideas
were never important before the
1890s, but they were three or four
decades before the 1930s
De-industrialization fear was a
major factor determining tariff
policy in the periphery long before
the 1930s, joining grain-invasion
fear in the industrial core
Revenue need was always a
major explanation for high tariffs in
the periphery
Where geography was conquered,
high tariffs compensated the losers
Strategic efforts to improve the
terms of trade were always at work
pushing up tariffs. Those forces
were only resolved after WWII
Want to Learn More?
You can down load papers at Williamson’s web site
www.economics.harvard.edu/~jwilliam/
“Closed Jaguar, Open Dragon: Comparing Tariffs in Latin America and
Asia before World War II,” NBER Working Paper 9401 (December
2002). With Michael Clemens.
“Always Protectionist? Latin American Tariffs from Independence to
Great Depression,” Journal of Latin American Studies vol. 36, part 2
(May 2004), pp. 205-32. With J. Coatsworth.
“Why Did the Tariff-Growth Correlation Reverse After 1950?” Journal
of Economic Growth vol. 9, no. 1 (March 2004), pp. 5-46. With M.
Clemens.
“Globalization in Latin America Before 1940,” in V. Bulmer-Thomas, J.
Coatsworth and R. Cortés Conde (eds.), The Cambridge Economic
History of Latin America: Volume II: The Long Twentieth Century
(Cambridge: Cambridge University Press, 2006). With Luis Bértola.
“Explaining World Tariffs 1870-1938: Stolper-Samuelson, Strategic
Tariffs and State Revenues,” in R. Findlay, R. Henriksson, H. Lindgren
and M. Lundahl (eds.), Eli F. Heckscher, International Trade, and
Economic History (Cambridge, Mass.: MIT Press, 2006).
End of Advertisement
Part 2. The political economy of
immigration
Actually, I think there’s a better way to pursue this.
Namely, ask
“Why Have Trade and Immigration Policies Always
Differed in Labor-Scarce Economies?”
a question Tim Hatton and I asked recently in The
New Comparative Economic History (Cambridge,
Mass.: MIT Press, forthcoming 2007).
The Policy Paradox
• A century ago, trade policies were restrictive
and immigration policies were liberal.
• Today, trade policies are liberal while
immigration policies are restrictive.
• So, why have policies towards trade and
immigration always differed in labor-scarce
economies?
The simple HS-SS 2x2x2 model
• As Robert Mundell (1957) pointed out 50
years ago, in the labor-scarce economy
open immigration and free trade policies
both lower wages, while restricted
immigration and protection both raise
wages. Thus, open immigration will offset
the distributional impact of protection, and
free trade will offset the impact of
immigration restriction.
What accounts for the Paradox?
Better theory could make the Paradox evaporate:
After all, the predictions of HO-SS factor endowment
model might be overturned by modifying it with
specific factors, increasing returns or Ricardian
differences in productivity.
Better political economy could make the Paradox
evaporate: After all, factor income distribution isn’t
the only determinant of policy.
We favor better political economy.
Road Map
• The evolution of immigration restriction (less to
more) and trade restriction (more to less) over two
centuries.
• Declining immigrant positive selectivity, declining
immigrant ‘quality’ and rising immigrant restriction.
• The fiscal implications of the rise of the welfare state
for trade and immigration policy.
• The role of suffrage in changing the median voter.
• Forming public opinion and public policy.
• The bottom line.
Refrain: World trade policy was restrictive
in the First Global Century
Figure 1
Average Tariff, 35 countries 1870-1998
25
Percent
20
15
10
5
0
1870
1880
1890
1900
1910
1920
1930
1940
Year
1950
1960
1970
1980
1990
Refrain: Trade policy was very restrictive in laborscarce countries during the First Global Century
Figure 2
Unweighted Average of Average Tariffs Before World War 2
45
40
US
Lat Am
Offshoots
Euro Periph
Core
Asia
35
Tariff Percent
30
25
20
15
10
5
0
1865 1870 1875 1880 1885 1890 1895 1900 1905 1910 1915 1920 1925 1930 1935
Year
Liberal trade policy
in the Second Global Century
• Since 1950, labor-scarce countries have
gone open.
• The fact that the Third World stayed closed
for so long (especially Latin America and
eastern Europe) is irrelevant given our
focus on trade policy in labor-scarce
countries.
Open immigration policy in labor-scarce
economies in the First Global Century
• Index of policy stance for 5 major immigration
countries: Australia, Argentina, Brazil,
Canada and US (Timmer and Williamson
1998).
• Index = -5 for tough quotas (including
exclusion of those from poor countries), 0 for
neutrality and +5 for generous subsidies.
Figure 3
New World Immigration Policy Index 1860-1913
4
Restrictiveness Index
3
2
1
0
-1
-2
-3
1860 1865 1870 1875 1880 1885 1890 1895 1900 1905 1910 1915 1920 1925 1930
Year
Rising immigration restriction in the Second
Global Century
• Quotas of the interwar period continued (after
temporary guestworker experiments)
• But labor-scarce countries ‘leveled the playing
field’ by removing racist restrictions on poor
Third World immigrant source countries
• The fact that OECD immigration has soared is
not inconsistent with restriction. After all, the
immigration would have been 5, 10 or 20
times higher without it.
• Thus, the Paradox: liberal trade policy and
restrictive immigration policy coexist today, in
the second global century; and liberal
immigration policy and restrictive trade policy
coexisted in the first global century.
• Why? Is the HO-SS model irrelevant, or is the
political economy of policy formation much
more complex?
The long run decline in immigrant positive
selectivity and ‘quality,’ and the rise in
immigration restriction
• In the early 1800s, European overseas emigration was a trickle
and it was very positively selected.
• They came from the relatively developed northwest where an
early industrial revolution had raised living standards.
• They came from the middle and upper parts of the income and
wealth distribution.
• Even if the labor-scarce host countries only wanted
immigrants like themselves, they did not have to use
discriminatory policy to get that result. Long distances, high
transportation costs and source country poverty kept out the
poor from backward east and southeast Europe.
As you have seen, US native-born
‘quality’ rose 1870-1930
Quality Proxies for the US Population 1870-1930
____________________________________________
Enrollment
Attendance % 17 Year olds
Illiteracy
Rates Per 100
Rates per
Graduating
Rate
Population
Student
High School
_________________________________________________________________________
1870
48.4
78.4
2.0
20.0
1880
57.8
81.1
2.5
17.0
1890
54.3
86.3
3.5
13.3
1900
50.5
99.0
6.3
10.7
1910
59.2
113.0
8.6
7.7
1920
64.3
121.2
16.3
6.0
1930
69.9
151.7
28.8
4.3
__________________________________________________________________________
… while the ‘quality’ of US immigrants fell.
The Occupations of US Immigrants
Occupation
Skilled
Farmers
Unskilled
Miscellaneous
18201831
61
23
16
--
18321846
40
33
26
--
18471854
24
33
43
--
18551864
36
23
41
0
18651873
31
18
51
1
18731880
30
18
48
5
18811893
24
14
60
3
18941898
30
12
55
3
Percent male
70
62
59
58
62
63
61
57
… more on the fall of immigrant ‘quality’ as their
numbers rose.
Table 2
GDP Per Capita Ratio: Average Source Country/Destination Country
Country
US
Canada
Argentina
1860s
95.4
154.8
114.2
1870s
92.3
183.1
110.2
1880s
73.3
159.4
89.8
1890s
64.0
136.7
68.4
1900s
49.5
107.0
54.6
Country
US
Canada
Australia
Germany
UK
1950s
49.1
64.5
73.4
95.6
--
1960s
40.8
60.0
75.4
70.1
--
1970s
29.8
40.8
64.5
61.1
75.3
1980s
24.0
33.7
55.5
51.1
83.1
1990s
22.4
30.8
49.0
44.7
86.2
… and the long run relative quality decline
continues. Relative wage and relative education:
immigrants vs native-born US 1960-1990.
1960
1970
1980
1990
Percentage earnings differential relative to the native-born
All Immigrants
Earnings unadjusted
4.1
0.1
9.7
16.3
Earnings adjusted
1.3
1.7
7.1
10.0
Recent immigrants
Wage unadjusted
13.9
18.8
32.8
38.0
Wage adjusted
16.2
19.8
24.1
26.9
Percentage point difference in educational attainment relative to native-born
All immigrants
Education > 16 years
3.5
2.4
0.0
Education < 12 years
3.2
14.3
22.1
Recent immigrants
Education > 16 years
12.9
7.5
4.9
Education < 12 years
5.6
13.1
20.4
… and falling immigrant labor quality has left its
mark on rising immigrant poverty rates.
Poverty rates (%)
US 1959
Foreign-born
Native-born
US 1999
Foreign-born
Native-born
14.2
20.9
17.4 (up 3.1)
11.8 (down 9.1)
Summary of the road trip so far
• Immigrant restriction since 1895 has been driven in
large part by the long run decline in potential
immigrant quality and their rising numbers.
• The long run decline in immigrant quality would have
been far greater without the rise of restrictions.
• Immigration policy is much tougher now simply
because there are vastly more potential immigrants
from poor countries to keep out.
Changing fiscal implications of trade over two centuries.
• Alexander Hamilton thought “the tariff was more
important as a tool of fiscal policy than as an
instrument for promoting manufactures.”
• Customs duties were a major source of central
government revenue in the 19th century, especially in
labor-scarce and land-abundant overseas countries
where low population and taxpayer density made
other forms of taxation inefficient.
• Tariff revenues are unimportant in labor-scarce
countries today.
To repeat: tariffs were an important
source of revenue then, but not now
Customs duties/total tax revenues:
Latin America 1820-1890
66%
US 1850s
90
Australia 1850s
90
Ave. 7 labor-scarce 1890s58
vs. UK and France 1890s
<20
vs. OECD 1970s
4
Immigration, the rise of the welfare state
and fiscal burdens.
• Immigration had little fiscal impact in the first
global century before the rise of the welfare
state. The fiscal impact rose in the second
global century with the decline in immigrant
quality and the rise of the welfare state.
• Thus, tariffs brought fiscal benefits and
immigrants brought no fiscal costs before
1914. The opposite is true today.
The rise of the welfare state (implying rising
net fiscal effects of immigrants).
Figure 4
Welfare State Spending 1880-1980
30
US
Canada
France
Germany
UK
Sweden
Percent of GDP
25
20
15
10
5
0
1880
1890
1900
1910
1920
1930
Year
1960
1970
1980
Who gets the goodies?
Welfare Dependency and Personal Characteristics in the EU 1994-6
(differences between immigrants and EU nationals)
Country
Percentage point difference
between immigrants and EU
nationals in receipt of
Difference in characteristics between
immigrants and EU nationals
Unemp.
Benefit
Low
educated
High
educated
Family
Benefit
Pensions
Age
(years)
No. of
children
Germany
1.6
--
--
21.2
5.5
8.6
0.54
Denmark
24.5
5.3
17.9
14.7
0.6
7.8
0.47
Netherlands
7.0
7.9
14.9
22.7
5.3
7.7
0.65
Belgium
6.7
1.1
6.1
10.6
14.1
2.5
0.12
France
4.9
16.7
12.8
22.5
7.2
3.6
1.10
UK
0.6
0.6
23.4
15.4
21.2
8.7
0.85
Austria
8.9
8.1
18.0
7.8
12.2
10.6
0.35
Finland
31.7
0.2
12.7
12.3
17.5
7.4
0.04
Suffrage and the changing median voter.
Two questions matter, not just one:
[1] Who stood to gain, and who stood to lose from
trade and immigration policy?
[2] Who had the vote?
It could be argued that US (male, white) workers
dictated policy in all 19th c labor-scarce countries
since they had the vote, and other countries competing
for trade and immigrants had to follow suit.
Figure 5
Percent of Adults Voting: New World, 1850-1940
70
60
Percent Voting
50
40
US
30
Canada
20
Argentina
10
Chile
0
1850
1860
1870
1880
Brazil
1890
1900
Year
1910
1920
1930
1940
Figure 6
Percent of Adults Voting: Europe, 1850-1940
100
90
80
Percent
70
60
50
40
30
France
20
UK
Germany
10
Netherlands
Italy
0
1850
Sweden
1860
1870
1880
1890
1900
Year
1910
1920
1930
1940
Bringing the summary of the
road trip up to date
Three factors conditioned trade and immigration
policies in overseas labor-scarce economies in
the first global century:
[1] revenue needs for development and war;
[2] sending country poverty constraints that
kept immigration moderate and selective for
most of the period;
[3] often-limited political franchise.
What about public opinion and policy today?
Two questions:
[1] Is public opinion more negative
towards immigration than trade? If not,
why are trade policies more liberal?
[2] What are the individual and country
characteristics that determine public
opinions?
The International Social Survey Programme
• Interviewer’s question: Would you like to see
immigration increased a lot (1), a little (2), stay the
same (3), reduced a little (4), reduced a lot (5)?
• Same question asked of imports.
• ISSP conducted in 24 countries in 1995, of which 14
are labor-scarce OECD.
• O’Rourke and Sinnott (2001, 2006), Mayda (2004),
Mayda and Rodrik (2005) and O’Rourke (2006) use
country fixed effects; Hatton and Williamson (2007)
use explicit country characteristics.
Anti-global attitudes in the 1990s
Attitudes Toward Immigration and Trade, 1995/6
Country
Australia
Austria
Canada
Germany
Great Britain
Ireland
Italy
Japan
Netherlands
New Zealand
Norway
Spain
Sweden
USA
All countries
AntiImmigration
opinion
3.768
3.808
3.311
4.270
4.060
3.073
4.148
3.373
3.822
3.737
3.845
3.385
3.970
3.880
3.770
Anti-Imports
opinion
3.999
3.907
3.292
3.283
3.772
3.664
3.599
2.939
2.930
3.401
3.146
3.889
3.254
3.765
3.480
Correlation
coefficient
0.271
0.267
0.284
0.370
0.325
0.178
0.243
0.219
0.272
0.310
0.240
0.180
0.253
0.249
0.237
No of
observations
2318
923
1310
1630
955
919
1020
1000
1864
950
1333
1014
1132
1090
17458
Note: Based on data from the 1995 ISSP module on national identity. Figures are average attitudes on a
five point scale where respondents were asked whether immigrants or imports should be increased a lot
(1), increased a little (2), kept the same (3), reduced a little (4), or reduced a lot (5).
Note! Anti-immigrant opinion = 3.77 and anti-import opinion =
3.48. How do we reconcile their similarity with the fact that trade
policy is liberal and immigration policy is restrictive?
Negative
Public
Opinion
Immigration
Imports
Immigration or Imports
Determinants of anti-global attitudes
Explanatory
Variable
Individual-level variables
‘Patriotism’
‘Chauvinism’
Foreign-born
2nd Generation Immigrant
Female
Age/100
Married
Highly Educated
Employed
Country-level variables
Log GDP Per Capita
Inequality
Log Population
Welfare Expenditure /GDP
Share of Popn Foreign
Imports/GDP
OECD Trade/GDP
R2
No of obs
(1)
Anti-Immigration Opinion
(2)
Anti-Imports Opinion
0.055
0.374
-0.035
-0.283
0.035
0.009
0.038
-0.219
-0.008
(1.81)
(8.23)
(0.32)
(6.21)
(1.13)
(0.07)
(1.77)
(7.13)
(0.51)
0.201
0.397
-0.130
0.085
0.304
-0.001
0.029
-0.280
-0.032
0.692
1.850
0.077
0.047
0.044
(2.58)
(2.26)
(1.51)
(7.26)
(3.13)
-0.294 (0.57)
4.043 (2.23)
-0.072 (0.64)
0.207
14820
(7.39)
(13.7)
(1.99)
(2.11)
(11.3)
(1.08)
(1.40)
(7.32)
(1.07)
0.006 (0.28)
-0.009 (0.93)
0.219
14820
… and the trend in policy reflected the rise in
anti-immigrant opinion 1960s-2001
Government Immigration Policies, 1976-2001
(Percent of governments aiming to restrict immigration more)
Year
All Countries
More Developed Countries
Less Developed Countries
Source: United Nations (2002), p. 18.
1976
7
18
3
1986
20
38
15
1996
40
60
34
2001
40
44
39
Based on Heckscher-Ohlin and Stolper-Samuelson
thinking, we have a policy paradox for labor-scarce
economics then and now:
• A century ago, restricted trade and unrestricted
immigration.
• Now, restricted immigration and unrestricted trade.
• The immigration policy fundamentals at work:
As the cost of migration fell and the ability of poor
workers to finance the move rose, the number of
immigrants soared and their quality fell. With the rise
in both, restrictive policy emerged.
Today, anti-immigration opinion is driven by:
● the scale of immigration, which represents a labor
market threat;
● the size of the welfare state combined with low
immigrant quality, which represents the welfare burden;
● and the universal franchise, which assures that these
concerns are reflected in policy.
Why isn’t anti-immigrant opinion even more negative?
● it would be without the restriction!
● and the median voter today is far less threatened by
labor market competition than a century ago since the
‘quality’ gap between native-born and immigrant has
gotten much bigger over time.
What about trade policy and opinion?
● the revenue motive was central a century
ago, but irrelevant today;
● the median voter faces a far smaller threat
from low-skill labor embodied in imports from
poor countries;
● and there is a greater balance between losers
and winners from trade since trade is far more
intra-industry today than it was a century ago.
It’s also easier for the losers to out-migrate
from declining sectors today.
Why do richer countries, cet. par., have more antiimmigrant opinion but not more anti-import feeling?
• For the median voter, the threat is the fiscal
implications of the welfare state.
• This threat is far greater today than a century ago
because of the combination of the rise of the welfare
state and the release of the poverty trap which held
back the pool of potential poor and less skilled
immigrants (compared with the native-born).
• These underlying fundamentals will not diminish
over the next half century.
The end of the Copenhagen
short course.
Many thanks!