Transcript nation” and “

The Economics of Nationalism and Trade
Xiaohuan Lan (CKGSB)
Ben Li (Boston College)
2012.11.05
1998 NATO’s Bombing of Chinese Embassy in Belgrade
2008 Olympics Torch Relay in Paris
2012 Diaoyu Island
An economic theory of nationalism (I)




Forming a nation:
Benefits: lower costs of domestic trade, larger domestic
market
Cost: living with people with different cultures or races.
Equilibrium: optimal size of a nation
Alesina, Spolaore, and Wacziarg (2000)
An economic theory of nationalism (II)

Uneven globalization shocks in regions
within a nation

The shocked regions rely less on the domestic market
Disagreement with the current country size as optimal
(prefer a smaller country)
Less nationalistic


Empirical Evidence

A one standard deviation increase in the economic openness
reduces nationalism by 0.2 standard deviation

Across 200 cities in China in 2009: the unique data of Chinese
Political Compass (CPoC)
Before and after the membership of WTO, across 20 provinces in
China from 2001 to 2007: the World Value Survey
Across 15 countries from 2001 to 2007: the World Value Survey


Channels: not through protectionism, pride of national culture, or
other related political ideologies
Literature: non-economic consequences of international trade


Wars: Martin, Mayer and Thoenig (2008, 2012); Skaperdas and
Syropoulos (2001)
Military spending: Acemoglu and Yared (2010)

Institutional quality: Acemoglu, Johnson, Robinson (2005)
Slavery trade and mistrust in Africa: Nunn and Wantchekon (2011)

Of course, a large literature on protectionism

Literature: nationalism affects economic behavior




Boycott of French wines in the U.S. in 2003, no effect on sales:
Ashenfelter, Ciccarella, and Shatz (2007)
Boycott of France in the U.S. reduced bilateral trade: Michaels and
Zhi (2010)
Boycott of French cars in China in 2008: sales fell. Hong et.al.
(2011)
Adverse shocks of Sino-Japan relations on the bilateral trade in
2005 and 2010: Fisman, Hamao, and Wang (2012)
A large literature of nationalism in other social sciences

Lots of case studies, general weak empirical evidence

A model from the perspective of economics
empirical tests



Our focus on the within-country analysis: history,
ethnicity, geopolitics, and trade policy are constant.
Separate different channels
The Model (I)

N symmetric regions, each of which (i) uses a local specific factor K(i) to
produce a tradable good X(i).

Each region also makes and consumes a non-tradable aggregate good:

Within-border trade is costless. Cross-border trade incurs an iceberg cost

Utility function: consumption increases with the size of country S(i)
The Model (II)

Optimal country size

Nationalism in Home is the endorsement of the S* as the best size of the
domestic market.

Congruent with nationalism defined in social sciences
Handler (1988) on nationalism
“Nationalism is an ideology about individuated being. It is an ideology
concerned with boundedness, continuity, and homogeneity
encompassing diversity [...] [O]ur notions of “nation” and “state” imply
similar senses of boundedness, continuity, and homogeneity
encompassing diversity. The state is viewed as a rational,
instrumental, power-concentrating organization. The nation is
imagined to represent less calculating, more sentimental aspects of
collective reality. Yet both are, in principle, integrated: well-organized
and precisely delimited social organisms. And, in principle, the two
coincide.”
The Model (III)

A globalization shock hits some regions in each country:

the optimal size of the country for the these regions is now:

Two effects: a growth effect and a “relative-importance” effect

Conditional on income, globalization-hit regions have weaker nationalism
because it no longer endorses the same domestic market as other
regions
Through the economic openness: imports+exports

Illustration 1
Home
Foreign
(Country
formation)
Home
Foreign
(Globalization)
15
Illustration 3
Home
(Globalization
weakens nationalism)
16
Chinese Political Compass


the prototype: the UK website Political Compass
test questions: customized to the Chinese socio-economic context.

appraise 50 statements on a four-point scale. Responses mapped using
an algorithm to generate a test report.
– authoritarianism vs. libertarianism (politically)
– conservatism vs. liberalism
(culturally)
– collectivism vs. neoliberalism
(economically)

Strictly anonymous; IP addresses are identified; year 2009; mainland
only.
The CPoC Self-Evaluation (I)
The CPoC Self-Evaluation (II): [-2, 2] from left to right
The CPoC Self-Evaluation (III): Compare yourself with
other people
Political regime
Economy
Culture
20
Representativeness of the CPoC sample

Representative of the Chinese population? No.

Representative of the Chinese internet users?
29% of the whole population in 2010
Younger, richer, more educated, more urban residents
Current or future Chinese middle class




Altogether, more than 300,000 participants in CPoC
In our 2009 sample, 54,602 participants with a mainland IP address,
across 303 cities/prefectures (out of 333 cities)

Various checks on the sample selection

Measure Nationalism
[N1] assertion of the national unity (Gellner 1999)
National unity and territorial integrity are the interests of paramount priority for a
society. (国家的统一和领土完整是社会的最高利益)

[N2] protecting national interests from other countries (Hobsbawm,1990)
Given sufficient comprehensive national power, China has the right to take any
measure to protect its interests. (如果国家综合实力许可，那么中国有权为了维护自
己的利益而采取任何行动)

[N3] militarism (Posen, 1993)
All students, regardless of whether they are in college, high school, or
elementary school, should attend the military training arranged by the
government.(无论中小学生或大学生，都应参加由国家统一安排的军训)

[N4] anti-foreign sentiments (Gries, 2005)
Western countries, headed by the United States, will not really allow China to
become a world-class powerful nation.(以美国为首的西方国家不可能真正容许中国
崛起成为一流强国。)

The Index of Nationalism: simple average of the four
Correlation matrix
N1
N2
N3
N4
Number of Observations=54,602
N1
N2
N3
1
0.339
1
0.392
0.237
1
0.384
0.299
0.289
Principle Component Analysis
N4
1
The Distribution of Chinese Nationalism
(Mean=2.907, Standard Deviation=0.123)
OLS Results: a one standard deviation increase in the economic
openness reduces nationalism by 0.2 standard deviation
(number of obs=200)
Econ-open
Nationalism
index
Nationalism
index
Nationalism
index
Nationalism
index
-0.124***
-0.102**
-0.113**
-0.113**
(0.037)
(0.039)
(0.053)
(0.055)
0.036
0.042
0.042*
(0.022)
(0.027)
(0.025)
Geo-open
Other control
NO
NO
YES
YES
Political
propaganda
NO
NO
No
-0.067
(0.209)
Other controls: GDP per capita, College education, Rural population, Gov’t
budget in GDP
On Political Propaganda
“Official ideologies of states and movements are
not guides to what it is in the minds of even the
most loyal citizens or supporters.”
Eric Hobsbawm (1992)
Nationalism and Protectionism
Measure Protectionism:
 [P1] High tariffs should be imposed on foreign counterparts of domestic
products, in order to protect national industries (应当对国外同类产品征收
高额关税来保护国内民族工业)
 [P2] Foreign capital in China should be restricted from developing at
(their) will.(在华外国资本应该受到限制，不能任意发展)

Regress the index of nationalism on P1 and P2, then take residuals
Nationalism residuals
Econ-open
-0.106**
(0.049)
All other controls
YES
Instrumental Variable




Each city sells and buys different products from different trade partners
The prediction power of the gravity equation on trade volumes
IV: distance-weighted GDP of trade partners
Data: product-destination-city-year trade data from the China Customs
Statistical Information Service
Exogeneity: the GDP of foreign countries is beyond the control of a
certain city; constant trade policies across China
 Concern: nationalism drives the selection of trade partners?
(various robustness check: without the US, Japan, the EU, OECD countries,
War8 countries, NATO countries etc.)

2SLS Results: a one standard deviation increase in the economic
openness reduces nationalism by 0.27 standard deviation
(number of obs=200)
Nationalism Nationalism Nationalism Nationalism Nationalism
index
index
index
index
Residuals
Econ-open
-0.158***
-0.138***
-0.160**
-0.159**
-0.158**
(0.037)
(0.046)
(0.065)
(0.065)
(0.063)
0.028
0.034
0.034
0.032
(0.025)
(0.028)
(0.028)
(0.024)
Geo-open
Other
control
NO
NO
YES
YES
YES
Political
propaganda
NO
NO
No
YES
YES
First-stage
11.434***
10.587***
8.559***
8.606***
8.606***
(1.224)
(1.212)
(1.290)
(1.285)
(1.285)
Cultural Origin of Nationalism: Smith (1999), Anderson (2001)
Measure attitudes towards Chinese traditional culture

[Chinese Medicine] Traditional Chinese medicine exceeds modern mainstream
medicine in a number of ways. (中国传统医学具有一些现代主流医学不能比拟的
优势)

[Zhou-yi] Zhou-yi and Ba-gua are great achievements of our forefathers and
can explain a wide range of phenomena. (周易八卦是老祖宗智慧的结晶，能够
有效的解释很多事情)

[Confucianism] Modern China needs Confucianism. (现代中国社会需要儒家思想
)
The channel is not culture
Medicine
Zhou-yi
OLS
OLS
2SLS
-0.036
-0.012
2SLS
Confucianism
Nationalism
Residuals
OLS
OLS
2SLS
ECON
-0.054
OPEN
(0.055) (0.073) (0.038) (0.056) (0.057) (0.072) (0.054)
Not
0.075
0.005
0.021
2SLS
-0.098* -0.179***
(0.067)
through channels of other related political ideologies either:
collectivism, populism, conservatism etc.
Other concerns: driven by extreme opinions…
Analysis within China, over time





Control for regional fixed effects
The natural experiment of becoming a member of the WTO before
and after 2001
Data: World Value Survey in China (WVS) in 2001 and 2007
Advantages: the independent data source; widely used; individuallevel demographic variables
Disadvantages: small number of observations (2, 038); provincelevel analysis
The Difference-in-difference Strategy

Unbalanced effects of the WTO membership

High (trade) exposure regions:
the Yangtze River Delta and the Pearl River Delta
two trade centers: Shanghai and Hongkong
4 provinces located in the two deltas: Shanghai, Zhejiang,
Jiangsu, Guangdong




Low (trade) exposure regions: other 16 provinces
1.5
Unbalanced Effect of the WTO on the Trade Volume
openness
1
WTO Membership
.5
High-exposure provinces
0
Low-exposure provinces
1995
2001
2007
year
2010
Low-exposure provinces
1.9
2
nationalism
2.1
2.2
Unbalanced Change in Nationalism
1.8
High-exposure provinces
2001
2007
year
DID Estimates
(individual demographics: gender, age, marital status,
education, income)
Province-level
High-exposure dummy
Individual-level
0.077
0.057
(0.051)
(0.059)
-0.103**
-0.076
(0.047)
(0.049)
-0.155**
-0.215***
(0.072)
(0.064)
Log(GDP per capita)
YES
YES
Demographics
NO
YES
Number of obs
40
1,806
After WTO dummy
Interaction
Fixed-Effects estimates: a one standard deviation increase in
the economic openness reduces nationalism by 0.21 standard
deviation
Cross-Country Analysis





The rationale applies to cross-country analysis
Must control for country fixed effects
infer causality? trade policy could be endogenous
WVS 2001 and 2007, 15 countries
Data of GDP and trade: the Penn World Table
(v 7.0)
2.4
Levels of Nationalism in 2007
Turkey
2.2
USA
India
2
MoroccoChina
Mexico
South Africa
Canada
Indonesia
1.8
Spain
1.6
Korea
Sweden
Argentina
Japan
Chile
.2
Change in nationalism v.s. change in openness
2001 to 2007
Turkey
.1
USA
Japan
Canada
South Africa
Mexico
Spain
0
Chile
Indonesia
Argentina
Sweden
-.1
Korea
India
-.2
Morocco
China
-.2
-.1
0
.1
detrended change of openness from 2001 to 2007
.2
Turkey and Argentina
Turkey


Full membership negotiation with the EU started in 2004, accompanied by
the high uncertainty in Turkey of being finally accepted by the EU.
A sense of double-standards skepticism and mistrust significantly
strengthened the country's nationalism (Keyman and Yilmaz, 2006)
Argentina


economic collapse in 2001, default on external debts in 2002 seriously
damaged its national pride.
“for a while it was as though Argentina had no possible future and was
doomed to descend into poverty and shame” (Domingues, 2006).
Country Fixed Effects Estimates: a one standard deviation
increase in the economic openness reduces nationalism by 0.25
standard deviation
Country-level
Econ-open
Individual-level
-0.648**
-0.708**
(0.274)
(0.289)
Country Fixed Effects
YES
YES
Year Fixed Effects
YES
YES
Log(GDP per capita)
NO
YES
Demographics
NO
YES
Number of obs
30
73,866
Concluding Remarks
(On Scottish independence) ... just 21% of Scots would favour
independence if it would leave them £500 ($795) a year worse
off, and only 24% would vote to stay in the union even if they
would be less well off sticking with Britain. Almost everyone else
would vote for independence if it brought in roughly enough
money to buy a new iPad, and against it if not.
The Economist, April 14th, 2012