Soft power and exports

• Published date: 01 November 2019
• Author: Andrew K. Rose
• Date: 01 November 2019
• DOI: http://doi.org/10.1111/roie.12435

Rev Int Econ. 2019;27:1573–1590. wileyonlinelibrary.com/journal/roie

|

1573

© 2019 John Wiley & Sons Ltd

Received: 30 May 2019

|

Revised: 1 July 2019

|

Accepted: 8 July 2019

DOI: 10.1111/roie.12435

ORIGINAL ARTICLE

Soft power and exports

Andrew K.Rose

Haas School of Business,University of

California, Berkeley, CA

Correspondence

Andrew K. Rose, Haas School of Business,

University of California, Berkeley, CA

94720‐1900.

E‐mail: arose@haas.berkeley.edu. http://

faculty.haas.berkeley.edu/arose

Abstract

This paper seeks to help establish a stylized fact: a country’s

exports rise when its leadership is approved by other coun-

tries. I show this using a standard gravity model of bilateral

exports, a panel of data from 2006 through 2017, and an an-

nual Gallup survey that asks people in up to 157 countries

whether they approve of the job performance of the leader-

ship of China, Germany, Russia, the United Kingdom and

the United States. Holding other things constant, a coun-

try’s exports are higher if its leadership is approved by the

importer—“soft power” promotes exports. The soft power

effect is statistically and economically significant; a 1%

increase in leadership approval raises exports by around

two‐thirds of a percent. This effect is reasonably robust, and

different measures of soft power deliver similar results. I con-

servatively estimate that the >20 percentage point decline in

foreign approval of American leadership between 2016 (the

final year of Obama’s presidency) and 2017 (Trump’s first

year) lowered American exports by at least U.S.$3 billion.

JEL CLASSIFICATION

F14; F59

1

|

INTRODUCTION AND MOTIVATION

The objective of this paper is modest: it tries to help establish an empirical linkage between the pop-

ularity of a country’s leadership and international trade flows. Using standard techniques, I find that

countries whose leadership is approved abroad tend to sell more exports, holding all else equal. More

generally, I hope to demonstrate that “soft power”—the ability of one country to attract or persuade

citizens in another—is a significant determinant of export demand. Countries that are more favorably

viewed by foreigners tend to sell more.

1574

|

ROSE

2

|

METHODOLOGY AND GRAVITY DATA

I am interested in whether countries are affected in any tangible way by fluctuations in soft power.

In particular, I test whether changes in foreign perceptions of soft power affect export sales, all else

equal. In doing so, I hope to establish a new stylized fact linking trade to soft power.

I use a standard gravity model of international trade to account for other influences on bilateral

exports besides soft power. In particular, I pursue “theory‐consistent estimation” of the gravity equa-

tion, closely following the suggestions in the authoritative survey by Head and Mayer (2014). I focus

on their “LSDV” (least squares with time‐varying country dummy variables) technique that they show

works well in many situations. I estimate:

where ln(Xijt) denotes the natural logarithm of the nominal value of bilateral exports from country i to

country j at time t, measured as the average of FOB exports from i to j and CIF imports into j from i;

SOFTPOWERijt denotes a bilateral measure of the soft power that i is perceived to have in country j

at time t; D is the great‐circle distance between i and j; Lang is a binary variable that is unity if i and

j have a common language and zero otherwise; Cont is unity if i and j share a land border and zero

otherwise; RTA is unity if i and j belong to the same regional trade agreement and zero otherwise;

Colony is unity if i colonized j or vice versa and zero otherwise; β is a vector of nuisance coefficients;

{λit} is a complete set of time‐varying exporter binary variable fixed effects; {ψjt} is a complete set of

time‐varying importer binary variable fixed effects; and εij represents the myriad other influences on

exports, assumed to be well behaved.

The coefficient of interest is γ, the effect (on bilateral exports) of the exporter’s soft power over the

importer, ceteris paribus. I estimate this equation with least squares, using standard errors robust to

clustering by dyadic (ij) pair.

In practice, most of the variation in exports is explained by the “monadic” country–time fixed ef-

fects (one set each for the exporter and importer), which control a host of other influences on bilateral

exports. For instance, any general short‐term effect on American sales arising from the 2017 ascen-

dance of President Trump is accounted for by the 2017 American exporter fixed effect. Similarly,

any effect on Egyptian imports arising from the 2011 Arab Spring is taken out by the 2011 Egyptian

importer fixed effect. Anything that is specific and common to a country and a year—such as the size

of its economy, population, culture, policy uncertainty, leadership, or military spending, for either the

exporter or the importer—is accounted for by the fixed effects. Any effect of bilateral soft power on

exports is estimated conditional on—and above and beyond—both these fixed effects, and the other

determinants in Equation 1. This is a demanding requirement.

2.1

|

Gravity data

I rely on annual trade data drawn from the Direction of Trade data set assembled by the International

Monetary Fund. The data set covers bilateral trade between over 200 IMF country codes between

1948 and 2017.1

Bilateral trade on FOB exports and CIF imports is recorded in U.S. dollars. To this,

I add a number of dyadic variables necessary to estimate the gravity model, including distance, the

presence of a common land border and language, colonial history, and date of independence (these

are taken from the CIA’s World Factbook). I obtain data from the World Trade Organization to create

(1)

ln (

Xijt

)

=gSOFTPOWERijt +𝛽1ln

(

Dij

)

+𝛽2Langij +𝛽3Contij +𝛽4RTA

ijt

+𝛽

5

Colony

ij

+

{

𝜆

it}

+

{

𝜓

jt}

+𝜀

ijt