Voter and Legislator Responses to Localized Trade Shocks from China in Brazil

• Author: Daniela Campello,Francisco Urdinez
• DOI: 10.1177/0010414020970233
• Published date: 01 June 2021
• Date: 01 June 2021
• Subject Matter: Articles

https://doi.org/10.1177/0010414020970233

Comparative Political Studies

2021, Vol. 54(7) 1131 –1162

© The Author(s) 2020

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/0010414020970233

journals.sagepub.com/home/cps

Article

Voter and Legislator

Responses to Localized

Trade Shocks from

China in Brazil

Daniela Campello1

and Francisco Urdinez2

Abstract

This paper examines whether localized trade shocks from China influence

Brazilians’ views on integration with the country. We test the following

hypotheses: (1) as trade shocks are localized, views on trade should form

at the local, rather than at the individual level, and (2) as localized trade

shocks affect both workers and companies in a same region, they should

also influence legislators’ views on China. Our analyses find support for both

claims, but only among losers from Chinese trade. Residents and legislators

from localities hurt by import shocks tend to hold negative views about

economic ties with China, whereas neither residents nor legislators from

localities benefitted by export shocks exhibit more positive views about the

country. Our paper contributes to the literature on the politics of trade

by incorporating meso-level theories of trade preference formation and

by establishing the conditions under which the interests of constituencies

should shape legislators’ views on international trade.

Keywords

globalization, trade preferences, Brazil, China, trade shocks

1Getúlio Vargas Foundation, Rio de Janeiro, Brazil

2Pontificia Universidad Católica de Chile, Santiago, Chile

Corresponding Author:

Francisco Urdinez, Political Science Institute and Center for International Studies, Pontificia

Universidad Católica de Chile, Campus San Joaquin, Macul 7820436, Chile.

Email: furdinez@uc.cl

970233CPSXXX10.1177/0010414020970233Comparative Political StudiesCampello and Urdinez

research-article2020

1132 Comparative Political Studies 54(7)

After more than a decade of careful empirical research, there is little evi-

dence that self-interest drives individuals’ views on trade (Hainmueller &

Hiscox, 2006; Kaltenthaler & Miller, 2014; Sabet, 2016). Likewise, there is

weak support for the expectation that voter self-interest influences legisla-

tors’ sentiment towards international trade (Fordham & McKeown, 2003).

These results have increasingly challenged the Open Economy Politics

(OEP) paradigm (Lake, 2009), prompting new venues of research on the

determinants of trade attitudes.

At the individual level, many authors have turned their attention to the

role of identities and values such as cosmopolitanism, or tolerance, on

trade preferences (Hafner-Burton et al., 2017; Mansfield & Mutz, 2009).

Others have resorted to experimental studies to demonstrate that self-inter-

est does matter, but only as long as individuals are aware of it, which is not

always true (Butler & Nickerson, 2011; Rho & Tomz, 2017). At the legis-

lative level, concerns for foreign policy (Milner & Tingley, 2011), igno-

rance of voter preferences (Butler & Nickerson, 2011) and conflicting

preferences between business and labor within a constituency (Karol,

2007) are frequent explanations for the disconnection between constituent

self-interest and representatives’ views on trade.

In this paper, we join a strand of the literature that maintains that the miss-

ing link between interests and trade views lies on OEP’s unrealistically nar-

row understanding of self-interest (Fordham & Kleinberg, 2012; Mansfield

& Mutz, 2009). We argue that when trade shocks occur along sectoral lines,

and sectors are geographically clustered—when trade shocks are therefore

localized—, self-interest is determined by income shocks that occur at the

local level. As such, their effects are not restricted to those employed in the

affected industries but spread to other residents working in the service sector,

a logic recently explored by Ardanaz et al. (2013). We further contend that,

because localized trade shocks affect the interests of business and residents

within a same geographic constituency, they are also likely to shape their

representatives’ views on international trade.

We test these claims in a study of the impact of trade shocks from China

on Brazilians’ perceptions of the risks and opportunities associated with the

country, both at mass and legislative levels. We want to know whether people

are able to connect income shocks related to trade and particular trading part-

ners and their own material well-being to form the attitudes we hypothesize,

that is, whether residents and representatives from localities harmed (bene-

fited) by Chinese trade see ties with the country more negatively (positively)

than those living in and representing unaffected areas.

In doing so, we extend a recent literature that broadens the agenda on

determinants of trade views—originally focused on support for liberalization

Campello and Urdinez 1133

vs. protectionism—towards understanding how phenomena such as voting

patterns (Jensen et al., 2017; Margalit, 2011), support for nationalist and iso-

lationist parties (Colantone & Stanig, 2018b), and views of other states as

friendly or threatening (Kleinberg & Fordham, 2010) vary between winners

and losers from trade openness.

Our empirical strategy combines multiple sources, including recently pub-

lished estimates of localized trade shocks from China in Brazil (Costa et al.,

2016), surveys of legislators and of public opinion, electoral and census data,

and content analyses of legislative speeches. Our unity of analysis, to which

we also refer as a locality, is the microregion1—a metropolitan area that inte-

grates municipalities into local economies and local labor markets and is

therefore ideal to capture local labor market effects of trade. As in 2017, the

Brazilian Institute of Geography and Statistics (IBGE) identified a total of

558 microregions (after 2017 named immediate geographical regions) and

5,570 municipalities in Brazil.

Results reveal that localized trade shocks from China do influence

Brazilians’ views about the country, but this effect is restricted to losers from

trade. Whereas residents in localities harmed by large import shocks are more

likely to see ties with China as posing a risk to Brazil, there is no evidence

that residents in areas benefited by export shocks are any more inclined to see

it as an opportunity. This impact is magnified in the case of respondents who

explicitly identify China as a major competitor to local production, suggest-

ing that the connection between interests and risk perceptions is not spurious.

This pattern reinforces previous findings in the economics literature (Baldwin

& Robert-Nicoud, 2007; Mayda & Rodrik, 2005) and is consistent with pros-

pect theory’s proposition that losses loom larger than equivalent gains (Jervis,

1992; Kahneman & Tversky, 1979).2

Self-interest strictly determined by factor ownership or sector of employ-

ment is not relevant under any specification. Rather, a positive association

between skill (a scarce resource in Brazil) and pro-China responses suggests

that education is less an expression of factor ownership, and more of a cos-

mopolitan attitude, possibly associated with personal exposure to pro-trade

arguments and/or to other nations and cultures (Ardanaz et al., 2013;

Hainmueller & Hiscox, 2006; Kleinberg & Fordham, 2010).

We obtain strikingly similar results at the legislator level; constituency

interests shape legislator views of economic ties with China regardless of

how we measure them. Legislators whose constituents suffered import

shocks are more inclined to portray ties with the country as a risk. Likewise,

representing losing constituents increases the probability that a legislator

joins the Brazil-China Legislative Caucus, created to set the legislative

bilateral agenda between the countries. These effects are also observable in