Exposure to Offshoring and the Politics of Trade Liberalization:
Debate and Votes on Free Trade Agreements in the US House of
Representatives, 2001–2006
ERICA OWEN
Texas A&M University
The movement of jobs overseas—known as “offshoring”—is one of the most politically charged aspects of globalization in
developed countries. The high salience of offshoring reflects growth in competitive pressures from globalization that directly
target individuals’ jobs. I argue that, in a world of fragmented production, how vulnerable one’s occupation is to offshoring
helps explain patterns of protectionist sentiment not otherwise accounted for by existing work. I expect that the
negative consequences of offshoring for workers will be particularly salient. As a result, vulnerability to offshoring is likely
to generate protectionist sentiment toward free trade among constituents. Therefore, legislators whose constituents are vulnerable
to offshoring should prove more likely to oppose trade liberalization. I measure vulnerability to offshoring at the
district level using data from the US Census. I analyze roll call votes on free trade in the House of Representatives between
2001 and 2006. I find that the larger the share of their constituents who are vulnerable to offshoring, the more likely legislators
are to vote against free trade and to discuss the costs of trade for workers in floor debates. My results suggest that the
dynamics of offshoring constitute an overlooked dimension to the political economy of trade.
Scholars usually treat the political economies of trade and
of foreign direct investment (FDI) as separate subjects.
However, the increasing role of multinational firms in global
production networks, and concomitant rise in intra-firm
trade flows, blurs the distinction between the two (Pandya
2016, 470).1
Trade increasingly involves the exchange of
intermediate goods and services, with value added in a number
of different countries (Grossman and Rossi-Hansberg
2008, 1978). This fragmentation of production generates
new distributional pressures and exposes previously sheltered
segments of the economy to international economic
competition. These new pressures have important implications
for the mass politics of globalization, including voters’
demands for protection or redistribution at the individual2
and aggregate levels,3
as well as how elected representatives
craft policies in response to those demands.4
Offshoring5
is one of the most visible and politicized aspects
of globalization in the United States and other
developed countries (e.g., Blinder 2006; Drezner 2004;
Mankiw and Swagel 2006). A recent study finds that a majority
of the US public consistently opposes offshoring
(Mansfield and Mutz 2013, 597). Offshoring helps drive
protectionist sentiment and rhetoric in US politics. In
1992, Ross Perot suggested that NAFTA would lead to a
“giant sucking sound” of American jobs moving across the
border. In 2004, presidential candidate Senator John
Kerry denounced “Benedict Arnold CEOs” that ship
American jobs overseas. Offshoring was again an important
issue in the 2012 presidential election. Concerns
about offshoring played a major role in the 2016 presidential
primaries and general election. Indeed, President
Trump promised, and has pursued, policies to halt and reverse
offshoring, in part by renegotiating free trade agreements
(FTAs). Furthermore, systematic research provides
evidence that offshoring generates electoral costs for incumbents
(Margalit 2011).6
Offshoring activities have increased not only in manu-
facturing,7
but also across a broader range of more skillintensive
activities, including service activities, which
scholars have traditionally viewed as non-tradable (see
Hays 2009; Iversen and Wren 1998).8
Thanks to declining
transportation and telecommunication costs, firms are increasingly
able to separate the production of a good or
service over space and time, which in turn expands the
scope of tradable jobs (see Blinder 2007; Feenstra 2010;
Jensen and Kletzer 2005). One widely cited estimate suggests
that 29 percent of jobs in the US workforce are
potentially offshorable (Blinder 2007), implying that
Erica Owen is an assistant professor in the Department of Political Science
at Texas A&M University. She is also a visiting researcher at the University of
Zurich during Fall 2016. She was previously a Postdoctoral Fellow at the
Niehaus Center for Globalization and Governance at Princeton University.
Author’s note: Part of this work was done while I was a visiting researcher at
the University of Zurich. I thank Despina Alexanidou, Bill Clark, John
Freeman, Matt Fuhrmann, Ben Graham, Lloyd Gruber, Kim Hill, Nan Jia,
Noel Johnston, Yoo Sun Jung, Alison Kingsley, Mike Koch, Quan Li, Pablo
Pinto, Dennis Quinn, Weiyi Shi, participants in the Program on International
Conflict and Cooperation at Texas A&M APSA 2013, IPES 2013, the anonymous
reviewers, and the editors of ISQ for helpful suggestions and comments.
Special thanks to Will Lowe for assistance with text analysis.
1
For recent work on multinationals and trade, see Jensen, Quinn, and
Weymouth (2015b) and Baccini, Pinto, and Weymouth 2017.
2
See Margalit (2012); Owen and Johnston (Forthcoming); Walter (2017).
3
See Owen and Quinn (2016).
4
See Burgoon (2012); Hellwig (2015); Hwang and Lee (2014).
5
Although the terms “outsourcing” and “offshoring” are sometimes used
interchangeably, offshoring is the correct term to describe the phenomena of
interest because it refers to production located abroad regardless of the
boundary of the firm. In contrast, outsourcing refers to production that
occurs outside the boundary of the firm regardless of location (for example,
customer service call centers located in the United States or India).
6
See related work by Jensen, Quinn, and Weymouth (2015a).
7
Materials offshoring is the subject of a large literature (Feenstra 2010;
Feenstra and Hanson 1999).
8
For important exceptions, see Chase (2008) and Wren and Rehm (2013).
Owen, Erica. (2017) Exposure to Offshoring and the Politics of Trade Liberalization: Debate and Votes on Free Trade Agreements in the US House of Representatives, 2001–2006.
International Studies Quarterly, doi: 10.1093/isq/sqx020
VC The Author (2017). Published by Oxford University Press on behalf of the International Studies Association.
All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
International Studies Quarterly (2017) 61, 297–311
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
offshoring has the potential to impact the welfare of a
large share of the US labor force.9
This is also likely to be
true in other developed countries (see Walter 2017).
The continued salience of offshoring reflects, in part,
the growing importance of competitive pressures from
globalization directed toward individual jobs rather than
toward entire industries or individuals of different skill levels.
The ability of firms to split up the production chain
means that individuals in different occupations employed
in the same industry or even in the same firm may face
different competitive pressures from globalization. For example,
consider the offshoring of software development
by firms like Microsoft, and of accounting by firms like
Toys “R” Us and New York Life (Preston 2015). Research
by trade and labor economists highlights the importance
of occupation characteristics—including offshorability—
in shaping labor market outcomes in a global economy
(see for example Blinder 2007; Ebenstein et al. 2014;
Jensen 2011). Yet occupation characteristics, which play a
central role in the comparative political economy literature
as a site of preference formation (Kitschelt and
Rehm 2014; Rehm 2009; Wren and Rehm 2013), are
largely ignored by scholars of international political
economy.10
Drawing on the recent trade literature, I introduce vulnerability
to offshoring—an occupation characteristic—to
explain new patterns of protectionist sentiment, including
the circumstances in which skilled workers in exporting
industries (like software developers) or service occupations
(like accountants) would be threatened by trade
liberalization. In contrast to less educated, blue-collar
workers who have faced competition from trade for several
decades, many newly offshorable occupations are
higher-skill, higher-wage white-collar jobs. Yet other
skilled workers, such as those in managerial positions
(like CEOs), as well as less skilled workers in childcare or
food service, for instance, provide location-specific services.
They therefore remain relatively invulnerable to offshoring.
As a result, occupation-based exposure to trade
differs from pressures based on skill level, industry of employment,
or firm (see Blinder 2007; Ebenstein et al.
2014; Oldenski 2014).
What does this mean for the politics of trade liberalization?
I argue that vulnerability to offshoring is an
important source of protectionist sentiment over trade liberalization
in developed countries. Although there are individual
winners and losers from offshoring, I argue that
the aggregate vulnerability to offshoring is a key source of
protectionist sentiment among constituents. As politicians
who seek re-election, legislators must factor this vulnerability
into their trade policy positions, because liberalization
facilitates offshoring by reducing the costs of crossborder
production. Thus, legislators with constituents vulnerable
to offshoring will prove more likely to oppose
trade liberalization. My argument accounts for patterns of
protectionist sentiment at odds with the expectations of
both the factor endowments model—which suggests that
less skilled workers in developed countries will oppose
trade liberalization—and the sectoral model—which
predicts industry-based cleavages in which workers in
import-competing industries will oppose trade
liberalization.
My argument implies that legislators are, first, more
likely to vote against trade liberalization and, second,
more likely to discuss the costs of trade for labor, when a
larger share of their constituents are vulnerable to offshoring.
I test my claims by examining support for free trade
in roll call votes and in speeches for eight bills on FTAs
and trade promotion authority for the US House of
Representatives between 2001 and 2006. FTAs are one of
the most prominent forms of trade liberalization and thus
provide an ideal context to examine the influence of offshoring
on the political economy of trade liberalization. I
choose this period because it corresponds to the emergence
of political concern about offshoring in services
(Guisinger 2017; Mankiw and Swagel 2006; Margalit
2011), and because the United States signed FTAs with
both developed and developing countries during this
time. I use US Census data, in combination with the offshorability
classification developed by Blinder (2007), to
calculate a measure of constituency exposure to offshoring
at the congressional district level. As my argument expects,
I find that when their constituents are more vulnerable
to offshoring, representatives are less likely to vote in
favor of FTAs and more likely to discuss the costs of trade
for workers.
Related Literature
How do constituents’ interests shape legislators’ positions
on trade policy? A number of studies examine the determinants
of legislators’ support for free trade or protection
in the context of roll call votes in the US House of
Representatives and Senate.11
This literature posits that
politicians, motivated by the desire for re-election, must
balance the benefits of trade for the economy as a whole
against the costs and benefits borne by specific economic
interests.
These economic interests may be organized or diffuse
(Grossman and Helpman 2002; Olson 1971). Many scholars
examine the demand for trade policy by interest
groups, including labor unions.12
More relevant to this
paper, however, is work on diffuse labor interests.
Scholars mainly draw from two competing theories of the
material interests: Heckscher-Ohlin and Ricardo–Viner.13
Heckscher-Ohlin expects that coalitions in favor of—and
in opposition to—trade form along factor lines (Rogowski
1989). It expects that owners of relatively abundant factors
will benefit from trade, while owners of relatively scarce
factors will be hurt by trade (Stolper and Samuelson
1941). Thus, cleavages will form between owners of abundant
factors in support of trade and owners of scarce factors
opposed to free trade. For instance, Hiscox (2002a,
2002b) examines the influence of class cleavages between
labor, capital, and land on trade votes between 1987 and
9
See also Jensen (2011). Actual offshoring numbers likely understate the
labor market consequences because we can never know the number of jobs
that would have otherwise been created in the home country if not for fragmented
production (Blinder 2009).
10
Exceptions include Hays (2009), Owen and Johnston (Forthcoming),
and Walter (2017).
11
See also recent work on position-taking in the form of bill sponsorship
(Fordham and Kleinberg 2012; Galantucci 2014).
12
Much of this literature draws on the protection-for-sale framework
(Grossman and Helpman 1994). See also Baldwin and Magee (2000),
Fordham and McKeown (2003), and McGillivray (2004). For emphasis on the
protectionist influence of organized labor interests, see Baldwin and Magee
(2000), Conybeare and Zinkula (1996), and Steagall and Jennings (1996).
13
Note that this approach is not limited to roll call votes on trade, but
also, for instance, support for international financial organizations (Broz
2008) as well as foreign aid (Milner and Tingley 2011).
298 Exposure to Offshoring and the Politics of Trade Liberalization
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
1994.14
Although the traditional factors in this framework
are land, labor, and capital, it is flexible enough to include
additional factors, such as skilled and unskilled
labor (Midford 1993; Rogowski 2006). In developed countries,
which enjoy an abundance of skilled labor and scarcities
of unskilled labor, cleavages should form between
pro-trade skilled workers and protectionist unskilled workers.
A large number of studies support this expectation.
They find that legislators are more likely to support free
trade when their constituency includes a greater share of
skilled labor (Bailey 2001; Choi 2015; Conconi, Facchini,
and Zanardi 2014; Ladewig 2006; Milner and Tingley
2011).
In contrast to factor-based coalitions, sectoral models
drawing from Ricardo-Viner anticipate trade coalitions
that form around industry interests. In particular, limited
mobility within the economy means some factors are
“stuck” in a given industry, and therefore individuals in
import-competing (exporting) industries are expected to
oppose (support) free trade. Thus, a larger share of constituents
in import-competing (exporting) industries will
lead to lower (greater) support for free trade (e.g., Alt
and Gilligan 1994; Choi 2015; Hiscox 2002a). As is well
known in the literature, the level of factor mobility determines
whether coalitions will be factor- or sector-based
(Hiscox 2002a, 2002b).15
Empirical evidence finds support
for both approaches.
The above summarizes the literature with a narrow
focus on constituency characteristics. Additional accounts
of legislators’ positions include legislator characteristics
like partisanship and ideology, the economic interests of
firms (Milner 1988; Nollen and Quinn 1994), foreign policy
goals (Milner and Tingley 2015), institutional rules
(Bailey, Goldstein, and Weingast 1997; Goldstein 1986),
differences between the House and Senate (Conconi,
Facchini, and Zanardi 2014; Karol 2007), and electoral incentives
(Feigenbaum and Hall 2015).
Theory
My theory of the political economy of trade with offshoring
draws on the trade in tasks literature as a framework.
This allows us to understand the political economy of
trade in a world with fragmented production. The key intuition
of the tasks approach is that every good and service
is the product of a combination of a set of tasks. By definition,
trade in tasks is possible when the tasks required to
produce a good or service can be unbundled; that is,
when certain steps of the domestic production process
can be performed by foreign labor. If trade in tasks is impossible,
goods will be wholly produced in a particular
country (Grossman and Rossi-Hansberg 2008). However,
when trade in tasks is possible, firms in the home country
can offshore segments of the production process of a
good—both exported and import-competing—to the foreign
country.16
The presence of a wage gap between workers in the
home and foreign country drives the extent to which offshoring
occurs.17
In the production of a good, firms will
offshore tasks performed by both high- and low-skill
labor up to the point that the marginal cost of producing
each task at home equals the cost of doing so
abroad (Grossman and Rossi-Hansberg 2008). Trade liberalization
reduces the cost of offshoring by making it
cheaper to import the intermediate inputs generated by
the offshored tasks.
What are the characteristics of tasks that can potentially
be offshored? Many scholars offer empirical definitions of
this concept (for example, Blinder 2007; Crino 2010;
Jensen and Kletzer 2005; Oldenski 2014).18
Following
Blinder (2007), I define offshorable tasks as those that do
not depend on a specific location and that do not require
face-to-face interaction; in other words, offshorable tasks
can be provided from a distance. Many manufacturing
tasks (for example assembly of clothing) and services tasks
(for example customer service, bookkeeping) can be
transported physically or over a wire. These tasks are offshorable,
and thus more exposed to trade. On the other
hand, location-specific tasks (for example farming) or
interpersonal tasks (for example hairstylists, personnel
management) are non-offshorable. In the remainder of
this section, I discuss offshorability and the welfare consequences
of trade, the aggregation of preferences with
respect to trade in tasks, and the impact on the political
economy of trade.
Offshorability and Trade Competition
The nature of cleavages due to competition from trade in
tasks depends on the level of labor market mobility.19
If
there is full labor market mobility, then the distributional
consequences of offshoring will fall along factor lines, because
workers reallocate skills to tasks in response to trade
pressures until wages within each skill level are the same
(Acemoglu and Autor 2011). In contrast, I expect that
labor market mobility is limited,20
with tasks that are clustered
within occupations. For instance, a computer
programmer (an occupation) may perform a number of
different tasks (testing and trials, updating code). As a result,
when labor market frictions make it difficult for workers
to switch occupations, occupation characteristics shape
who trade benefits and harms.21
Individuals in offshorable occupations are exposed to
competition from foreign labor. This puts downward pressure
on wages, increases the risk of displacement, and
reduces job security. Those in the most offshorable occupations
pay a wage penalty, even after controlling for education
and industry (Blinder 2007). Workers exposed to offshoring
at the occupation level experience large negative
wage effects (Ebenstein et al. 2014). Furthermore, individuals
in offshorable jobs are also more likely to lose their
job, and thus face additional costs due to short-term costs
of dislocation and possibly reduced future earnings
(Jensen and Kletzer 2005).
14
See also Ladewig (2006) for a similar approach.
15
See also the work of Choi (2015), Jeong (2009), and Ladewig (2006).
16
The trade in tasks model is particularly well suited to explain the distributional
consequences of increases in the extensive margin of trade (new
things are tradable), in contrast to the canonical models, which explain increases
in the intensive margin of trade (more trade in the existing set of
goods).
17
Models of offshoring often start from the premise that wages are lower
in the foreign (developing) country for workers of all skill levels relative to the
home (developed) country because of a technology gap or lagging productivity
(Grossman and Rossi-Hansberg 2008, 1988).
18
For a review of the tasks approach in labor economics, see Autor (2013).
19
Relaxing the assumption of labor market mobility has important implications
for distributional consequences, as shown by Hiscox (2002a).
20
For recent models of labor market frictions, see Davis and Harrigan
(2011), Egger and Kreickemeier (2009), and Helpman, Itskhoki, and Redding
(2010).
21
See Kambourov and Manovskii (2008), who demonstrate mobility across
occupations is lower than across industries.
ERICA OWEN 299
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
The costs of trade in tasks are even greater for those individuals
displaced by offshoring. Workers whose jobs are
moved offshore experience loss of income and reduced
earnings potential, and these workers are more likely to
be re-employed in a job of lower quality and pay.22
These
costs are greatest when individuals are forced to change
occupations. For instance, one study estimates that
occupation-switching due to trade competition leads to
wage losses of 12 to 17 percent (Ebenstein et al. 2014).
However, the possibility of trade in tasks also creates
opportunities for some workers to benefit through
onshoring (i.e., tasks that are exported). Although offshorability
is often associated with negative labor market
outcomes, this conflates the possibility that a task can be
provided from abroad with the likelihood that it will be
offshored. In particular, many developed countries have a
surplus in trade in services and would likely benefit from
further liberalization of services that are trade intensive in
certain high-skill, creative tasks (Acemoglu and Autor
2011; Jensen 2011). Thus, workers in some offshorable
occupations may benefit from the creation of jobs in the
home economy as a result of trade in tasks (Ebenstein
et al. 2014; Hummels et al. 2014).
Exactly who wins and loses from offshoring at the individual
level is the subject of a growing body of research.
Many scholars agree that workers in developed countries
who perform routine tasks are negatively affected by trade,
especially as exposure to offshoring increases (Acemoglu
and Autor 2011; Ebenstein et al. 2014; Goos, Manning,
and Salomons 2014; Oldenski 2014). Routine tasks follow
a set of rules or scripts, and thus can readily be taught to
foreign workers. For application to individual-level preferences,
see Owen and Johnston (Forthcoming), who find
that those in routine-task-intensive jobs are more protectionist,
especially as offshorability increases. Walter (2017)
offers an alternative account of the distributional consequences
of trade based on new new trade theory, in which
trade benefits high-skill workers in offshorable occupations
while harming low-skill workers in offshorable
occupations.
Aggregate Labor Preferences
Given that offshoring benefits some and harms others at
the individual level, how does constituency offshorability
shape aggregate preferences toward trade? In the aggregate,
I argue that higher levels of offshorability should be
associated with greater protectionist sentiment for two
main reasons. First, the literature on loss aversion
(Kahneman and Tversky 1979) demonstrates that individuals
place more weight on loss of income than gains of
the same amount. Therefore, the losses associated with
offshorability should receive more weight than the gains,
and we should expect those hurt by offshoring to care
more about trade policy (Freund and Ozden 2008; Tovar
2009).23
This can produce a status quo or protectionist
bias with respect to preferences (Conconi, Facchini, and
Zanardi 2014). This means that even if the number of winners
and losers from offshoring were exactly equal,24
the
protectionist sentiment generated by losers would be
higher than support for free trade generated by winners
(Guisinger 2017).
Second, in the short term, the costs generated by trade
in tasks are likely to outweigh the benefits. Even if a country
benefits from trade overall, this is typically a long-run
outcome, and there are large gross costs associated with
the reallocation of workers in the short run (Blinder
2009). For instance, job turnover due to trade causes
some people to bear the costs of unemployment. The predominance
of long-run full employment models means
that scholars have primarily emphasized the impact of
trade on wages and the composition of jobs (Jensen
2011). This overlooks the costs of dislocation and transition
(see Kletzer 2009), as well as the possibility that trade
may produce non-trivial spells of unemployment and reduce
long-run aggregate employment (Blinder 2009;
Go¨rg 2011). The impact of short-term costs and long-run
gains on aggregate preferences is exacerbated by the fact
that there may be greater uncertainty about the identity
of the winners than losers. This uncertainty can generate
support for protectionism even if welfare improves for the
majority (Conconi, Facchini, and Zanardi 2014).
Thus, if individuals exposed to trade via occupation offshorability
place more emphasis on the costs, and/or if in
the short run, the costs are greater than the benefits, then
offshorability will generate protectionist sentiment in the
aggregate. I assume that workers in non-offshorable occupations
do not hold a preference related to offshoring because
they are not exposed to this type of trade competi-
tion.25
Therefore, I expect that higher levels of
offshorability in a constituency will lead to greater protectionist
sentiment in the aggregate.
Offshoring in the Political Economy of Trade
Offshoring, an important aspect of the public’s concern
about free trade (Guisinger 2017), introduces a new dimension
to the political economy of trade because distributional
pressures occur along occupation lines. In line
with previous literature, I assume that politicians are motivated
by the desire for re-election and must balance the
benefits of trade for the economy as a whole with the costs
and benefits to certain economic actors. Workers’ interests
may be organized into unions or associations, or interests
may be diffuse and thus determined by a latent shared
interest with respect to trade (Grossman and Helpman
2002; Olson 1971).26
Due to the distributional pressures
of offshoring, we should expect both organized and diffuse
labor interests to form along occupation lines.
Although anecdotal evidence suggests there is some lobbying
by occupation-based interest groups, voters’ interests
as workers are central to politicians’ electoral calculus. I
therefore focus on diffuse channels of representation
(while controlling for contributions from labor groups
broadly defined). I return to the issue of occupationbased
interest groups below.
What incentives do elected officials have to care about
diffuse labor interests, and specifically those related to
22
See Davis and Harrigan (2011), who argue that trade reduces the number
of “good” jobs (those that pay a wage premium) in the economy.
23
This is consistent with the “loser’s paradox,” where special interests fight
more to avoid loss than to reap gain (Tovar 2009), and declining industries
are rewarded with more protection.
24
Expectations about aggregate gains from trade depend heavily on the assumptions
of the model of trade, especially the assumption of full employment
(Dean 2015; Egger and Kreickemeier 2009). When the assumption of
full employment is relaxed, trade can lead to an increase in long-run unemployment
(Egger and Kreickemeier 2009, 205). See also Helpman and
Itskhoki (2010).
25
However, there is evidence that offshoring is widely disliked by a majority
of people (Mansfield and Mutz 2013). Additional evidence suggests that individual
welfare expectations about trade are based on the impact of trade on
the local community (Guisinger 2017).
26
For a review of labor influence on trade policy, see Owen (2015a).
300 Exposure to Offshoring and the Politics of Trade Liberalization
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
offshorability? Looking at skilled labor, Bailey (2001)
argues that elected officials need to consider diffuse interests,
because these interests represent votes. Elected
officials must anticipate that other actors, like electoral
challengers or hostile interest groups, may try to mobilize
these interests. Following this logic, I argue that elected
officials’ positions on trade liberalization will vary systematically
according to how vulnerable their constituencies
are to offshoring because they must anticipate the possible
mobilization of these interests in future elections.27
Elected officials incorporate these interests into their behavior
when there is a latent shared interest and some
actor willing to mobilize the latent interest if neglected
(Bailey 2001). The case of offshoring fits these two conditions
because there is clear opposition to this dimension
of international trade and many examples of politicization
of the issue of offshoring. During the 2004 election, there
were 113 ads mentioning imports and outsourcing
(Guisinger 2017, 286). Campaign ads in battleground
states criticized President Bush for supporting outsourcing
(Mankiw and Swagel 2006). For example, an ad titled
“It’s About Jobs,” announced that: “During the past three
years it’s true George W. Bush has created more jobs.
Unfortunately, they were created in places like China.
Bush’s policies have encouraged the loss of nearly three
million jobs. He supported tax breaks to corporations that
shipped jobs overseas” (Appleman 2015). These examples
suggest that concerns about offshoring are salient and
that challengers raise the issue during campaigns.
Because legislators can preemptively respond to voters’
interests, mobilization does not even need to be an overt
process (Bailey 2001, 46). With offshorability shaping a latent
interest, elected officials have incentives to consider
the impact of trade on those in offshorable jobs when
forming trade policy positions.
This incentive remains even if citizens do not precisely
understand the mechanics of the relationship between
offshoring and trade or how these phenomena affect
them. It is necessary only that citizens link offshoring and
trade. Voters may not understand economic theory, but
instead may form opinions about free trade based on
what is happening in their lives, including the possibility
of job losses due to trade and offshoring (Freeman 2009,
67).28
Given the extent to which the public is concerned
about jobs being shipped overseas as a result of trade
(Mankiw and Swagel 2006, 1041), there are strong reasons
to suspect that attitudes toward offshoring will shape attitudes
toward FTAs. In support of this idea, Guisinger
(2017, 68) finds that preferences toward trade generally
and trade defined as offshore outsourcing follow
similar patterns. Additionally, Owen and Johnston
(Forthcoming) find that occupation characteristics—
including offshorability—shape preferences over trade.29
Therefore, legislators should be less likely to support
trade liberalization when their constituency is more vulnerable
to offshoring. Although a legislator’s position on
trade liberalization is unobserved, we can learn about legislators’
latent trade policy positions by examining
legislative speech and votes. Votes are the end result of
the legislative process, but position-taking is another important
part of the legislative process (Box-Steffensmeier,
Arnold, and Zorn 1997; Galantucci 2014; Grimmer 2013).
Speeches during legislative debates are intended for a
broad audience, including constituents, the media, and
other representatives (Pearson and Dancey 2011, 912). If
vulnerability to offshoring generates protectionist sentiment
in the aggregate, as I argue, then I should find support
for the following observable implications of the
argument:
H1: Legislators with a larger share of their constituency vulnerable
to offshoring will be less likely to vote in favor of trade
liberalization.
H2: Legislators with a larger share of their constituency vulnerable
to offshoring will be more likely to discuss the costs of trade for
labor during debates on liberalization.
Although this argument is not the first to suggest that
constituency vulnerability to trade affects legislators’ positions,
the offshoring model introduces a new dimension
of vulnerability based on occupation, and thus offers an
explanation of patterns of protectionism not captured by
factoral and sectoral models. As noted above, the factor
endowments model suggests that owners of the same
factor will have the same preferences over trade (for example,
low-skill workers all support protection). In the aggregate,
this implies that the influence of low-skill workers
on legislators’ policy positions will be homogeneous
across other dimensions (including offshorability or industry).
The offshoring model is based on the idea that
task content of occupations varies within skill levels30
and
that mobility across occupations is limited. It suggests that
within skill levels, there will be differences in attitudes toward
free trade and thus heterogeneity in terms of constituents’
influence on legislators’ policy positions within skill
levels.31
Even sectoral models, with limited mobility across
industries, do not capture the possibility that firms in an
export-oriented industry may offshore certain segments of
the production process, hurting some individuals even in
expanding or exporting industries.32
In contrast, the offshoring
model suggests that individuals in the same industry
could hold different preferences with respect to trade.
The potential explanatory power of the offshoring
model is best understood when considering interests that
might support free trade according to the canonical models,
but would not do so under the offshoring model.
For example, recent years have seen an increase in
occupation-based lobbying on trade and offshoring. The
Communication Workers of America, for instance, are
very active in lobbying to protect call center worker jobs
27
This is consistent with recent work by Feigenbaum and Hall (2015),
which suggests that members of Congress are responsive to constituents’ economic
interests on trade policy.
28
See also Guisinger (2017).
29
In contrast, Mansfield and Mutz (2013) do not find evidence of a link
between offshorability and trade preferences. This is likely due to the fact that
offshorability in and of itself will have an ambiguous effect on trade preferences
for individuals (Owen and Johnston Forthcoming; Walter 2017).
30
See Acemoglu and Autor (2011) and Oldenski (2014), who demonstrate
that canonical models of factor endowments cannot explain polarization of
labor markets in developed countries like the United States. For more extensive
discussion on the difference between tasks and skills, see Acemoglu and
Autor (2011) and Baumgarten, Geishecker, and Go¨rg (2013).
31
If the offshoring model is correct, we should therefore expect to see legislators
respond differently to low-skill workers in offshorable versus nonoffshorable
occupations. I discuss this further below in the Empirical Strategy
and Background section.
32
For instance, workers in business services occupations like bookkeeping,
data entry, and supply management face downward wage pressure from
abroad regardless of whether an industry is import-competing or exportoriented.
See Grossman and Rossi-Hansberg (2008).
ERICA OWEN 301
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
in the United States,33
including reported lobbying disclosures
on major trade legislation considered in the next
section of this paper (Center for the Study of Responsive
Politics 2010). Not only is this an example of occupationspecific
lobbying, but communications workers would
also be considered sheltered from trade in canonical models.
Highly skilled occupation associations also engage in
lobbying against trade. For instance, the International
Federation of Professional and Technical Engineers
also lists offshoring and trade as top concerns.34
Similarly,
the AFL-CIO Professional Employees Department lobbied
against the Chilean and Singaporean FTAs, as well as
H-1B visas. More recently, the American Institute of
Certified Public Accountants began lobbying on trade and
immigration (Center for Responsive Politics 2010). This
anecdotal evidence supports the idea that these groups
lobby with respect to occupation, and in ways that cannot
be explained by previous accounts of distributional
consequences.
To be clear, this article does not test microfoundations
of the trade in tasks model for workers. Rather it develops
a theory of how the politics of trade in the aggregate have
changed due to offshoring. If the underlying microfoundations
of the offshoring model are as posited (namely that
differences exist within skill levels and industries), then we
should see evidence consistent with the hypotheses above. I
therefore test the predictions of the offshoring model
alongside those suggested by the existing literature.
Finally, legislators must weigh interests regarding offshorability
against those who stand to benefit from a given
FTA, including consumers, exporters, and multina-
tionals.35
Arguments about pro-liberalization interests are
well established in the literature, and so I defer further
discussion of these factors to model specification.
Empirical Strategy and Background
I test my argument by looking at votes on free trade agreements
(FTAs) because FTAs are high-profile trade votes
(Mansfield and Milner 2012; Mansfield, Milner, and
Rosendorff 2002). I examine roll call votes and floor debates
pertaining to FTAs in the House of Representatives
during the 107th, 108th, and 109th Congresses from 2001
to 2006. In robustness checks, I examine roll call votes on
all trade bills during the same period.
Background on FTAs in the 107th–109th Congresses
This study primarily examines eight roll call votes and corresponding
floor debates across three Congresses for the
implementation of seven free trade agreements and one
bill on the adoption of Trade Promotion Authority
(TPA). The vote on Trade Promotion Authority (H.R.
3005) merits special discussion because it is not a vote on
a specific FTA, but rather a vote to grant the President the
power to negotiate reciprocal agreements with partner
countries (formerly known as “fast-track” authority).
Subject to certain consultation provisions, Congress must
then vote on the agreement without amendment or filibustering.
TPA is often viewed as essential to the successful
conclusion of FTAs (Conconi, Facchini, and Zanardi 2012).
Prior to 2001, fast-track authority was in effect from 1974 to
1994, but efforts to renew fast-track during the Clinton
administration were unsuccessful. During his campaign,
President George W. Bush stated that the renewal of TPA
would be one of the top priorities of his administration.
In December 2001, the House approved its version of the
TPA bill by a narrow margin of 215–214.36
Although TPA
itself does not directly reduce barriers to trade, it facilitates
the negotiation of trade agreements and each of the
FTAs considered in this study was implemented under
TPA. The 108th Congress voted on trade agreements
with Australia, Chile, Morocco, and Singapore, while the
109th Congress voted on the Dominican Republic–
Central America Free Trade Agreement (DR-CAFTA),
and agreements with Bahrain and Oman. Table 1 presents
the vote tallies for each bill included in the main
analysis. All the agreements passed; however, there is a
large amount of heterogeneity of support across the bills.
In particular, the vote margins for TPA and the FTAs
with Chile, Singapore, DR-CAFTA, and Oman passed by
much smaller margins than the other FTAs.
During debates on implementing legislation, representatives
who opposed trade liberalization frequently
emphasized the cost of these agreements to American
workers generally and with respect to offshoring in particular.
Not only did representatives consistently discuss
lost manufacturing jobs, they also raised concerns about
the new threat to white-collar, high-wage service jobs of
being moved overseas. As one example, Representative
Stark (D-California) suggests that in addition to 500,000
manufacturing jobs lost due to NAFTA,
3.5 million white-collar jobs and $136 billion in
wages will shift from the United States to low-cost
countries in the next 10 years. So all of those, in addition
to the 100,000 high-tech jobs we have already
lost in California, Silicon Valley, those jobs will become
obsolete under the Bush administration’s
course for free trade. It will not just be IT jobs. We
will see a shift in financial service jobs, research and
development jobs, service call center jobs and insurance
jobs. (United States–Chile Free Trade
Agreement Implementation Act 2003, H7471)
Table 1. Votes on FTAs in the House of Representatives
Bill Vote Date
TPA, House version 215–214 12/6/2001
Chile 271–156 7/24/2003
Singapore 271–155 7/24/2003
Australia 314–109 7/14/2004
Morocco 323–99 7/22/2004
DR-CAFTA 271–215 7/28/2005
Bahrain 327–95 12/7/2005
Oman 221–205 7/20/2006
33
http://www.cwa-union.org/national-issues/secure-sustainable-jobs, accessed
February 1, 2017.
34
http://www.ifpte.org/issues/, accessed February 1, 2017. See also the
Washington Alliance of Technology Workers, formed in 1998 by contract workers
at Microsoft (techsunite.org, accessed February 1, 2017), and the Alliance@IBM
(https://web.archive.org/web/20160504150653/http://www.endicottalliance.org/,
accessed February 1, 2017), which both focused on offshoring as a key policy area.
35
The preferences of those in non-offshorable jobs are less straightforward.
Their preferences may be shaped by skill, education (Hainmueller and Hiscox
2006), consumer interests (Baker 2005), or sociotropic, media, and elite effects
(Guisinger 2017; Mansfield and Mutz 2009; Naoi and Kume 2011).
36
Ultimately, conference resolution with the Senate version folded the
House TPA bill into a larger, comprehensive trade bill known as the Trade
Act of 2002 (H.R. 3009). I consider the first vote on TPA because subsequent
votes on the same issue are closely tied to the original vote. The results are robust
to either vote.
302 Exposure to Offshoring and the Politics of Trade Liberalization
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
In addition to offshoring, representatives also raised
concerns during all of the debates about the weakness
and enforceability of the provisions for labor standards.
The debate transcripts provide suggestive evidence that legislators
consider diffuse labor interests—particularly vulnerability
to offshoring—during the formation of their
trade policy positions.
Measuring Labor’s Economic Interests
The main independent variable is share of the labor
force vulnerable to offshoring. I measure Offshorability as
the percent of workers in the congressional district in
offshorable occupations. I use occupation data from a 5
percent weighted sample from the 2000 Census.
Individuals’ occupations are coded according to
Blinder’s (2007, 11) ranking of offshorability, which defines
occupations that cannot be offshored as those that
(1) must be performed at a specific domestic location
(like a farm) or (2) require face-to-face contact with endusers
(like a taxi driver). District offshorability is measured
as the sum of workers in offshorable and highly
offshorable jobs as a percent of the labor force. I
matched workers to congressional districts based on a
concordance with public-use microdata areas. Further
details on the construction of this measure and other
variables are available in the supplementary information,
along with maps of the distribution of offshorability
across districts. Offshorability ranges from 11.3 percent
to 27.8 percent, and has a mean of 18.5 and a standard
deviation of 2.9 percent. Figure 1 shows that district offshorability
is normally distributed.
As discussed, the factor endowments and sectoral models
offer possible alternative/additional measures of diffuse
labor interests. First, to account for comparative
advantage, as suggested by the factor endowments model,
I control for the percent of skilled workers. This is measured
as the percent of the population with a college degree.
Educational attainment is a common proxy for skill
in analyses of voting (see Bailey 2001; Milner and Tingley
2011). I expect the greater share of constituents who are
college educated, the more likely legislators will be to
support FTAs.37
Percent college has a mean of 21.1 percent
and a standard deviation of 11.6 percent. Data on educational
attainment were collected from the National
Historic Geographical Information System (NHGIS).38
One common alternative measure of skill in the survey literature—based
on average occupation wage (Mansfield,
Mutz, and Silver 2014)—is not an appropriate measure
for this analysis because I distinguish between skill, an
endowment of an individual (and not specific to an occupation),
and occupation characteristics like offshorability,
which are specific to a particular job.
Second, as suggested by sectoral models, I control for industry
cleavages. The primary measures of industry interests
are the shares of employment in export-oriented and
import-competing manufacturing industries.39
As the share
of employment in export-oriented (import-competing)
industries increases, I expect legislators will be more likely
to support (oppose) free trade according to sectoral models.
I define those industries with a positive trade balance
as export-oriented and those with a negative trade balance
as import-competing.40
I calculate data on employment
across industries at the district level using the 2000, 2002,
and 2004 County Business Patterns. The shares of employment
in import-competing and export-oriented industries
are logged to account for skewness. I consider additional
measures of industry interests in robustness
checks presented in the supplementary information.
In addition to being conceptually distinct from skill
and industry, offshorability differs from these two concepts
in the data. In particular, note that offshorability is
only moderately correlated with skill and industry cleavages:
offshorability is positively correlated with district skill
level (q ¼ 0.35), log of employment shares in importcompeting
(q ¼ 0.34), and export-oriented manufacturing
industries (q ¼ 0.20). A correlation scatter plot for the
0
.05
.1
.15
Density
10 15 20 25 30
Percent offshorable
Figure 1. Distribution of the Percent Offshorable in Districts
37
A college education may also shape attitudes to trade for non-material
reasons, as suggested by Hainmueller and Hiscox (2006) and Mansfield and
Mutz (2009). A positive coefficient would be consistent with both the material
and ideational views of the impact of educational attainment.
38
See Minnesota Population Center (2011).
39
Data on trade in services are extremely limited (Jensen 2011).
40
Data available from Schott (2008).
ERICA OWEN 303
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
measures of labor interests and the full correlation matrix,
as well as additional summary statistics, are provided in
the supplementary files.
Analysis of Roll Call Votes
To test Hypothesis 1, I examine roll call votes pertaining to
FTAs in the 107–109th House of Representatives. The dependent
variable is coded one for those representatives
who voted in favor of the agreement and zero for those
who voted against.41
I estimate panel logistic regression
models, pooling votes across Congresses (see Broz 2011;
Milner and Tingley 2011; Nollen and Quinn 1994). In addition
to the measures of labor interests, I include a dummy
variable for members of the Republican Party, who I expect
to be more likely to vote in favor of free trade. I include
the level of unemployment and expect that higher levels of
unemployment will reduce the likelihood that a legislator
votes in favor of free trade (Mansfield and Busch 1995). I
also include dummy variables for region (Nollen and
Quinn 1994), as well as fixed effects for each vote. I estimate
logistic regression with robust standard errors to account
for heterogeneity across representatives.
Main Results
The results, presented in Table 2, support the hypothesis
that when a greater share of constituents is vulnerable to
offshoring, legislators are more likely to vote against trade
liberalization. In Model 1, the coefficient on offshorability
is negative and statistically significant, as hypothesized. A
1 percent increase in vulnerability to offshoring reduces
the odds of voting in favor of free trade by 6.7 percent.42
The results also support the predictions of factor endowments
theory: the coefficient on percent college is positive
and statistically significant. A 1 percent increase in the
share of constituents with a college education increases
the odds of voting in favor by approximately 3.9 percent.43
The results also suggest that district industry exposure to
import competition shapes legislators’ votes on FTAs. A 1
percent increase in the share of workers employed in
import-competing manufacturing industries reduces the
odds of voting in favor of free trade by approximately 2.0
percent.44
These results suggest that the effect of offshorability
is both statistically and substantively significant.
Moreover, the inclusion of district offshorability improves
the explanatory power of the model over a model that excludes
offshorability. To demonstrate, I present an analysis
of nested and non-nested models in Figure 2 in the appendix.
The findings suggest that offshorability contributes to
our ability to understand the political economy of free trade.
In terms of control variables, the coefficient on
Republican is positive and statistically significant, suggesting
Table 2. Analysis of votes on free trade
1 2 3 4 5
FTAsþTPA No TPA All bills % Free trade
% Offshorable À0.069*** À0.083*** À0.077*** À0.045*** À0.795***
(0.021) (0.023) (0.024) (0.010) (0.249)
% College 0.038*** 0.034*** 0.037*** 0.034*** 0.565***
(0.007) (0.008) (0.008) (0.005) (0.101)
Log share employment in exports À0.091 0.007 À0.110 0.037 1.066
(0.103) (0.114) (0.106) (0.052) (1.176)
Log share employment in imports À0.220** À0.321*** À0.181 À0.204*** À3.047**
(0.110) (0.122) (0.118) (0.059) (1.390)
Republican 3.113*** 2.477*** 2.895*** 1.328*** 23.248***
(0.120) (0.188) (0.126) (0.051) (1.222)
Unemployment À0.089** À0.144*** À0.108*** 0.017 0.122
(0.037) (0.041) (0.037) (0.021) (0.423)
West 0.815*** 0.831*** 0.763*** 0.432*** 7.109***
(0.139) (0.156) (0.148) (0.077) (1.794)
Midwest 0.454*** 0.254* 0.365** 0.413*** 6.559***
(0.134) (0.149) (0.142) (0.073) (1.730)
South 0.813*** 0.347** 0.773*** 0.252*** 4.422***
(0.133) (0.156) (0.139) (0.068) (1.625)
% Past corporate contributions 3.974***
(0.441)
Log % past labor contributions À0.018
(0.037)
108th
4.355
(2.909)
109th
À8.016*
(4.585)
Observations 3407 2985 2978 10536 1317
Log likelihood À1444.0 À1219.8 À1294.5 À5207.3 NA
v2
783.1 692.0 650.2 1822.4 NA
BIC 3026.2 2591.7 2716.9 10729.5 11750.5
% Correctly predicted 82.4 83.3 81.7 74.3 NA
Models 1–4 are logit regressions with vote fixed effects suppressed. Model 5 is an OLS regression. Robust standard errors in parentheses. *p < 0.1,
**p < 0.05, ***p < 0.01.
41
Data on votes were collected from Poole and Rosenthal (1997).
42
95% confidence interval: [–0.105, –0.028].
43
95% confidence interval: [0.024, 0.054].
44
95% confidence interval: [–0.035, –0.0003].
304 Exposure to Offshoring and the Politics of Trade Liberalization
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
that Republicans are more likely to vote in favor of free trade
than Democrats (and Independents). The coefficient on district
unemployment is negative and statistically significant;
when there is a higher level of unemployment, legislators are
less likely to vote in favor of free trade. Representatives from
the Midwest, South, and West are more likely to vote in favor
of free trade than are those representatives from the
Northeast. Overall, the model correctly predicts 82.4 percent
of votes by legislators, and offers a proportional reduction of
error of 30.0 percent over the naive model.
Campaign contributions are also an important determinant
of legislators’ votes on trade (see Magee 2010;
Milner and Tingley 2011; Nollen and Quinn 1994).
Because campaign contributions are endogenous to votes
(see Feinberg, Husted, and Reynolds 2011; Fordham and
McKeown 2003; Magee 2010), I use the percent of legislators’
campaign contributions received from corporate and
labor PACs during the previous session of Congress.45
Labor contributions are logged to account for skewness,
while corporate contributions are normally distributed.
Lagging contributions one period eliminates concerns of
endogeneity between votes and contributions. However,
this reduces the number of observations from 3,407 to
2,985. The results are presented in Model 2 of Table 2.
The coefficient on logged labor contributions is not statistically
different from zero,46
while the coefficient on corporate
campaign contributions is negative and statistically
significant. The results for offshorability remain robust:
the coefficient is positive and statistically significant from
zero. A 1 percent increase in offshorability reduces the
odds of voting in favor of free trade by 7.9 percent.47
Robustness
I perform a number of checks to demonstrate that the results
are robust. First, I examine whether the results are
robust to the exclusion of the vote on TPA, in case decisions
to vote on TPA are somehow unique from votes on
FTA implementation bills. These results are presented in
Model 3 of Table 2. The coefficient on offshorability remains
negative and statistically significant.48
Because the
FTAs under consideration in this paper also include investment
and other provisions, I examine whether offshorability
reduces support for free trade more generally.
Thus, in Model 4 of Table 2, I estimate a panel logistic regression
for votes on all 24 trade bills considered by the
107th–109th Congresses.49
The coefficient on offshorability
is again negative and statistically significant, though
smaller in magnitude. Finally, in Model 5, I examine the
impact of offshorability on representatives’ support for
free trade, measured as the percent of votes in favor of
free trade in a given Congress (based on the trade bills
identified in the previous model). Thus, higher values on
the dependent variable indicate greater support for free
trade. As in previous models, the coefficient on offshorability
is negative and statistically significant. A 1 percent
increase in offshorability leads to a 0.80 percent reduction
in the votes in favor of free trade. Additional robustness
checks are presented in the supplementary files, including
model specifications that examine offshorability and task
routineness, and whether the effect of offshorability is
conditional upon partisanship and unemployment. I also
consider alternative measures of industry trade interests,
as well as employment in manufacturing and agriculture.
The effect of offshorability remains negative and statistically
significant in all cases. The findings support those reported
here, namely that in roll call voting, a higher level
of district offshorability leads legislators to take a more
protectionist position on trade liberalization.
Additional Empirical Implications
In this section, I describe two additional steps taken to
examine the validity of the results. First, one implication
of this argument is that offshorability should not determine
legislators’ votes in earlier periods when information
and communication technology were less advanced and
less readily available. Thus, in the supplementary files, I
examine the impact of offshorability on votes on NAFTA,
the Uruguay Round, and GATT during the 103rd
Congress (1993–1994). As expected, I find that offshorability
does not have an effect on legislators’ support for
free trade during this period.
Second, one of the underlying premises of the paper is
that losers from offshoring will be more politically salient
than winners. As a means of testing this claim, I seek to
measure the expected winners and losers from offshoring in
the district separately.50
I draw on the argument of Walter
(2017), who suggests that high-skill workers in offshorable
occupations benefit from trade, while low-skill workers in offshorable
occupations are harmed by trade (see also Rommel
and Walter 2016). If losers are more salient than winners, an
increase in the share of the constituency who are both low
skill (less than a college degree) and in offshorable occupations
will reduce legislators’ support for free trade.
To test this, I include the percent of workers in offshorable
occupations with a college degree or higher and the
percent in offshorable jobs with less than a college degree.
I continue to control for the overall district factor endowments
by including the percent of the district with a degree.
Thus, these results examine whether there are differences
in the impact of offshorability on legislators’ votes
by skill level.51
The results are presented in Model 1 of
Table 3. I find that the coefficient on low-skill offshorability
is negative and statistically significant in Model 1. This
indicates that higher levels of vulnerability to offshoring
among low-skill constituents reduce the likelihood that a
legislator votes in favor of free trade. A 1 percent increase
in vulnerability to offshoring in low-skill workers reduces
the odds of voting in favor of free trade by 12.3 percent.
The coefficient on high-skill offshorability, however, is not
statistically different from zero. However, the coefficient
on overall skill remains positive and statistically significant,
as expected by the factor endowments model. These
findings further support the underlying argument, which
is that losers from trade are politically more important
than winners from trade. In the supplementary files, I
45
Data on PAC contributions were made available by Milner and Tingley
(2011).
46
The coefficient is negative and statistically significant when the measure
is not logged.
47
95% confidence interval: [–0.119, –0.037].
48
In the supplementary files, I present results for the analysis of individual
votes. The coefficient on offshorability is negative and statistically significant
in six of eight votes.
49
I thank an anonymous reviewer for this suggestion. A full list of roll call
votes is provided in the supplementary files.
50
Because this paper focuses on offshorability as a source of protectionist
sentiment in the aggregate, I do not take a position on the determinants of
the winners and losers from offshoring at the individual level. Instead, I examine
one proposed division here and one in the supplementary files.
51
In the supplementary files, I examine whether there are differences
within skill groups by offshorability and find that there are, suggesting that
the offshoring model produces new insights over Heckscher-Ohlin.
ERICA OWEN 305
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
examine a different measure of winners and losers from offshoring
based on task routineness, as suggested by Owen
and Johnston (Forthcoming), which also suggests that losses
from offshoring are more politically salient to legislators.
Position-Taking During Floor Debates
In this section, I examine the amount of attention
devoted to labor during floor debates on FTAs in the
House based on the content of speeches during debates.52
I use dictionary-based content analysis to analyze whether
attention to the costs of trade for workers in representatives’
speeches during floor debates on FTAs varies according
to the level of offshorability among constituents,
as suggested by Hypothesis 2. Content analysis is based on
the idea that certain topics are discussed using a certain
set of words (Quinn et al. 2010). Given limited time and
opportunity to speak on the floor, time allocated to discussing
the costs of trade for workers is a good measure of
the priority placed on labor interests.53
I measure the relative frequency of labor keywords as an
indicator of the importance placed on labor interests by
each speaker. I create the dictionary category “labor,”
which includes the following terms: workers, jobs, offshor*,
outsourc*, white-collar, “exporting jobs,” and “shipping
jobs.” The asterisk captures variations of the stem words. I
perform pre-processing of the text in R,54
and content analysis
is performed using Yoshikoder (Lowe 2011). I selected
these keywords to capture attention to the link between
trade liberalization, offshoring, and the costs for workers.
Although dictionary-based automated content analysis
is difficult to validate (Grimmer and Stewart 2013), I created
the above dictionary specifically for the topic of interest,
based on a close reading of the full corpus of debate
transcripts. Furthermore, I examined the keyword in context
for each of the dictionary words to ensure that these
words are used to discuss the impact of trade on workers
as expected.55
This alleviates concerns that the dictionary
may not be appropriate for the question of interest.
The dependent variable, Rate of labor speech, is equal to
the number of labor keywords used as a percent of the total
number of words spoken during the debate. As in the
above analysis, I pool debates and include fixed effects for
each vote. The total number of speeches across the eight
votes is 401. The rate of labor speech ranges from zero to
8.80 percent (with a mean of 1.50 percent and standard deviation
of 1.65). I log the rate of labor speech to account
for skewness. A difference of means test shows that the rate
of labor speech during debate is greater among representatives
who voted against an agreement than among those
voting in favor of the agreement; this suggests that the
measure captures a dimension of anti-trade sentiment in
addition to attention to trade costs for labor.
House floor debates are governed by special rules restricting
the time for debate. Debates were limited to two hours
for debates on FTA implementation, and debate on TPA
was limited to one hour. Each party is allocated a certain
amount of time that is controlled by the parties’ managers
of the bill (typically members of the committees with jurisdiction
over the bill). The rules restrict the number of
speakers, and as a result, the sample of speakers is not random.
Thus, the rate of labor speech is censored. I therefore
estimate a Heckman selection model of labor speech, conditional
on the decision to speak during the debate.
One challenge of selection models is that it is often difficult
to find determinants of selection that do not also
influence the outcome of interest. However, there is a substantial
literature in American politics devoted to the study
of who speaks in Congress. I build on the model of
Pearson and Dancey (2011), who identify a number of institutional
and representative demographic characteristics
that influence who speaks. I include dummy variables for
Republican Party membership, leadership positions on
relevant committees, and party leadership positions. Those
in leadership positions are more likely to speak. To account
for motivation to speak and the interest of bill managers to
represent different views, I include the absolute value of
the deviation of the ADA score from the mean as a measure
of ideological extremity.56
Finally, I include individual
dummy variables for female, black, and Hispanic representatives,
who are less likely to speak during floor debate.
The results for the selection equation are presented in
Table 5 in the appendix. Representatives are coded as one
if they spoke during the debate on the implementing legislation
of a particular bill. In the whole sample, there
are 401 uncensored observations and 2,756 censored ob-
servations.57
Republicans are less likely to speak than are
Democrats. Committee leaders and party leaders are more
Table 3. Logit regression of free trade on offshorability by skill level
1
% Offshorable, low skill À0.112***
(0.033)
% Offshorable, high skill À0.025
(0.029)
% College 0.022**
(0.011)
Log share employment in exports À0.036
(0.106)
Log share employment in imports À0.116
(0.125)
Republican 3.143***
(0.120)
Unemployment À0.097***
(0.037)
West 0.834***
(0.139)
Midwest 0.482***
(0.136)
South 0.844***
(0.133)
Observations 3407
Log likelihood À1442.5
v2
799.9
BIC 3031.3
% Correctly predicted 82.3
Robust standard errors in parentheses. Vote fixed effects suppressed.
*p < 0.1, **p < 0.05, ***p < 0.01.
52
Data were collected from Gentzkow and Shapiro (2015).
53
See Grimmer (2013, 628), who suggests that the proportion of press releases
allocated to different topics indicates legislators’ expressed priorities.
54
This includes the removal of punctuation and stopwords, as well as word
stemming.
55
This dictionary is conservative in terms of precision because it does not
include words like “protections,” “rights,” or “standards” that were also important
components of the debates, but which do not exclusively refer to labor. I
do not include immigration-related words that may tap a different underlying
issue.
56
The Americans for Democratic Action score, based on 20 key roll call
votes, is an alternative measure of ideology from DW-NOMINATE.
57
I code representatives as eligible to speak during debate if the representative
voted on the bill. Thus, observations are censored if the representative
voted on the bill but did not speak.
306 Exposure to Offshoring and the Politics of Trade Liberalization
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
likely to speak than non-committee and non-party leaders.
The selection model fits the data well, correctly predicting
88.7 percent of all observations, which is a small improvement
over the naive prediction.
Table 4 presents the results for the outcome of interest. In
the outcome equation, the key independent variables are
the measures of labor interests described above. I also include
a dummy variable for members of the Republican
Party, which I expect has a negative effect on labor-related
speech. I also include regional controls (for which the
Northeast is the reference category) and vote fixed effects.
As I demonstrate, the results of the analysis of labor speech
are similar regardless of whether selection is modeled or not.
In Model 1 of Table 4, the coefficient on offshorability
is positive and statistically significant, suggesting that as
constituency exposure to offshoring increases, the use of
labor keywords also increases. A 1 percent increase in offshorability
increases the rate of labor speech by 33.6 per-
cent.58
These findings provide support for Hypothesis 2.
The coefficient on the level of education is negative and
statistically significant, suggesting that legislators from districts
with a greater share of skilled labor pay less attention
to the costs of trade for labor in speeches. The coefficient
on Republican partisanship is not statistically significant.
The coefficients on industry interests (the shares employed
in import-competing and export-oriented manufacturing
industries) are not statistically significant.
Finally, I present the results of a simple regression analysis
of the rate of labor words in Model 2 of Table 4 to
demonstrate that the findings are not dependent upon
the specification of the selection model.59
As in the previous
analyses, the coefficient on offshorability is positive
and statistically significant, as suggested by Hypothesis 2.
Conclusion
Offshoring is an important aspect of international trade
and one that entails new distributional consequences for
labor. Vulnerability to offshoring is not simply a proxy for
skill level or industry, and offshoring in services exposes a
new category of workers to international competition. In
the aggregate, offshorability generates a new source of
protectionist labor preferences. As a result, legislators are
less likely to support trade liberalization when their constituents
face greater vulnerability to offshoring.
I tested this argument in the context of debates and roll
call votes on eight major trade bills in the US House of
Representatives between 2001 and 2006. I found that members
of Congress from districts with a larger share of constituents
vulnerable to offshoring are: (1) more likely to
discuss labor concerns during speeches, and (2) less likely
to vote in favor of FTAs. I also find support for alternative
theories of labor’s interests with respect to trade. In particular,
a greater share of constituents with a college degree
increases legislator support for free trade, while a greater
share employed in import-competing manufacturing industries
reduces support for free trade. Although I only
studied the United States, I expect that these dynamics operate
in other developed countries, especially other
English-speaking countries that face growing competition
from India (Blinder 2009). Further research should examine
the extent to which occupation characteristics—like
vulnerability to offshoring—shape policy and political
outcomes in other countries (see for instance Owen and
Johnston Forthcoming; Rommel and Walter 2016; Walter
2017).
My argument speaks to broader concerns about the political
economy of globalization. It highlights an overlooked
dimension to the welfare consequences of trade
for workers. My findings suggest that vulnerability to offshoring
may also affect votes designed to help workers facing
adjustment costs from trade and offshoring—such as
in the form of Trade Adjustment Assistance (TAA) and
other compensation policies (Rickard 2015; see also Walter
2010, 2017). In the US context, the results suggest that current
TAA will prove insufficient for compensating the losers
from trade, because eligibility for benefits is limited to those
whose jobs are explicitly offshored. In other words, TAA
does not help those jobs that might have otherwise been created
in the United States if not for fragmented production;
thus, TAA does not help those in occupations vulnerable to
offshoring. Furthermore, policies designed to target sectors
or even firms may miss important segments of the workforce
hurt by globalization (Baldwin 2006).
In an international system that constrains protection
(Oatley 2011), scholars of international political economy
must reconsider how redistributive policies can better compensate
losers from globalization if they want to help secure
continued support for openness. Without additional
and reformed redistribution, as suggested by Scheve and
Slaughter (2007), we are beginning to see increasingly protectionist
backlash against globalization in a number of
advanced economies.60
The election of Donald Trump
constitutes one of the more prominent examples of this.
Outspoken on the issues of trade agreements and
Table 4. Regression analysis of labor speech
1 2
% Offshorable 0.237*** 0.238***
(0.073) (0.075)
% College À0.077*** À0.077***
(0.029) (0.030)
Republican À0.738 À0.775*
(0.473) (0.454)
Log share employment in exports À0.485 À0.480
(0.385) (0.391)
Log share employment in imports À0.314 À0.333
(0.435) (0.432)
Unemployment À0.011 À0.012
(0.108) (0.111)
Midwest 1.755** 1.771**
(0.686) (0.696)
West 0.572 0.575
(0.672) (0.687)
South 0.911 0.902
(0.684) (0.697)
Constant À5.369*** À5.528***
(1.743) (1.693)
Observations 3156 403
Log likelihood À2239.8 À1129.6
BIC 4713.3 2361.3
Model 1 estimated with selection model. Robust standard errors in
parentheses. *p < 0.1, **p < 0.05, ***p < 0.01.
58
Calculated as e0.290.
59
Additionally, a Wald test is unable to reject the null hypothesis of independence
between the outcome and selection equations, perhaps because it is
difficult to improve upon the modal prediction of the decision to speak.
60
Preferences toward trade also depend on the level of welfare spending
in the domestic country (Schaffer and Spilker 2016). Moreover, in order to
successfully attenuate protectionist sentiment, compensation policy must consider
domestic institutions and the geographic distribution of interests
(Menendez 2016; Rickard 2012).
ERICA OWEN 307
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
companies that offshore, President Trump moved in the
first days of his administration to withdraw from the TPP
and vowed to renegotiate NAFTA. He appears to have the
support of congressional Republicans, despite their traditionally
pro-free-trade positions.
These arguments have additional implications for the
nature of political and economic organization of labor
interests and the potential influence of these interests.
This paper focuses on the role of diffuse labor interests.
But as the scope of offshoring increases over time, and as
fragmented production increases, we should see increased
lobbying activity by occupation-based special interest
groups. Fragmented production may not only change the
cleavages that form among workers, but may also increase
the costs of organizing for labor. The spread of workers
with shared interests across firms and industries may make
coordination more difficult (Busch and Reinhardt 2000).
In combination with mobile capital, this may reduce the
political and economic bargaining power of labor. Along
these lines, some scholars examine both the ways in which
globalization hurts the bargaining power of unions over
wages (Dumont, Rayp, and Willeme 2006) and the political
effectiveness of organized labor interests. In particular,
heterogeneity of interests among union members will
reduce the ability of labor unions to advocate a single policy
position (Owen 2015b). Governments must pay attention
to this downward pressure on the influence of labor
relative to capital; representation of labor interests is another
key factor in avoiding backlash in the form of economic
populism or isolationism, as in the case of Brexit or
the election of Donald Trump.
Supplementary Information
Supplementary Information is available at www.ericaowen.
com and the International Studies Quarterly data archive.
References
ACEMOGLU, DARON, AND DAVID AUTOR. 2011. “Skills, Tasks and
Technologies: Implications for Employment and Earnings.” In
Handbook of Labor Economics, vol. 4b, edited by Orley Ashenfelter
and David Card, 1043–171. Amsterdam: Elsevier.
ALT, JAMES, AND MICHAEL GILLIGAN. 1994. “The Political Economy of Trading
States: Factor Specificity, Collective Action Problems and Domestic
Political Institutions.” The Journal of Political Philosophy 1(2): 165–92.
APPLEMAN, ERIC. 2015. Democracy in Action. Accessed February 1, 2017.
www.gwu.edu/$action.
AUTOR, DAVID. 2013. “The ‘Task Approach’ to Labor Markets: An
Overview.” Journal for Labour Market Research 46(3): 185–99.
BACCINI, LEONARDO, PABLO M. PINTO, AND WEYMOUTH. STEPHEN 2017. “The
Distributional Consequences of Preferential Trade Liberalization:
Firm-level Evidence” International Organization 71(2): 373–395.
BAILEY, MICHAEL. 2001. “Quiet Influence: The Representation of Diffuse
Interests on Trade Policy, 1983–94.” Legislative Studies Quarterly
26(1): 45–80.
BAILEY, MICHAEL A., JUDITH GOLDSTEIN, AND BARRY R. WEINGAST. 1997. “The
Institutional Roots of American Trade Policy: Politics, Coalitions,
and International Trade.” World Politics 49(3): 309–38.
BAKER, ANDREW. 2005. “Who Wants to Globalize? Consumer Tastes and
Labor Markets in a Theory of Trade Policy Beliefs.” American
Journal of Political Science 49(4): 924–38.
BALDWIN, ROBERT. 2006. “Globalisation: The Great Unbundling(s).”
Economic Council of Finland. 20(3): 5–47.
BALDWIN, ROBERT, AND CHRISTOPHER MAGEE. 2000. “Is Trade Policy for Sale?
Congressional Voting on Recent Trade Bills.” Public Choice 105(1–
2): 79–101.
BAUMGARTEN, DANIEL, INGO GEISHECKER, AND HOLGER GO¨ RG. 2013.
“Offshoring, Tasks, and the Skill-Wage Pattern.” European Economic
Review 61: 132–52.
BLINDER, ALAN. 2006. “Offshoring: The Next Industrial Revolution?”
Foreign Affairs 85(2): 113–38.
———. 2007. “How Many U.S. Jobs Might Be Offshorable?” CEPS
Working Paper No. 142, March 2007, 1–44. Accessed July 2013.
http://www.princeton.edu/$ceps/workingpapers/142blinder.pdf.
———. 2009. “Offshoring: Big Deal, or Business as Usual?” In Offshoring of
American Jobs: What Response from U.S. Economic Policy?, edited by
Benjamin Friedman, 19–59. Cambridge, MA: MIT Press.
BOX-STEFFENSMEIER, JANET, LAURA ARNOLD, AND CHRISTOPHER ZORN. 1997.
“The Strategic Timing of Position Taking in Congress: A Study of
the North American Free Trade Agreement.” American Political
Science Review 91(2): 324–38.
BROZ, J. LAWRENCE. 2008. “Congressional Voting on Funding the
International Financial Institutions.” Review of International
Organizations 3(4): 351–74.
———. 2011. “The United States Congress and IMF Financing, 1944–
2009.” Review of International Organizations 6(3–4): 341–68.
BURGOON, BRIAN. 2012. “Partisan Embedding: How Trade, Investment,
and Immigration Affect Party Support for the Welfare State.”
Comparative Political Studies 45(5): 606–35.
BUSCH, MARC, AND ERIC REINHARDT. 2000. “Geography, International
Trade, and Political Mobilization in U.S. Industries.” American
Journal of Political Science 44(4): 703–19.
CENTER FOR THE STUDY OF RESPONSIVE POLITICS. 2010. Accessed July 15, 2016.
http://www.opensecrets.org/lobby/clientbills.php?id¼D000000095&
year¼2016.
CHASE, KERRY. 2008. “Moving Hollywood Abroad: Divided Labor Markets
and the New Politics of Trade in Services.” International
Organization 62(4): 653–87.
CHOI, YOUNGMI. 2015. “Constituency, Ideology, and Economic Interests in
U.S. Congressional Voting: The Case of the U.S.–Korea Free Trade
Agreement.” Political Research Quarterly 68(2): 266–79.
CONCONI, PAOLA, GIOVANNI FACCHINI, AND MAURIZIO ZANARDI. 2012. “FastTrack
Authority and International Trade Negotiations.” American
Economic Journal: Economic Policy 4(3): 146–89.
———. 2014. “Policymakers’ Horizon and Trade Reforms: The
Protectionist Effect of Elections.” Journal of International Economics
94(1): 102–18.
CONYBEARE, JOHN, AND MARK ZINKULA. 1996. “Who Voted Against the
NAFTA? Trade Unions Versus Free Trade.” World Economy 19(1):
1–12.
CRINO, ROSARIO. 2010. “Service Offshoring and White-Collar
Employment.” Review of Economic Studies 77(2): 595–632.
DAVIS, DONALD, AND JAMES HARRIGAN. 2011. “Good Jobs, Bad Jobs, and
Trade Liberalization.” Journal of International Economics 84(1):
26–36.
DEAN, ADAM. 2015. “The Gilded Wage: Profit-Sharing Institutions and the
Political Economy of Trade.” International Studies Quarterly 59(2):
316–29.
DREZNER, DANIEL. 2004. “The Outsourcing Bogeyman.” Foreign Affairs
83(3): 22–34.
DUMONT, MICHEL, GLENN RAYP, AND PETER WILLEME. 2006. “Does
Internationalization Affect Union Bargaining Power? An Empirical
Study for Five EU Countries.” Oxford Economic Papers 58(1): 77–102.
EBENSTEIN, AVRAHAM, ANN HARRISON, MARGARET MCMILLAN, AND SHANNON
PHILLIPS. 2014. “Estimating the Impact of Trade and Offshoring on
American Workers Using the Current Population Surveys.” Review
of Economics and Statistics 96(3): 581–95.
EGGER, HARTMUT, AND UDO KREICKEMEIER. 2009. “Firm Heterogeneity and
the Labor Market Effects of Trade Liberalization.” International
Economic Review 50(1): 187–216.
FEENSTRA, ROBERT. 2010. Offshoring in the Global Economy: Microeconomic
Structures and Macroeconomic Implications. Cambridge, MA: MIT
Press.
FEENSTRA, ROBERT, AND GORDON HANSON. 1999. “Globalization, Outsourcing
and Wage Inequality.” American Economic Review 86(2): 240–45.
FEIGENBAUM, JAMES, AND ANDREW HALL. 2015. “How Legislators Respond to
Localized Economic Shocks: Evidence from Chinese Import
Competition.” Journal of Politics 77(4): 1012–30.
FEINBERG, ROBERT, THOMAS HUSTED, AND KARA REYNOLDS. 2011. “Trade Policy
and Antitrust: Do Consumers Matter to Legislators?” Review of
International Economics 19(3): 525–38.
308 Exposure to Offshoring and the Politics of Trade Liberalization
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
FORDHAM, BENJAMIN, AND KATJA KLEINBERG. 2012. “How Can Economic
Interests Influence Support for Free Trade?” International
Organization 66(2): 311–28.
FORDHAM, BENJAMIN, AND TIMOTHY MCKEOWN. 2003. “Selection and
Influence: Interest Groups and Congressional Voting on Trade
Policy.” International Organization 57(3): 519–49.
FREEMAN, RICHARD. 2009. “Offshoring of American Jobs: What Response
from U.S. Economic Policy?” In Offshoring of American Jobs: What
Response from U.S. Economic Policy?, edited by Benjamin Friedman,
61–71. Cambridge, MA: MIT Press.
FREUND, CAROLINE, AND CAGLAR OZDEN. 2008. “Trade Policy and Loss
Aversion.” American Economic Review 98(4): 1675–91.
GALANTUCCI, ROBERT. 2014. “Temporary Payoffs or Trade Policy Priorities?
Congressional Support for Trade Adjustment Assistance.” Foreign
Policy Analysis (first view):1–21. http://galantucci.web.unc.edu/
files/2012/12/galantucci_fpa_2014.pdf.
GENTZKOW, MATTHEW, AND JESSE SHAPIRO. 2015. Congressional Record for
104th–110th Congresses: Text and Phrase Counts. Inter-University
Consortium for Political and Social Research (ICPSR).
GOLDSTEIN, JUDITH. 1986. “The Political Economy of Trade: Institutions of
Protection.” American Political Science Review 80(1): 161–84.
GOOS, MAARTEN, ALAN MANNING, AND ANNA SALOMONS. 2014. “Explaining
Job Polarization: Routine-Biased Technological Change and
Offshoring.” American Economic Review 104(8): 2509–26.
GO¨ RG, HOLGER. 2011. “Globalization, Offshoring and Jobs.” In Making
Globalization Socially Sustainable, edited by Mach Bacchetta and
Marion Jansen, 21–48. Geneva: International Labour Organization
(ILO) and World Trade Organization (WTO).
GRIMMER, JUSTIN. 2013. “Appropriators Not Position Takers: The
Distorting Effects of Electoral Incentives on Congressional
Representation.” American Journal of Political Science 57(3): 624–42.
GRIMMER, JUSTIN, AND BRANDON STEWART. 2013. “Text as Data: The Promise
and Pitfalls of Automatic Content Analysis Methods for Political
Texts.” Political Analysis 21(3): 267–97.
GROSSMAN, GENE, AND ELHANAN HELPMAN. 1994. “Protection for Sale.”
American Economic Review 84(4): 833–50.
———. 2002. Interest Groups and Trade Policy. Princeton, NJ: Princeton
University Press.
GROSSMAN, GENE, AND ESTEBAN ROSSI-HANSBERG. 2008. “Trading Tasks: A
Simple Theory of Offshoring.” American Economic Review 98(5):
1978–97.
GUISINGER, ALEXANDRA. 2017. American Opinion on Trade: Preferences
without Politics. New York, NY: Oxford University Press.
HAINMUELLER, JENS, AND MICHAEL HISCOX. 2006. “Learning to Love
Globalization: Education and Individual Attitudes Toward
International Trade.” International Organization 60(2): 469–98.
HAYS, JUDE. 2009. Globalization and the New Politics of Embedded Liberalism.
New York: Oxford University Press.
HELLWIG, TIMOTHY. 2015. Globalization and Mass Politics Globalization and
Mass Politics: Retaining the Room to Maneuver. New York: Cambridge
University Press.
HELPMAN, ELHANAN, AND OLEG ITSKHOKI. 2010. “Labor Market Rigidities,
Trade and Unemployment.” Review of Economic Studies 77(3):
1100–1137.
HELPMAN, ELHANAN, OLEG ITSKHOKI, AND STEPHEN REDDING. 2010. “Inequality
and Unemployment in a Global Economy.” Econometrica 78(4):
1239–83.
HISCOX, MICHAEL. 2002a. International Trade and Political Conflict:
Commerce, Coalitions and Mobility. Princeton, NJ: Princeton
University Press.
———. 2002b. “Commerce, Coalitions, and Factor Mobility: Evidence
from Congressional Votes on Trade Legislation.” American Political
Science Review 96(3): 593–608.
HUMMELS, DAVID, RASMUS JøRGENSEN, JAKOB R. MUNCH, AND CHONG XIANG.
2014. “The Wage Effects of Offshoring: Evidence from Danish
Matched Worker-Firm Data.” American Economic Review 104(6):
1597–629.
HWANG, WONJAE, AND HOON LEE. 2014. “Globalization, Factor Mobility,
Partisanship and Compensation Policies.” International Studies
Quarterly 58(1): 92–105.
IVERSEN, TORBEN, AND ANNE WREN. 1998. “Equality, Employment and
Budgetary Restraint: The Trilemma of the Service Economy.” World
Politics 50(4): 507–46.
JENSEN, J. BRADFORD. 2011. Global Trade in Services: Fear, Facts, and
Offshoring. Washington, DC: Peterson Institute for International
Economics.
JENSEN, J. BRADFORD, AND LORI KLETZER. 2005. “Tradable Services:
Understanding the Scope and Impact of Services Offshoring.”
Brookings Trade Forum 73–133.
JENSEN, J. BRADFORD, DENNIS QUINN, AND STEPHEN WEYMOUTH. 2015a.
“Winners and Losers in International Trade: The Effects on U.S.
Presidential Voting.” International Organization.
———. 2015b. “The Influence of Firm Global Supply Chains and Foreign
Currency Undervaluations on US Trade Disputes.” International
Organization 69(4): 913–47.
JEONG, GYUNG-HO. 2009. “Constituent Influence on International Trade
Policy in the United States, 1987–2006.” International Studies
Quarterly 53(2): 519–40.
KAHNEMAN, DANIEL, AND AMOS TVERSKY. 1979. “Prospect Theory: An Analysis
of Decision Under Risk.” Econometrica 47(2): 263–92.
KAMBOUROV, GUEORGUI, AND IOURII MANOVSKII. 2008. “Rising Occupational
and Industry Mobility in the United States: 1968–97.” International
Economic Review 49(1): 41–79.
KAROL, DAVID. 2007. “Does Constituency Size Affect Elected Officials’
Trade Policy Preferences?” Journal of Politics 69(2): 470–82.
KITSCHELT, HERBERT, AND PHILIPP REHM. 2014. “Occupations as a Site of
Political Preference Formation.” Comparative Political Studies 47(12):
1670–706.
KLETZER, LORI. 2009. “Comments.” In Offshoring of American Jobs: What
Response from U.S. Economic Policy? edited by Benjamin Friedman,
85–91. Cambridge, MA: MIT Press.
LADEWIG, JEFFREY W. 2006. “Domestic Influences on International Trade
Policy: Factor Mobility in the United States, 1963 to 1992.”
International Organization 60(1): 69–103.
LOWE, WILL. 2011. “Yoshikoder: Cross-Platform Multilingual Content
Analysis.” Java software version 0.6.4. Accessed January 19, 2015.
http://conjugateprior.org/software/yoshikoder/.
MAGEE, CHRISTOPHER SEAN PATRICK. 2010. “Would NAFTA Have Been
Approved by the House of Representatives under President Bush?
Presidents, Parties, and Trade Policy.” Review of International
Economics 18(2): 382–95.
MANKIW, N. GREGORY, AND PHILLIP SWAGEL. 2006. “The Politics and
Economics of Offshore Outsourcing.” Journal of Monetary Economics
53(5): 1027–56.
MANSFIELD, EDWARD, AND MARC BUSCH. 1995. “The Political Economy of
Nontariff Barriers: A Cross-National Analysis.” International
Organization 49(4): 723–49.
MANSFIELD, EDWARD, AND HELEN MILNER. 2012. Votes, Vetoes, and the Political
Economy of International Trade Agreements. Princeton, NJ: Princeton
University Press.
MANSFIELD, EDWARD, HELEN MILNER, AND PETER ROSENDORFF. 2002. “Why
Democracies Cooperate More: Electoral Control and International
Trade Agreements.” International Organization 56(3): 477–513.
MANSFIELD, EDWARD, AND DIANA MUTZ. 2009. “Support for Free Trade: SelfInterest,
Sociotropic Politics and Out-Group Anxiety.” International
Organization 63(3): 425–57.
———. 2013. “Us vs. Them: Mass Attitudes toward Offshore Outsourcing.”
World Politics 65(4): 571–608.
MANSFIELD, EDWARD D., DIANA C. MUTZ, AND LAURA R. SILVER. 2014. “Men,
Women, Trade, and Free Markets.” International Studies Quarterly
59(2): 303–15.
MARGALIT, YOTAM. 2011. “Costly Jobs: Trade-Related Layoffs, Government
Compensation, and Voting in U.S. Elections.” American Political
Science Review 105(1): 166–88.
———. 2012. “Lost in Globalization: International Economic Integration
and the Sources of Popular Discontent.” International Studies
Quarterly 56(3): 484–500.
MCGILLIVRAY, FIONA. 2004. Privileging Industry: The Comparative Politics of
Trade and Industrial Policy. Princeton, NJ: Princeton University
Press.
MENENDEZ, IRENE. 2016. “Globalization and Welfare Spending: How
Geography and Electoral Institutions Condition Compensation.”
International Studies Quarterly 60(4): 665–76.
MIDFORD, PAUL. 1993. “International-Trade and Domestic Politics—
Improving on Rogowski’s Model of Political Alignments.”
International Organization 47(4): 535–64.
ERICA OWEN 309
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
MILNER, HELEN. 1988. Resisting Protectionism: Global Industries and the
Politics of International Trade. Princeton, NJ: Princeton University
Press.
MILNER, HELEN, AND DUSTIN TINGLEY. 2011. “Who Supports Global
Economic Engagement? The Sources of Preferences in American
Foreign Economic Policy.” International Organization 65(1): 37–68.
———. 2015. Sailing the Water’s Edge: The Domestic Politics of American Foreign
Policy. Princeton, NJ: Princeton University Press.
MINNESOTA POPULATION CENTER. 2011. National Historical Geographic
Information System: Version 2.0. Minneapolis: University of
Minnesota.
NAOI, MEGUMI, AND IKUO KUME. 2011. “Explaining Mass Support for
Agricultural Protectionism: Evidence from a Survey Experiment
During the Global Recession.” International Organization 65(4):
771–95.
NOLLEN, STANLEY, AND DENNIS QUINN. 1994. “Free Trade, Fair Trade,
Strategic Trade and Protectionism in the U.S. Congress, 1987–
1988.” International Organization 48(3): 491–525.
OATLEY, THOMAS. 2011. “The Reductionist Gamble: Open Economy
Politics in the Global Economy.” International Organization 65(2):
311–41.
OLDENSKI, LINDSAY. 2014. “Offshoring and the Polarization of the US
Labor Market.” Industrial and Labor Relations Review 67(3): 734–61.
OLSON, MANCUR. 1971. The Logic of Collective Action: Public Goods and the
Theory of Groups. Boston: Harvard University Press.
OWEN, ERICA. 2015a. “Labor and Protectionist Sentiment.” In The Oxford
Handbook of Political Economy of International Trade, edited by Lisa
Martin, 119–37. Oxford: Oxford University Press.
———. 2015b. “The Political Power of Organized Labor and the Politics
of Foreign Direct Investment in Developed Democracies.”
Comparative Political Studies 48(13): 1746–80.
OWEN, ERICA, AND NOEL JOHNSTON. Forthcoming. “Occupation and the
Political Economy of Trade: Job Routineness, Offshorability and
Protectionist Sentiment.” International Organization.
OWEN, ERICA, AND DENNIS QUINN. 2016. “Does Economic Globalization
Influence the U.S. Policy Mood? A Study of U.S. Public Sentiment,
1954–2011.” British Journal of Political Science 46(1): 95–125.
PANDYA, SONAL. 2016. “Political Economy of Foreign Direct Investment.”
Annual Review of Political Science 19: 455–75.
PEARSON, KATHRYN, AND LOGAN DANCEY. 2011. “Speaking for the
Underrepresented in the House of Representatives: Voicing
Women’s Interests in a Partisan Era.” Politics & Gender 7(4):
493–519.
POOLE, KEITH, AND HOWARD ROSENTHAL. 1997. A Political-Economic History of
Roll Call Voting. New York: Oxford University Press.
PRESTON, JULIA. 2015. “Toys ‘R’ US Brings Temporary Foreign Workers to
U.S. To Move Jobs Overseas.” New York Times, September 29.
Accessed February 2, 2017. http://www.nytimes.com/2015/09/30/
us/toys-r-us-brings-temporary-foreign-workers-to-us-to-move-jobs-over
seas.html?_r¼0.
QUINN, KEVIN, BURT MONROE, MICHAEL COLARESI, MICHAEL CRESPIN, AND
DRAGOMIR RADEV. 2010. “How to Analyze Political Attention with
Minimal Assumptions and Costs.” American Journal of Political Science
54(1): 209–28.
REHM, PHILIPP. 2009. “Risks and Redistribution an Individual-Level
Analysis.” Comparative Political Studies 42(7): 855–81.
RICKARD, STEPHANIE. 2012. “Electoral Systems, Voters’ Interests and
Geographic Dispersion.” British Journal of Political Science 42(4): 855–77.
———. 2015. “Compensating the Losers: An Examination of
Congressional Votes on Trade Adjustment Assistance.” International
Interactions 41(1): 46–60.
ROGOWSKI, RONALD. 1989. Commerce and Coalitions: How Trade Affects
Domestic Political Alignments. Princeton, NJ: Princeton University Press.
———. 2006. “Trade, Immigration and Cross-Border Investment.” In The
Oxford Handbook of Political Economy, edited by Donald Wittman and
Barry Weingast, 814–30. New York: Oxford University Press.
ROMMEL, TOBIAS, AND STEFANIE WALTER. 2017. “The Electoral Consequences
of Offshoring: How the Globalization of Production Shapes Party
Preferences in Multi-Party Systems.” Comparative Political Studies.
First view.
SCHAFFER, LENA, AND GABRIELE SPILKER. 2016. “Adding Another Level
Individual Responses to Globalization and Government Welfare
Policies.” Political Science Research and Methods 4(2): 339–426.
SCHEVE, KENNETH, AND MATTHEW SLAUGHTER. 2007. “A New Deal for
Globalization.” Foreign Affairs 86(4): 34–47.
SCHOTT, PETER K. 2008. “US Manufacturing Exports and Imports by SIC
or NAICS Category and Partner Country, 1972 to 2005.” Accessed
November 8, 2015. http://faculty.som.yale.edu/peterschott/sub_
international.htm.
STEAGALL, JEFFREY, AND KEN JENNINGS. 1996. “Unions, PAC Contributions,
and the NAFTA Vote.” Journal of Labor Research 17(3): 515–21.
STOLPER, WOLFGANG, AND PAUL SAMUELSON. 1941. “Protection and Real
Wages.” Review of Economic Studies 9(1): 58–73.
TOVAR, PATRICIA. 2009. “The Effects of Loss Aversion on Trade Policy: Theory
and Evidence.” Journal of International Economics 78(1): 154–67.
UNITED STATES–CHILE FREE TRADE AGREEMENT IMPLEMENTATION ACT. 2003.
Congressional Record. H7459–H7489.
WALTER, STEFANIE. 2010. “Globalization and the Welfare State: Testing the
Microfoundations of the Compensation Hypothesis.” International
Studies Quarterly 54(2): 403–26.
———. 2017. “Globalization and the Demand-Side of Politics. How
Globalization Shapes Labor-Market Risk Perceptions and Policy
Preferences.” Political Science Research Methods 5(1): 55–80.
WREN, ANNE, AND PHILIPP REHM. 2013. “Service Expansion, International
Exposure, and Political Preferences.” In The Political Economy of the
Service Transition, edited by Anne Wren, 248–81. Oxford: Oxford
University Press.
Appendix
Figure 2 presents BIC scores, where smaller values on the
BIC indicate greater explanatory power. In Figure 2,
“factor” refers to percent college, “industry” refers to
employment shares in exporting and import-competing
manufacturing industries, and “offshoring” refers to the
percent of the district in offshorable occupations. The
“full” model includes factor, industry and offshoring variables
(as specified in models above).
Table 5. Selection model of decision to speak
1
Republican À0.405***
(0.064)
Committee leader 1.224***
(0.154)
Party leader 0.402*
(0.206)
ADA ideological extremism 0.003
(0.003)
Female 0.122
(0.084)
Black À0.140
(0.106)
Hispanic À0.167
(0.140)
108th
À0.675***
(0.081)
109th
À0.623***
(0.083)
Constant À0.608***
(0.139)
q À0.038
(0.191)
ln(r) 1.385***
(0.030)
First-stage selection model for Model 1 in Table 4. Robust standard
errors in parentheses. *p < 0.1, **p < 0.05, ***p < 0.01.
310 Exposure to Offshoring and the Politics of Trade Liberalization
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018
Factor
Factor+industry
Factor+offshoring
Full model
3015 3020 3025 3030 3035
BIC
Figure 2. BIC for different model specifications
ERICA OWEN 311
Downloaded from https://academic.oup.com/isq/article-abstract/61/2/297/4064142
by Univesiteitsbibliotheek Utrecht user
on 16 January 2018