The Political Economy of Reshoring: Evidence from the

Semiconductor Industry

Richard Clark, Sarah Kreps, and Adi Rao*

May 1, 2023

Abstract

States have increasingly taken steps to “reshore” manufacturing in industries that re-
located abroad amidst pressures of globalization. Although scholars have carefully
studied individual trade preferences, they have focused primarily on support for free
trade and perceptions of offshoring. Less work has examined attitudes toward poli-
cies that seek to reverse these trends. We investigate public support for reshoring,
advancing hypotheses about the role of national security considerations and economic
self-interest in shaping why and under what conditions the public backs incentives
and investments that bring manufacturing home. We field two survey experiments
to representative samples of the American public, finding that U.S. citizens support
reshoring in most cases, irrespective of political party. Only the prospect of severe
price increases and reshoring from geographically proximate allies moderate these re-
sponses, suggesting publics do not view reshoring as simply the inverse of offshoring.
Our findings extend recent research on the globalization backlash and help to explain
decisive bipartisan action in the United States to reshore semiconductors.

Keywords: globalization; trade, reshoring, semiconductors, China

* RichardClark ([email protected]) is Assistant Professor of Government at Cornell University. Sarah

Kreps is John L. Wetherill Professor of Government at Cornell University. Adi Rao is Ph.D. candidate at Cornell
University. We thank Ryan Brutger, Sarah Baurle Danzman, Aycan Katitas, Abe Newman, Tom Pepinsky, and Chris
Way for helpful comments and suggestions.
 Reshoring, defined as "the decision to bring back to the home country production activities

earlier offshored," has been on the rise in the United States over the last decade (Barbieri, Keshk

and Pollins 2009, 79). The non-profit Reshoring Institute, established to reverse offshoring trends

of the 1990s and early 2000s, touts a total of 1.3 million jobs recovered since 2010.1 Although

globalization had increasingly integrated world economies and raised global economic output, it

also produced distributional consequences in the form of lost manufacturing jobs (Rodrik 2018;

Baccini and Weymouth 2021). Arguments in favor of reshoring gained steam in 2021, as the

pandemic created supply chain shortages in semiconductor production, leading to inconsistent

access to cars and laptops that relied on these chips and helping make the case for reshoring as

a solution to such disruptions. If these manufacturing processes were conducted at home, the

argument went, the United States would be less reliant on other countries’ pandemic policies.

Reshoring, then, was thought to not only redress the distributional consequences that had been

realized by globalization-fueled offshoring but also to provide supply chain resilience in the face

of myriad threats, including the pandemic and mounting geopolitical tensions with China. In

June 2021, the Biden Administration released a supply chain review with a promise to "strengthen

American supply chains to promote economic security, national security, and good-paying, union

jobs here at home."2 The Biden White House has frequently invoked this "Made in America"

policy as a hallmark of the Administration, in stark contrast to the decades of globalization in

which manufacturing jobs fled abroad as industries invested in labor-rich countries to keep costs

low for consumers and pad their bottom lines.

The framing of reshoring as a way to bolster economic security is emblematic of a broader glob-
1
Reshoring Initiative. “2021 Data Report.” https://bit.ly/3jFetEt
2
The White House. 2021. https://bit.ly/3lmNCxs

1
alization backlash, one that has increasingly received scholarly treatment (Colantone and Stanig

2018; Autor et al. 2020; Mansfield, Milner and Rudra 2021; Broz, Frieden and Weymouth 2021;

Ballard-Rosa, Jensen and Scheve 2022). Existing studies, however, offer little evidence for a con-

nection between public opinion and globalization’s retreat. For instance, though anti-globalization

candidates and policies have achieved increased success, reviews of public attitudes toward glob-

alization (Walter 2021) show little change in support for free trade and investment in the United

States.3 While scholars have examined political attitude formation in the context of highly salient

offshoring events (Margalit 2011; Rommel and Walter 2018; Autor et al. 2020), they have been

relatively silent on the topic of reshoring.4

To what extent does the public support or oppose reshoring, which has been referred to as a

form of de-globalization,5 and what explains the basis of those attitudes? Given the inextricable

links between governments’ trade policies, their regulation of foreign direct investment, and firms’

decisions to reshore under supply chain complexity (Kim, Liao and Miyano 2023), we draw on

arguments about the political economy of trade, security externalities, and public attitude formation

to advance two sets of factors that may shape public opinion of reshoring.

First, security considerations may affect how Americans think about reshoring. Offshoring,

which prompts firms to move abroad in search of cheaper production costs, often increases coun-

tries’ reliance on imports from their adversaries. Trade, meanwhile, is associated with economic

gains that can be used to strengthen a country’s military (Gowa and Mansfield 1993), and publics

are attuned to this (Carnegie and Gaikwad 2022). Further, an adversarial country could intention-

ally withhold access to a crucial trade component, a form of weaponized interdependence (Farrell
3
Gallup. 2020. https://bit.ly/3AGbrVs.
4
Though see Vaughn and Weldzius (2021).
5
The Hill. 2023. https://bit.ly/40P4dKr

2
and Newman 2019) that leaves the target country vulnerable economically as well as militarily. It

follows that individuals are more likely to seek reshoring from an adversarial trade partner than a

friend. We therefore expect that Americans should be more favorable to reshoring when the trade

partner is an adversary versus an ally.

Second, we observe that individual attitudes toward trade and globalization may hinge more

on economic self-interest. Although trade policy may be difficult to parse with respect to one’s

own material well-being (Rho and Tomz 2017; Mutz 2021), we expect that sudden price hikes and

shortages of critical imported goods clarify the salience and impact of trade in integrated supply

chains. In these volatile economic and high information settings, respondents should elicit support

for reshoring as a way to reduce consumer costs and ensure the availability of goods.6

We apply these contentions to the reshoring of semiconductor manufacturing — a tightly in-

tegrated, globalized supply chain. We pre-registered hypotheses related to security and economic

self-interest considerations and carried out a two-part survey experiment on diverse samples of

Americans. In the first study, we varied the country from which manufacturing was to be reshored

(China or South Korea) and the product to be reshored (steel or semiconductors). Our theory

anticipates that support for de-globalization policies should be higher for reshoring from China

(an adversary) and for semiconductors (which at the time of our study was a salient product with

direct consumer impact in terms of price and availability). Strikingly, we find that support for

reshoring among Americans is high in absolute terms regardless of the country from which it is

to be reshored, the product to be reshored, and respondents’ political party (Democratic or Re-

publican). This suggests that underlying support for de-globalization policies is quite high despite

significant elite and public polarization in American politics more generally (Green, Palmquist and
6
See Margalit (2011); Casler and Clark (2021).

3
Schickler 2002; Druckman, Peterson and Slothuus 2013).

The results from the first study motivate the second, which seeks to understand the basis of

such high support for reshoring. Here, we probe potential mechanisms of support, drawing on the

literature on attitudes toward trade and globalization to investigate whether respondents are moti-

vated by a desire to bring jobs back to the U.S., consumer price concerns, a feeling of nostalgia for

past American superiority in manufacturing, or xenophobic skepticism toward foreign out-groups.

To probe such mechanisms, we manipulate several additional factors with theoretical linkages to

public perceptions of trade and globalization, namely price increases (low or high), variation in the

types of jobs created by reshoring (blue or white collar), and the trade partner (China or Canada).7

We find evidence that Americans are sensitive to large price increases, and they are less willing

to reshore from nearby allies like Canada, whose geographic proximity and demographic compo-

sition impact respondents’ perceptions of “foreignness” (Mutz 2021).8 This suggests a role for

both consumer cost considerations and xenophobia, though support for reshoring remains high in

absolute terms regardless of the conditions assigned and across partisan divides.

Our research contributes to several important literatures. First, insofar as reshoring drives

protectionism, it speaks to debates about public opinion of trade, security externalities, and the re-

lationship between the two (Gowa and Mansfield 1993; Carnegie and Gaikwad 2022). We find that

security ties may not shape public opinion over reshoring in cases where allies are geographically

(or demographically) distant; respondents only reduce their support for reshoring (and support

trade) when the producing country is a close ally and regional partner. This may be indicative of a

shift in how citizens think about trade with allies and adversaries in the context of complex supply
7
On sensitivity to price, see Scheve and Slaughter (2001); Mayda and Rodrik (2005); on blue collar jobs and
nostalgia in an era of deindustrialization, see Kojola (2019); Goldstein, Ballard-Rosa and Rudra (2021); on support
for trade with allies versus adversaries, see Carnegie and Gaikwad (2022).
8
Also see Hemmer and Katzenstein (2002) on the role of collective identity in designing multilateralism

4
chains and trade in security-relevant goods.9 It also raises the prospect of ethnocentrism in trade

preferences insofar as the public prefers to trade with countries that are geographically proximate

as well as demographically and culturally similar (Bankert, Powers and Sheagley 2022).

Second, we extend research on the globalization backlash (Mansfield, Milner and Rudra 2021),

showing how underlying public opinion of reshoring matches the policy environment in the United

States — both policymakers and the public seem enthusiastic to bring manufacturing back home.10

Finally, also build on a nascent literature on the political economy of supply chains, though we

shift the focus from labor rights (Greenhill, Mosley and Prakash 2009; Mosley 2010; Malesky

and Mosley 2018), expropriation (Johns and Wellhausen 2016), and environmental considerations

(Kolcava, Smith and Bernauer 2022) to public opinion of reshoring under supply chain complexity.

Supply chains have increasingly become central to many policy debates pertaining to globalization,

and yet International Relations scholarship on the topic has not kept pace. Our research aims to

spark interest in how academics, policymakers, the public, and firms interpret and address kinks in

these increasingly complicated chains, suggesting that autarky is a likely response.

The Case of Semiconductors

As our baseline case, we focus on semiconductor reshoring — an ideal typical case of global-

ized supply chains. Toward the end of the 20th century, the digital revolution drove demand for

silicon chips beyond the capacity of American suppliers, while lower trade barriers led to incen-

tives for specialization and offshoring to increase the speed and lower the cost of manufacturing.

The result was a complex and highly specialized supply chain consisting of over 1,000 steps, with
9
Also see Davis and Meunier (2011); Farrell and Newman (2019).
10
On symmetry between elite and public opinion, see Kertzer (2020).

5
chips crossing national borders 70 times before reaching a customer.11 Taking advantage of the

technical expertise of countries such as Taiwan and South Korea, several American semiconductor

firms went “fabless,” meaning that they design and sell the hardware but contract out the produc-

tion of the semiconductor chips to a specialized manufacturer, or foundry. Such chips are essential

in the making of everything from core national security hardware (e.g., fighter planes) to everyday

devices (e.g., smartphones, cars, and washing machines) that are crucial for a thriving economy.

As a result of these trends, over the last several decades, the United States has steadily lost

market share in semiconductor manufacturing. In 1990, the U.S. controlled 37% of semiconductor

fabrication. By 2021, that share was 12%, as it shed market share to countries such as Taiwan,

Korea, and China amidst a wave of offshoring.12 The United States had done little to stem the

tide until 2021, when a confluence of events led to a policy shift. COVID-19 lockdowns disrupted

supply chains, causing acute semiconductor chip shortages. The Biden Administration estimated

that the pandemic chip shortage cost the United States one percent of its GDP — attesting to the

integral relationship between chips and economic growth. Further, escalating geopolitical tensions

with China raised concerns about the sustainability of integrated supply chains over time.

Questions of existing and future availability galvanized support for U.S. legislation that in-

cluded tens of billions of dollars for firms to bolster semiconductor manufacturing capacity domes-

tically as well as hundreds of billions for scientific research in this area. The Chips and Science

Act passed the Senate in July 2022 with a lopsided bipartisan vote of 64-33. Legislative efforts to

reshore this industry in the name of resilience led some observers to conclude that “semiconduc-

tor shortages end an era of globalization” (Schewe 2021). Reshoring would inevitably reduce the
11
Center for Security and Emerging Technology. “The Semiconductor Supply Chain.” 2021. https://bit.ly/
3RLy1Up
12
Consultancy.Asia. 2021. https://bit.ly/3cxkSKS.

6
volume of trade between countries like the U.S. and China as well.

The issue of semiconductor manufacturing reshoring then stands to provide valuable insights

into public attitudes toward policies intended to unwind globalization, which have gone largely un-

studied despite the pronounced steps taken toward reshoring in recent years. We seek to understand

not just aggregate levels of support but also the basis and elasticity of that support.

Hypotheses

We pre-registered the hypotheses derived below with AsPredicted before deploying our first

survey; additional information about pre-registration can be found in Appendix §1. Building on

the existing literature about public attitudes in international political economy, we advance two

main sets of hypotheses, the first having to do with security and the second with economic self-

interest.

We especially draw from scholarly research on public opinion of trade. While reshoring rests

on firms’ foreign direct investment strategies, its most observable implications for the public are in

the trade domain. Indeed, reshoring reduces a country’s reliance on foreign imports. In doing so,

it shifts job creation from abroad to the home country. Consumers may hear about firms returning

home to manufacture goods within their country’s borders; existing work shows that offshoring

events are often highly salient and reverberate within affected communities (Margalit 2011; Ace-

moglu et al. 2016), and so reshoring events that seek to reverse the effects of offshoring should be

similarly salient. Moreover, political rhetoric emphasizing the importance of manufacturing goods

at home is common in the U.S. context, as highlighted above. For these reasons, we believe trade

to be the appropriate lens through which to study attitudes towards reshoring.

7
 First, we theorize that security concerns weigh on public attitudes and that the public will be

more receptive to reshoring when the trade partner is an adversary than when the partner is an

ally. Research recognizes that trade and other economic relationships, from aid flows to member-

ship in international organizations, often mirror security relationships (Gowa and Mansfield 1993;

Mansfield and Bronson 1997; Davis and Pratt 2020). Such patterns are attributed to at least two

considerations. First, countries might utilize the economic gains from trade to bolster their mil-

itaries (Gowa and Mansfield 1993; Mansfield and Bronson 1997). These concerns resonate with

the public, which has expressed preferences for trading with allies over adversaries in part due to

the association between trade gains and military advantages (Carnegie and Gaikwad 2022).

Second, citizens may also believe that trade flows are less likely to be weaponized by an ally

than an adversary. Globalization has forged tight networks of economic interdependence between

countries, which scholars argue begets coercive potential and security vulnerabilities (Krasner

1976; Farrell and Newman 2019; Bauerle Danzman 2021). In the semiconductor context, for ex-

ample, the United States is increasingly reliant on China, whose manufacturing share is more than

double that of the U.S.13 Given the importance of chips for national security hardware and an array

of consumer goods, citizens may perceive an asymmetric reliance on China to be dangerous since

China could restrict U.S. access to them. South Korea, Taiwan, and Japan have also grown their

share of chip manufacturing in recent decades but are allies, making the threat of weaponization

less plausible. We hypothesize that citizens likely favor reshoring from a foe that they fear could

weaponize American reliance than from a longstanding ally such as South Korea.14 In short, given
13
Boston Consulting Group. 2020. “Government Incentives and U.S. Competitiveness in Semiconductor Manufac-
turing.” https://bit.ly/3YnozJ8
14
Recent polling from the Chicago Council shows that a plurality of Americans are aware that South Korea is an
ally or necessary partner of the United States — awareness of alliance ties is on par with that of Germany, France, and
Japan. The same polling shows that similar percentages of Americans identify China as a rival or adversary, making
both countries good cases for our theory. See “Americans on Their Allies, Partners, and Their Rivals.” Chicago

8
the growing skepticism of and competition with China in the U.S. and the multi-decade economic

and security partnership with South Korea, we expect the following:

Hypothesis 1. Respondents will be more likely to support reshoring when the trade partner is

China than when the trade partner is South Korea.

We also posit an important role for economic self-interest concerns; literature in this vein

broadly suggests that individuals will support an economic policy when they see economic ben-

efit to doing so. Though scholars have long shown that individuals have difficulties pinpointing

the downstream economic consequences of trade policies (Guisinger 2009; Rho and Tomz 2017),

recent work suggests that individuals may do so when the information environment, inclusive of

media and elite cues, provides them the requisite information (Casler and Clark 2021). We build

on this literature, highlighting the importance of information for individuals’ assessments of trade

policy and reshoring in particular. Indeed, Gallup found that a majority of Americans experienced

supply chain problems in 2021, including shipping delays and price increases.15 The industries for

which such supply chain kinks were particularly acute should therefore be top-of-mind for citizens.

The media, meanwhile, covered the semiconductor supply problem and connected it to consumer

price inflation and low product availability; Figure 1 shows that semiconductors rose from relative

obscurity to a more mainstream issue around the time of our studies, nearly converging with cover-

age on steel. Steel coverage, conversely, decreased from its historic high during the Trump Era.16

The result, then, was a high information environment in which consumers could ably connect trade

policy over semiconductors to their own economic self-interest.

We compare support for reshoring semiconductor chips with steel production to help isolate the
Council. 2021. https://bit.ly/3lWjNEs.
15
Gallup. 2021. https://bit.ly/3OeGXzF
16
A similar plot with Google Trends data appears in the Appendix, Figure A1, with accompanying discussion.

9
 News Articles Over Time Involving Term 'Semiconductor' Versus 'Steel'
5000

Study 1

Study 2
4000
Number of Articles

3000

2000

Steel

1000

Semiconductors

2010 2015 2020

Year

Figure 1: Salience of Semiconductors versus Steel. This figure shows public interest from
AB/INFORM searches for news sources with the term "semiconductor" versus "steel" from the
years 2010 to 2022. ABI/INFORM maintains a database of millions of articles across nearly 10,000
news outlets (especially business-oriented outlets, e.g., Wall Street Journal, The Economist). Ver-
tical lines demarcate the timing of our two studies. Source: ABI/INFORM, https://about.
proquest.com/en/products-services/abi_inform_complete/. Accessed 23
Apr. 2023.

10
effect of short-term economic self-interest considerations. Steel and semiconductors are necessary

for the production of security-relevant items, including land vehicles and military aircraft. The

manufacturing of both are similarly comparative disadvantaged industries in the United States.

We contend that Americans should put a premium on reshoring chip production because of the

salient connection to the availability and cost of household goods such as cars, appliances, and

smart phones. Steel also faced pandemic shortages but was less inextricably linked to the type of

direct, consumer impacts that we expect individuals would associate with chip shortages.17 Indeed,

the media trends plot in Figure 1 suggests that news articles on steel and semiconductors nearly

converge during the time of our studies.18 As such, we specifically hypothesize the following:

Hypothesis 2. Respondents will be more likely to support reshoring when the industry is semicon-

ductors than when the industry is steel.

Initial Survey

To test these hypotheses, we conducted a survey experiment to a diverse sample of 2,000 Amer-

icans in early December 2021 using Lucid.19 We offer further discussion of respondent screening

procedures, the advantages of using Lucid, and experimentation during the COVID-19 pandemic

in Appendix §2. We also follow best practices for screening out low quality responses; we offer

further explanation in Appendix §3. Our final data set includes just over 1,500 responses.

We randomly assign participants to one of four conditions. Respondents receive information

about one of two industries (steel or semiconductors) and one of two countries (China or South
17
Though President Trump did seek to revive American steel production in 2018 by levying tariffs on Chinese steel
and aluminum and instituting tax cuts for American producers. See York, Erica. “President Trump Announces Two
Steep Tariffs on Steel and Aluminum.” Tax Foundation. March 2, 2018. https://bit.ly/2OsPTX1
18
Also see the Google Trends plot in Appendix Figure A2,
19
Our research accords with APSA’s ethics principles — for discussion, see Appendix §6.

11
Korea) that correspond to our hypotheses as described above. The specific text of the vignette is

as follows: “In recent years, the United States has lost global market share of the [semiconduc-

tor or steel] manufacturing industry to [China or South Korea]. The United States government

has periodically proposed incentives and investments to reshore more manufacturing – in other

words bring more manufacturing home – from [China or South Korea]." While there are several

forms of industrial policy that promote reshoring with varying effectiveness, including tax breaks,

subsidies, tariffs, and import controls, we are agnostic about the specific government policy that

drives firms to reshore. We utilize language that mirrors media coverage and political statements,

which emphasize government “incentives and investments” in general terms.20 We believe this is

the level at which that most Americans consume information about and form opinions on gov-

ernment initiatives to spur reshoring, especially given the public’s low awareness of the technical

aspects of trade and investment policies (Rho and Tomz 2017) and their focus on outcomes over

processes (Snyder and Borghard 2011). Our approach is also mindful of the advantages of abstract

and concise experimental vignettes (Brutger et al. 2022).

Our main dependent variable is measured based on responses to the following question: “Do

you support or oppose the series of incentives and investments to prop up the US [steel or semicon-

ductor] industry?" Respondents answer on a five-point scale from "Strongly oppose" to "Strongly

support." We also ask standard demographic questions as well as a nationalism question, posed as

follows: “When someone says something bad about the American people, how strongly do you

feel it is as if they said something bad about you?" and it is measured on five-point scale from

“Not strongly at all” to “Extremely strongly.” Descriptive statistics for our sample can be found in
20
See e.g., “The IRA Will Bring Some Manufacturing Back to the US.” Forbes. 2023. https://bit.ly/
3JX0NOi.

12
Appendix Table A1. Appendix Figure A3, meanwhile, shows how our dependent variable is dis-

tributed across respondents. Last, a detailed description of relevant survey questions can be found

in Appendix §3.

To start, we examine levels of support broken out by treatment arm, finding patterns that con-

tradict our hypotheses. These results are shown in panel (a) of Figure 2. Americans support

reshoring from both China and South Korea — 66 percent of respondents expressed support for

bringing manufacturing home from each country despite apparent differences in security ties be-

tween them and the United States. Moreover, we find evidence that U.S. citizens are more support-

ive of reshoring the production of steel than semiconductors, though support for reshoring both

remains high in absolute terms. This findings perhaps suggest an important role for nostalgia poli-

tics (Jardina 2019; Clark, Khoban and Zucker 2022), or an expressed desire for the U.S. to return

to an era of past manufacturing prowess. We further probe this explanation in our follow-up study.

We also calculate the percent of respondents that support reshoring across political party affilia-

tions. Strikingly, we do not identify major differences based on respondents’ partisanship; average

support for reshoring is nearly identical among Democrats and Republicans, as panel (b) of Figure

2 shows, although considerably lower for Independents. These patterns mirror bipartisan elite con-

sensus on reshoring chips. In an era of heightened polarization (Fiorina and Abrams 2008; Iyengar

and Westwood 2015), reshoring appears to be a rare area of partisan and elite-public agreement

(Kertzer 2020; Dellmuth et al. 2021).

Next, we more systematically test the main effects of our treatments on respondents’ levels of

support for reshoring. We find little support for each of our hypotheses. With respect to H1, we

find a treatment effect of 0.042 (p = 0.163) for the China treatment relative to the South Korea

13
 0.8 0.8

0.7 0.71 0.7

0.7 0.66 0.66 0.7

Support

Support
0.62

0.6 0.58
0.6

0.5
rs 0.5
a cto
ina ore el u t an en
t
Ch t hK Ste co
nd o cra blic e nd
So
u mi em pu ep
Se D Re Ind

(a) Treatment (b) Partisanship

Figure 2: Support for Reshoring by Treatment and Party (Experiment 1). Figure 2 shows the
share of respondents that offered support for reshoring by responding either “Strongly support” or
“Moderately support” to our outcome question. 95 percent confidence intervals are included.

condition.21 With respect to H2, we find a negative and statistically significant treatment effect

for semiconductors versus steel (-0.115, p = 0.002). This corresponds to a roughly three percent

decrease in support for reshoring when moving from the steel to semiconductor condition.

We also model our results in a regression framework, both with and without the inclusion of

relevant covariates, and they are similar (Table 1) — we identify no effect of the China treatment

and a negative and statistically significant effect for the semiconductor one. Several covariates

achieve statistical significance as well. Those who are nationalistic, men, and older individuals are

all more supportive of reshoring.

That nationalistic individuals favor reshoring is unsurprising since they see their identity as

closely tied to that of the country as a whole. Moreover, that men are more supportive of reshoring

contradicts existing work on public attitudes toward different aspects of globalization such as im-

migration and trade (Scheve and Slaughter 2001; Mayda and Rodrik 2005; Mansfield and Mutz

2009; Guisinger 2009). On the other hand, it is possible that reshoring elicits enthusiasm from
21
This is a bootstrapped treatment effect from 1,500 draws, though results are nearly identical for a simple
difference-in-means. The same is true for our semiconductor treatment.

14
men since they were the primary losers from the offshoring of manufacturing industries like steel

(Clark, Khoban and Zucker 2022), and similar trends like decarbonization (Bush and Clayton

2023), while growing service sectors like healthcare mostly employ women (Winant 2021). More-

over, while we identify no effect of partisanship at least between the two major parties (Indepen-

dents are less supportive than partisans), we find that respondents who reported voting for Trump

in 2020 are more likely to endorse reshoring — perhaps more as a result of their ethno-nationalist

and populist ideological dispositions than their political ideology (Rodrik 2018; Broz, Frieden and

Weymouth 2021).

Last, we detect a positive association between both income and education and support for

reshoring. Canonical theories of trade expect wealthier and more highly educated Americans to

be more supportive of free trade (and possibly less supportive of reshoring as a distortion of free

trade) given the abundance of skilled labor in the U.S. (Scheve and Slaughter 2001). Further, these

two factors are proxies for cosmopolitanism, which should also be associated with higher levels of

support for economic interconnectedness rather than isolationism (Hainmueller and Hiscox 2006;

Mansfield and Mutz 2009). However, it could be that more educated respondents are more aware

of recent debates around issues of supply chain security and product availability. We leave more

conceptual work about the relationship between free trade, reshoring, and demographics for future

study.

On the whole, these results illustrate high levels of support for reshoring across several cross-

sections of the American public. Though we find evidence that Trump voters, who were found

to be distinctly anti-trade in their preferences (Ojeda and Telles 2021), and highly nationalistic

respondents are more likely to support reshoring, we find no clear divides between the two major

parties, and the evidence contradicts our theoretical contentions regarding security and self-interest

15
 Support for reshoring
(1) (2)
China 0.045 0.027
(0.042) (0.041)

Semiconductor −0.116∗∗∗ −0.099∗∗

(0.042) (0.041)

National superiority 0.087∗∗∗

(0.033)

Male 0.154∗∗∗
(0.042)

Age 0.005∗∗∗
(0.001)

Income 0.013∗
(0.007)

Education 0.044∗∗∗
(0.015)

Republican −0.001
(0.057)

Trump voter 0.178∗∗∗

(0.056)

Constant 3.925∗∗∗ 3.097∗∗∗

(0.035) (0.107)

Observations 1,895 1,895

Adjusted R2 0.004 0.055
∗
Note: p<0.1; ∗∗ p<0.05; ∗∗∗ p<0.01

Table 1: Experiment 1 Results.

calculations.

Follow-up Study

Our initial survey does not allow us to determine what specifically drives such high support for

reshoring, both across industries and countries of origin. Therefore, we field a second survey with

16
Qualtrics on a representative sample of 2,250 Americans in September 2022 to better understand

why Americans are supportive of reshoring manufacturing from abroad and what factors might

cause this support to decline. We discuss the Qualtrics sample in greater detail in Appendix §4. In

this survey, we hold the industry of employment constant (semiconductors) to probe mechanisms

in a highly salient context.

Motivated by the results of the first survey, we derive three additional hypotheses about pub-

lic attitudes toward reshoring. We again pre-registered the hypotheses with AsPredicted before

deploying our survey (Appendix §1). First, we explore why support for reshoring was high both

when the target country was an ally (South Korea) and adversary (China) in the results from the

first study. We posit a role for ethnocentrism here — preferences over trade with allies may be

shaped in part by whether the populace looks more or less foreign (Sabet 2016; Mutz 2021) and is

perceived to hold similar values and adhere to similar cultural norms. We note that general feelings

for China among the general public tend to be low.22 In our survey, the average "warmth" from a

5-point Likert for China was 2.29 compared to 3.50 for Canada.23 While reliance on a rival likely

invokes fears of volatility in securing critical goods (Bauerle Danzman 2021), trust in allies may

be shaped by skepticism of out-groups that feel more foreign. As Mansfield and Mutz (2009) note,

individuals may have “a tendency to think less of those who are racially or ethnically different from

one’s own group.” As such, in our second survey, we contrast support for reshoring from China

and Canada, with Canada representing a demographically similar and geographically proximate
22
See: Greenwood, Shannon. “Most Americans Have ‘Cold’ Views of China. Here’s What
They Think About China, In Their Own Words.” Pew Research Center’s Global Attitudes Project, 30
June 2021, https://www.pewresearch.org/global/2021/06/30/most-americans-have-cold-views-of-china-heres-what-
they-think-about-china-in-their-own-words/.
23
As a post-treatment question, we asked respondents, "How do you feel towards [TREATMENT COUNTRY]?"
Numbers reported are thus the average feeling toward China in the China treatment group; and average feeling toward
Canada in the Canada treatment group.

17
ally of the U.S. — shared borders and similar demographic makeups should reduce feelings of

foreignness. This comparison allows us to contrast our findings to the results from the first survey

and observe whether racial, ethnic, and geographic differences alter reshoring perceptions holding

alliance constant.24

Hypothesis 3. Respondents will be more likely to support reshoring when the trade partner is

China than when the trade partner is Canada.

Additionally, Americans might be concerned about economic self-interest but in ways that we

could not easily capture in the first survey. Reshoring drives price increases over the medium-term

because manufacturing is moved away from comparative-advantage countries where labor and in-

puts are typically cheaper. If economic self-interest drives attitudes, Americans may be concerned

about such price increases, especially given high inflation that emerged in the latter part of the

pandemic. This economic concern could manifest at the individual-level or in skepticism about

the state of the national economy as a whole (Mansfield and Mutz 2009); both are compatible with

an economic self-interest story. We therefore introduce information about price, with respondents

receiving information about price increases of either 10 or 25 percent as a result of reshoring, and

we expect Americans to be sensitive to, and less supportive of, reshoring as prices increase.

Hypothesis 4. Respondents will be less likely to support reshoring when the treatment specifies

that it will drive price increases, with support declining in the magnitude of the price increase.

Next, given the absence of evidence for security and economic self-interest considerations in

the first survey, we also consider sociocultural explanations for public support for trade. In particu-

lar, we examine whether support for reshoring is associated with nostalgic sentiment — a yearning
24
While there are other factors that vary between Canada and China, including factor endowments and levels of de-
velopment, we account for relevant individual-level economic factors (e.g., education, income) in subsequent models.

18
for a period of past American manufacturing dominance.25 If this drives our results, respondents

should wish to protect and grow jobs that are perceived to be particularly “American,” which schol-

ars map onto blue-collar industries that involve working with one’s hands (Goldstein, Ballard-Rosa

and Rudra 2021). We capture this by manipulating whether respondents receive information about

the effect of reshoring on jobs (creating either white or blue collar jobs). Support for the creation

of blue over white collar jobs, holding job creation itself constant, would be indicative of a de-

sire to return to an era of past economic greatness in manufacturing rather than to leverage capital

abundance and comparative advantages (Scheve and Slaughter 2001; Mayda and Rodrik 2005). To

account for the latter factors, we also directly control for skill endowments in subsequent tests.

Hypothesis 5. Respondents will be more likely to support reshoring when the treatment specifies

an increase in blue-collar jobs rather than white-collar jobs.

The vignette in the follow-up survey pivots to questions of supply chains, as illustrated in

Figure 3, which shows the condition in which chips are reshored from Canada, creating blue-collar

jobs and driving a 25 percent price increase. The dependent variable is the same as the first study;

full question text can be found in Appendix §3. Descriptive statistics appear in Appendix Table

A2, and a histogram of support for reshoring can be found in Appendix Figure A4.

To begin, we present the results by treatment arm (Figure 4, panel a) which reveals variation

that largely aligns with our expectations. Support for reshoring is higher when the source country is

China (64 percent) than Canada (60 percent). The treatment effect on the five-point scale associated

with our dependent variable is 0.104, and this difference is statistically significant (p = 0.008).26

Next, 63 percent of respondents support reshoring when it generates blue collar jobs, while only
25
See e.g., Clark, Khoban and Zucker (2022)
26
As above, this is a bootstrapped treatment effect from 1,500 draws, though results are similar for a simple
difference-in-means as the plot shows.

19
Figure 3: Sample Vignette. In this vignette, respondents are told that semiconductor manufactur-
ing will be reshored from Canada (a close regional ally), leading to the creation of a large number
of blue-collar jobs, though it will also lead to a 25 percent increase in the price of smartphones.

20
61 percent support it when it creates white collar jobs; the difference on our five-point scale fails

to achieve statistical significance at conventional levels, though it points in the expected direction

(effect = 0.046, p = 0.144). The largest effect, however, is driven by the price increase associated

with reshoring — only 58 percent support bringing manufacturing home when it results in a 25

percent price increase, though 66 percent are supportive when the price increase is only 10 percent.

The effect size here is 0.159 on a five-point scale (p = 0.0007). Substantively, the China effect

corresponds to a 3 percent increase in support for reshoring, and the price effect corresponds to a

4.5 percent increase. Since respondents are not significantly more supportive of reshoring to create

blue versus white collar jobs, and given their sensitivity to price increases, it appears respondents

are more attuned to economic self-interest than nostalgia politics in the context of semiconductors.

This is despite the fact that nostalgia may have driven support for reshoring steel in the first study.

Next, we disaggregate by partisanship, as is illustrated in panel b of Figure 4. Democrats and

Republicans once again express similar levels of support for reshoring, and Independents remain

more skeptical of moving semiconductor manufacturing to the United States. Still, reshoring ap-

pears to be a broad area of bipartisan consensus in an otherwise polarized polity.

Overall, the results from the second survey point to high levels of support for reshoring, but they

go further than the first study to probe the basis of that support. We find that the security preference

for partnering with allies abroad may be sensitive to geography and/or the specific demographic

composition of the country, with fewer differences between reshoring from China compared to

South Korea than from China compared to Canada. The finding is consistent with previous studies

that have pointed to trade preferences based on ethnocentric perceptions of "foreignness" (Sabet

2016; Mutz 2021; Bankert, Powers and Sheagley 2022) — Canada is more phenotypically similar

and geographically proximate than South Korea to the United States.

21
 0.8 0.8

0.7 0.66 0.7 0.67

0.65

Support
Support

0.64 0.63
0.6 0.61
0.58
0.6 0.6
0.53

0.5 0.5
r t
a r
oll
a
cra
t an en
a da in a c oll c 0 0 50 o blic e nd
an Ch lue hit
e 11 12 em pu ep
C B W D Re Ind

(a) Treatment (b) Partisanship

Figure 4: Support for Reshoring by Treatment and Party (Experiment 2). Figure 4 shows the
share of respondents that offered support for reshoring by responding either “Strongly support” or
“Moderately support” to our outcome question. Panel a shows treatment effects, including country,
white versus blue collar, and price increase based on a standard iPhone cost with 10 and 25 percent
cost increases. Panel b shows heterogeneous effects by party. 95 percent confidence intervals are
included.

We also find that the public is sensitive to price increases driven by the return of manufactur-

ing to countries like the U.S. that may have a comparative disadvantage in the production of the

corresponding goods. Though most individuals are willing to pay a premium for American-made

products, that willingness wanes as prices increase, offering evidence for economic self-interest

arguments above a certain threshold (Scheve and Slaughter 2001; Mayda and Rodrik 2005).27 This

may be why the government has aggressively supported chip manufacturing in the U.S. through

subsidies that help to control prices.28 Lastly, we find little evidence of preferential support for the

creation of blue- over white-collar jobs; this suggests nostalgia may have been a weaker driver of
27
We note that in-group preferences tend to follow sociotropic ones: those who respond that they believe that
reshoring would benefit the country are also likely to respond that reshoring would benefit people like them — these
two variables have a Pearson’s correlation coefficient of around 0.60. This suggests our results may be driven by
either consideration, or both. To capture sociotropic preferences on reshoring, we asked respondents, "To what extent
do you agree with the following statement: The series of incentives and investments to bring U.S. manufacturing in
semiconductors home from [TREATMENT COUNTRY] is good for the United States." And for in-group preferences,
we asked, "To what extent do you agree with the following statement: The series of incentives and investments to
bring U.S. manufacturing in semiconductors home from [TREATMENT COUNTRY] is good for people like me."
28
Lee, Jane Lanhee. “U.S. Congress Passage of Subsidies Prompts Chip Makers to Move on Projects." Reuters.
July 28, 2022. https://reut.rs/3Uh25as.

22
support for reshoring than ethnocentrism or price concerns.

We again model our results in a regression framework, both with and without the inclusion of

relevant covariates, and the findings are unchanged (Table 2). Respondents are more supportive of

reshoring from China and when price increases are modest. Covariate results, meanwhile, mostly

mirror those from the first study. Those who are nationalistic, men, older respondents, and richer

and more highly educated individuals are more supportive of reshoring, as in the first study. Last,

we notably identify no relationship between partisanship or self-reported voting for Trump and

support for reshoring — stark evidence of the bipartisan nature of this support.

Finally, we conduct ex-post analysis of heterogeneous treatment effects to further tease out

mechanisms. We find that support is highest among two sub-groups — those employed in man-

ufacturing and white respondents — as illustrated by Figure 5. In the eyes of those in the manu-

facturing sector, China may be perceived as the main country to which manufacturing jobs fled in

recent decades. The impact of Chinese imports on areas that served as former U.S. manufacturing

hubs has been severely negative (Margalit 2011; Colantone and Stanig 2018), driving economic

volatility and political polarization (Autor et al. 2020). It is no surprise then that these individuals

are especially eager to reshore semiconductor manufacturing from China. Notably, manufacturing

workers are also enthusiastic about reshoring blue collar jobs from Canada, likely because they

work in blue collar sectors themselves. These results indicate sensitivity to economic self-interest

considerations among manufacturing workers.

We also find that support for reshoring is much higher among white respondents than non-white

respondents. Declining industries subject to job losses amidst globalization disproportionately em-

ployed white men at their peak (Baccini and Weymouth 2021). White respondents may identify

reshoring as a means to bolster the economic welfare and social status of their racial in-group

23
 Support for reshoring
(1) (2)
∗∗
China 0.098 0.097∗∗
(0.041) (0.041)

Blue collar 0.046 0.043

(0.041) (0.041)

High price −0.154∗∗∗ −0.148∗∗∗

(0.041) (0.041)

National superiority 0.079∗∗

(0.032)

Male 0.191∗∗∗
(0.044)

Age 0.130∗∗∗
(0.027)

Income 0.024∗∗∗
(0.007)

Education 0.041∗∗∗
(0.015)

Republican 0.017
(0.057)

Trump voter 0.024

(0.056)

Constant 3.747∗∗∗ 2.886∗∗∗

(0.042) (0.113)

Observations 2,254 2,240

Adjusted R2 0.008 0.043
∗
Note: p<0.1; ∗∗ p<0.05; ∗∗∗ p<0.01

Table 2: Experiment 2 Results.

(Baccini and Weymouth 2021; Clark, Khoban and Zucker 2022). The race results also point to an

important role for perceptions of “foreignness” at the individual-level. Non-white respondents are

more supportive of reshoring from Canada — a predominantly white country with similar demo-

graphic make-up to the U.S. — compared to China while the opposite is true for white respondents.

24
This offers suggestive support for our posited ethnoracial mechanism.
HTE by Race HTE by Manufacturing Sector
Non−White White Manufacturing Non−Manufacturing

0.6 0.75

0.4 0.50
Job Type Job Type
Support

Support
blue collar blue collar

white collar white collar

0.2 0.25

0.0 0.00

Canada China Canada China Canada China Canada China

Country Treatment Country Treatment

(a) HTE by Race (b) HTE by Manufacturing Sector

Figure 5: Heterogeneous Treatment Effects by Race and Manufacturing Sector. Figure 5

shows mean standard errors of the country and jobs treatments across different subgroups.

Discussion

Our results suggest that attitudes toward reshoring are different from those toward trade —

they are more bipartisan and less sensitive to alliance ties and economic self-interest considera-

tions than we anticipated. It follows that examining reshoring through a trade lens may be inappro-

priate; reshoring appears to be a distinct phenomenon for Americans despite the observable trade

implications of moving industries from abroad back home.

When firms move back to a country from which they previously offshored production, it im-

pacts the country’s trade balance — the country imports fewer goods from abroad, and more prod-

ucts are made at home. Such a move cuts against traditional theories of free trade, which emphasize

the economic benefits of specialization according to comparative advantage (i.e., producing goods

25
in countries with abundant, cheap labor). But reshoring does not necessarily correlate with the

implementation of protectionist trade policies (e.g., higher tariffs or quotas on imports). Instead,

governments may grant firms tax cuts or subsidies to incentivize them to return home,29 or firms

may opt to return of their own accord as a result of supply chain disruptions (e.g., to locate them-

selves more closely to key inputs or consumer markets).

For these reasons, publics may not think about reshoring as the inverse of trade — it is a

decision made by firms, often in the context of complex supply chains, in response to market

forces, geopolitical tensions, exogenous shocks like pandemics, and countries’ industrial policies.

Instead, Americans endorse reshoring in most cases and across partisan divides, which may be

good news for Biden’s “Made in America” agenda.

Conclusion

This paper finds that Americans are broadly supportive of reshoring and that this support is

consistent across the two major political parties. We further probe the elasticity and basis of such

enthusiasm for reshoring in the U.S., finding that large price increases and reshoring from ge-

ographically proximate allies attenuates this support, though it remains high in absolute terms.

Our results suggest that while Americans may be willing to pay higher prices for American-made

goods, they are also more comfortable trading with close and demographically similar allies, and

they are sensitive to significant strain on their pocketbooks. We therefore extend and revise existing

literatures interested in the geopolitics of trade (Gowa and Mansfield 1993; Carnegie and Gaikwad

2022) and how security and economic self-interest inform trade opinion in high information envi-
29
Though such policies may violate the WTO’s national treatment provision and therefore also be classified as
protectionism.

26
ronments (Rho and Tomz 2017; Casler and Clark 2021).

Our research carries important implications for policymakers and practitioners, especially as

the United States has ramped up support for domestic semiconductor manufacturers and escalated

its semiconductor export restrictions toward China. First, we find that a strategy of “friendshoring”

may help the U.S. to balance the economic benefits derived from trade in comparative advantage

industries in ways that are compatible with the public’s preference for trade with allies. However,

an important caveat is that the ally may need to be geographically and demographically close to

the U.S. for support to be robust. Second, our findings imply, given the price sensitivity of the

public, that the policy may face headwinds if or as manufacturing at home drives up consumer

prices given higher labor costs. This means it may behoove policymakers to consider mitigation

strategies such as encouraging more inward immigration of both high and low skilled laborers that

are required for reshoring.

Although our research has taken important steps toward understanding a growing yet under-

studied economic policy, we suggest several potential paths for follow-on research. First, while

focused on semiconductor reshoring, we encourage scholars to study the reshoring of other indus-

tries, and specifically those that citizens may weigh differently than chips. For instance, direct

household products such as toilet paper or baby formula, or personal protective equipment that

presents clear health dividends, might present different logics of support. We urge further study

to understand more systematically whether there are patterns of support across reshoring domains,

perhaps through a conjoint experiment that varies the product, country, and change in cost.

Next, while the policy environment suggests broad alignment around protectionism among po-

litical elites and the public, the success of these investment and incentive policies requires buy-in

from industry elites, whose views may diverge with either of these two sets of actors (Kertzer

27
2020; Dellmuth et al. 2021). In the semiconductor context, the post-pandemic economic downturn

has dampened demand and sent stock prices plummeting, raising questions of whether and how

industry will respond to government incentives that make up only a fraction of the efficiency losses

associated with reshoring. Studying industry elites would shed light on the attitudes of key stake-

holders required for these policies to be realized. Scholars might also probe how policymakers

weigh public opinion against lobbying from upstream and downstream industries in this domain;

a large body of work shows that lobbyists can often dictate the terms of trade (Osgood et al. 2017;

Kim 2017). Understanding these dynamics, including the way they play out at the subnational

level, would add granularity to studies of the impact of these policies.

Lastly, while we identify bipartisan convergence regarding the attitudes we studied, previous

research suggests that the two major parties diverge in terms of the basis of their attitudes on

trade (Mutz 2021). We suggest further probing of why respondents from both parties support

reshoring but are split on trade more generally. In particular, future work might examine how

respondents assess different forms of industrial policy to promote reshoring, including subsidies,

tax breaks, and import controls. Our work shows that Americans value the outcome of such policies

— the return of jobs — but does not explore heterogeneity among various government tactics.

Relatedly, we urge additional research on the demographic correlates such as gender and education

that appear to operate in different directions for reshoring than aggregate trade, and research to

help tease out whether sympathy toward friendshoring is grounded in proximity or ethnocentrism,

perhaps by studying attitudes toward reshoring to Mexico, which has been a beneficiary of the

recent manufacturing pullback from China.30

30
Ben-Achour, Sabri. “Shifting Supply Chains Settle on Mexico.” Marketplace. May 2, 2022. https://bit.
ly/3EafFpx

28
References

Acemoglu, Daron, David Autor, David Dorn, Gordon H. Hanson and Brendan Price. 2016. “Import

Competition and the Great US Employment Sag of the 2000s.” Journal of Labor Economics

34(S1):141–198.

Autor, David, David Dorn, Gordon Hanson and Kaveh Majlesi. 2020. “Importing Political Polar-

ization? The Electoral Consequences of Rising Trade Exposure.” American Economic Review

110(10):3139–3183. NBER Working Paper 22637.

Baccini, Leonardo and Stephen Weymouth. 2021. “Gone For Good: Deindustrialization, White

Voter Backlash, and U.S. Presidential Voting.” American Political Science Review 115:550–567.

Ballard-Rosa, Cameron, Amalie Jensen and Kenneth Scheve. 2022. “Economic Decline, Social

Identity, and Authoritarian Values in the United States.” International Studies Quarterly 66(1).

Bankert, Alexa, Ryan Powers and Geoffrey Sheagley. 2022. “Trade Politics at the Checkout Lane:

Ethnocentrism and Consumer Preferences.” Political Science Research and Methods Forthcom-

ing.

Barbieri, Katherine, Omar M.G. Keshk and Brian Pollins. 2009. “Trading Data: Evaluating our

Assumptions and Coding Rules.” Conflict Management and Peace Science 26(5):471–491.

Bauerle Danzman, Sarah. 2021. Investment Screening in the Shadow of Weaponized Interde-

pendence. In The Uses and Abuses of Weaponized Interdependence, ed. Daniel Drezner, Henry

Farrell and Abraham Newman. Brookings, Washington, D.C.

Broz, J. Lawrence, Jeffry Frieden and Stephen Weymouth. 2021. “Populism in Place: The Eco-

nomic Geography of the Globalization Backlash.” International Organization 75(2):464–494.

Brutger, Ryan, Joshua D. Kertzer, Jonathan Renshon, Dustin Tingley and Chagai M. Weiss. 2022.

29
 “Abstraction and Detail in Experimental Design.” American Journal of Political Science Forth-

coming.

Bush, Sarah and Amanda Clayton. 2023. “Facing Change: Gender and Climate Change Attitudes

Worldwide.” American Political Science Review 117(2):591–608.

Carnegie, Allison and Nikhar Gaikwad. 2022. “Public Opinion on Geopolitics and Trade: Theory

and Evidence.” World Politics 74(2):167–204.

Casler, Don and Richard Clark. 2021. “Trade Rage: Audience Costs and International Trade.”

Journal of Conflict Resolution 65(6):1198–1230.

Clark, Richard, Roza Khoban and Noah Zucker. 2022. “Breadwinner Backlash: The Gendered

Political Consequences of Industrial Decline.” IPES 2022 Annual Meeting .

Colantone, Italo and Piero Stanig. 2018. “The Trade Origins of Economic Nationalism: Import

Competition and Voting Behavior in Western Europe.” American Journal of Political Science

62(4):936–953.

Davis, Christina and Sophie Meunier. 2011. “Business as Usual? Economic Responses to Political

Tensions.” American Journal of Political Science 55(3):628–646.

Davis, Christina and Tyler Pratt. 2020. “The Forces of Attraction: How Security Interests Shape

Membership in Economic Institutions.” Review of International Organizations Forthcoming.

Dellmuth, Lisa, Jan Aart Scholte, Jonas Tallberg and Soetkin Verhaegen. 2021. “The Elite-Citizen

Gap in International Organization Legitimacy.” American Political Science Review . Forthcom-

ing.

Druckman, James N., Erik Peterson and Rune Slothuus. 2013. “How Elite Partisan Polarization

Affects Public Opinion Formation.” American Political Science Review 107(1):57–79.

Farrell, Henry and Abraham Newman. 2019. “Weaponized Interdependence: How Global Eco-

30
 nomic Networks Shape State Coercion.” International Security 44(1):42–79.

Fiorina, Morris P. and Samuel J. Abrams. 2008. “Political Polarization in the American Public.”

Annual Review of Political Science 11:563–588.

Goldstein, Judith, Cameron Ballard-Rosa and Nita Rudra. 2021. “Trade as Villain: The Fading

American Dream and Declining Support for Globalization.” International Political Economy

Society Annual Meeting .

Gowa, Joanne and Edward D. Mansfield. 1993. “Power Politics and International Trade.” The

American Political Science Review 87(2):408–420.

Green, Donald P., Bradley Palmquist and Eric Schickler. 2002. Partisan Hearts and Minds: Polit-

ical Parties and the Social Identities of Voters. New Haven, CT: Yale University Press.

Greenhill, Brian, Layna Mosley and Aseem Prakash. 2009. “Trade-based Diffusion of Labor

Rights: A Panel Study 1986–2002.” American Political Science Review 103(4):669–690.

Guisinger, Alexandra. 2009. “Determining Trade Policy: Do Voters Hold Politicians Account-

able?” International Organization 63(3):533–557.

Hainmueller, Jens and Michael J. Hiscox. 2006. “Learning to Love Globalization: Education and

Individual Attitudes Toward International Trade.” International Organization 60(1):469–498.

Hemmer, Christopher and Peter J. Katzenstein. 2002. “Why is There No NATO in Asia? Col-

lective Identity, Regionalism, and the Origins of Multilateralism.” International Organization

56(3):575–607.

Iyengar, Shanto and Sean J. Westwood. 2015. “Fear and Loathing across Party Lines: New Evi-

dence on Group Polarization.” American Journal of Political Science 59(3):690–707.

Jardina, Ashley. 2019. White Identity Politics. Camridge University Press, New York, NY.

Johns, Leslie and Rachel Wellhausen. 2016. “Under One Roof: Supply Chains and the Protection

31
 of Foreign Investment.” American Political Science Review 110(1):31–51.

Kertzer, Joshua D. 2020. “Re-Assessing Elite-Public Gaps in Political Behavior.” American Jour-

nal of Political Science . Available on Early View.

Kim, In Song. 2017. “Political Cleavages within Industry: Firm-level Lobbying for Trade Liber-

alization.” American Political Science Review 111(1):1–20.

Kim, In Song, Steven Liao and Sayumi Miyano. 2023. “Why FDI and Trade Politics Should Be

Studied Together.”. Manuscript, Massuchusetts Institute of Technology.

URL: http://web.mit.edu/insong/www/pdf/fditrade.pdf

Kojola, Erik. 2019. “Bringing Back the Mines and a Way of Life: Populism and the Politics of

Extraction.” Annals of the American Association of Geographers 109(2):371–381.

Kolcava, Dennis, E. Keith Smith and Thomas Bernauer. 2022. “Cross-National Public Acceptance

of Sustainable Global Supply Chain Policy Instruments.” Nature Sustainability Forthcoming.

Krasner, Stephen D. 1976. “State Power and the Structure of International Trade.” World Politics

28(3):317–347.

Malesky, Edmund J. and Layna Mosley. 2018. “Chains of Love? Global Production and the Firm-

Level Diffusion of Labor Standards.” American Journal of Political Science 62(3):712–728.

Mansfield, Edward D. and Diana C. Mutz. 2009. “Support for Free Trade: Self-Interest, So-

ciotropic Politics, and Out-Group Anxiety.” International Organization 63(3):425–457.

Mansfield, Edward D., Helen V. Milner and Nita Rudra. 2021. “The Globalization Back-

lash: Exploring New Perspectives.” Comparative Political Studies 54(13):2267–2285. doi:

10.1177/00104140211024286.

URL: https://doi.org/10.1177/00104140211024286

Mansfield, Edward D. and Rachel Bronson. 1997. “Alliances, Preferential Trading Arrangements,

32
 and International Trade.” American Political Science Review 91:94–107.

Margalit, Yotam. 2011. “Costly Jobs: Trade-related Layoffs, Government Compensation, and

Voting in U.S. Elections.” American Political Science Review 105(1):166–188.

Mayda, Anna Maria and Dani Rodrik. 2005. “Why Are Some People (and Countries) More Pro-

tectionist than Others?” European Economic Review 49(6):1393–1430.

Mosley, Layna. 2010. Labor Rights and Multinational Production. Cambridge University Press,

Cambridge, UK.

Mutz, Diana C. 2021. Winners and Losers: The Psychology of Foreign Trade. Princeton University

Press, Princeton.

Ojeda, Raul Hinojosa and Edward Telles. 2021. “Trump Paradox: How Immigration and Trade

Affected White Voting and Attitudes.” Socius 7.

Osgood, Iain, Dustin Tingley, Thomas Bernauer, In Song Kim, Helen Milner and Gabriele Spilker.

2017. “The Charmed Life of Superstar Exporters: Survey Evidence on Firms and Trade Policy.”

Journal of Politics 79(1):133–152.

Rho, Sungmin and Michael Tomz. 2017. “Why Don’t Trade Preferences Reflect Economic Self-

Interest?” International Organization 71(S1):S85–S108.

Rodrik, Dani. 2018. “Populism and the Economics of Globalization.” Journal of International

Business 1:12–23.

Rommel, Tobias and Stefanie Walter. 2018. “The Electoral Consequences of Offshoring: How

the Globalization of Production Shapes Party Preferences.” Comparative Political Studies

51(5):621–658.

Sabet, Shahrzad. 2016. “Feelings First: Non-Material Factors as Moderators of Economic Self-

Interest Effects on Trade Preferences.”.

33
Scheve, Kenneth F. and Matthew J. Slaughter. 2001. “What Determines Individual Trade-Policy

Preferences?” Journal of International Economics 54(2):267–292.

Schewe, Eric. 2021. “Semiconductor Shortages End an Era of Globalization.” JSTOR Daily .

Snyder, Jack and Erica D. Borghard. 2011. “The Cost of Empty Threats: A Penny, Not a Pound.”

American Political Science Review 105(3):437–456.

Vaughn, Abigail and Ryan Weldzius. 2021. “Reshoring Global Supply Chains.”. Working paper.

URL: https://bit.ly/3UC45dL

Walter, Stefanie. 2021. “The Backlash Against Globalization.” Annual Review of Political Science

24:421–442.

Winant, Gabriel. 2021. The Next Shift: The Fall of Industry and the Rise of Health Care in Rust

Belt America. Cambridge, Mass.: Harvard University Press.

34
 Geopolitics and the Political Economy of Reshoring: Evidence
from Semiconductors

Online Appendices

Contents
1 Pre-Registration 2

2 Lucid, COVID-19, and Respondent Screening 2

3 Survey Procedures and Questions 3

3.1 Vignette (Survey 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.2 Vignette (Survey 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.3 Dependent variable (Surveys 1 and 2) . . . . . . . . . . . . . . . . . . . . . . . . 3
3.4 Covariates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

4 Qualtrics 5

5 Supporting Statistical Information 6

6 Discussion of Ethical and Human Subjects Principles 12

* Richard
Clark ([email protected]) is Assistant Professor of Government at Cornell University. Sarah
Kreps is John L. Wetherill Professor of Government at Cornell University. Adi Rao is Ph.D. candidate at Cornell
University. We thank Ryan Brutger, Sarah Baurle Danzman, Aycan Katitas, Abe Newman, Tom Pepinsky, and Chris
Way for helpful comments and suggestions.

1
1 Pre-Registration
All hypotheses were pre-registered with AsPredicted. An anonymous PDF of the pre-registration
for the first two hypotheses can be found here: https://aspredicted.org/1Z8_4WY.
An anonymous version of the pre-registration for the last three hypotheses can be found here:
https://aspredicted.org/K81_TKC.

2 Lucid, COVID-19, and Respondent Screening

Our study was fielded using Lucid’s “Theorem” respondent pool. Lucid uses a quota system
to target a nationally representative sample. Survey samples recruited with Lucid track well with
U.S. national benchmarks and are increasingly used in social science research, not for descriptive
work where probability samples are needed, but for assessing relationships between variables, as
is the objective here (Coppock and McClellan 2019).1
Given that our study was fielded during the Covid-19 pandemic, we recognize that the pan-
demic could affect our results. Research studying the effects of the pandemic on survey quality
and generalizability shows that respondent attentiveness declined during the pandemic among on-
line sample populations, and researchers are advised to use attention checks to screen for quality
respondents (Aronow et al. 2020). We use such pre-treatment checks in our study. That said,
(Peyton, Huber and Coppock 2020) find that studies conducted during the pandemic consistently
replicate pre-pandemic studies, though for some the effect sizes are smaller, which suggests that
the timing of our study may yield more conservative estimates than if we had conducted the same
study before the pandemic.
All respondents were provided with an informed consent form and had the option to opt-out of
the study. Consistent with best practices for data quality (Burleigh, Kennedy and Clifford 2018), re-
spondents were blocked from participating if they were located outside of the U.S. or were flagged
for using a Virtual Private Server (VPS). We collected responses until we had 2,000 respondents,
as outlined in out pre-registration.

1
See e.g., Kim and Margalit (2021).

2
3 Survey Procedures and Questions
Following the practices outlined by Burleigh, Kennedy and Clifford (2018), we implemented
techniques to block respondents from participating if they could not verify they were located in
the U.S. or they were using a Virtual Private Server (VPS) to mask their location. We also had
respondents answer a free response question unrelated to our study, which we used as an attention
check. Respondents who wrote gibberish or who entered text that was unrelated to the question
were removed from the sample, which resulted in 23 percent of the sample being excluded.

3.1 Vignette (Survey 1)

In recent years, the United States has lost global market share of the [semiconductor or steel]
manufacturing industry to [China or South Korea]. The United States government has periodically
proposed incentives and investments to reshore more manufacturing – in other words bring more
manufacturing home – from [China or South Korea].

3.2 Vignette (Survey 2)

As you know, Americans buy products that are either made in foreign countries or have parts
that are made in foreign countries. For example, the U.S. currently imports semiconductors in
large volumes from [China or Canada].
In recent months, these supply chains have become strained, and there is debate about whether
the U.S. should provide incentives and investments to American companies to bring their manu-
facturing home.
If implemented, supporters say these investments would bring large numbers of [blue or white
collar jobs] back to the United States. But some skeptics of these proposals have suggested that
they will increase prices of relevant goods by [10 or 25%] . This means that a smartphone, which
currently costs $1000, would cost[$1100 or $1250].
As a summary:
- The U.S. imports semiconductors from [Canada or China].
- Reshoring this industry would create [blue or white collar] American jobs.
- Reshoring this industry would drive [10 or 25%] higher prices for American consumers.

3.3 Dependent variable (Surveys 1 and 2)

• Do you support or oppose the series of incentives and investments to prop up the US [steel
or semiconductor] industry?

– Strongly oppose
– Oppose
– Neither support nor oppose
– Support
– Strongly support

3
3.4 Covariates
We randomized the order of all covariate questions with the exception of the Trump voter
variable, which we asked at the end of the survey to avoid differential attrition. Questions were the
same for both surveys that we fielded unless otherwise noted.

• Nationalism: When someone says something bad about the American people, how strongly
do you feel it is as if they said something bad about you? Measured on five-point scale from
“Not strongly at all” to “Extremely strongly.”

• Male: binary equal to 1 if respondent best identified themselves as male.

• Age (Survey 1): self-identified age in integer form.

• Age (Survey 2): self-identified age on a three-point scale (18-34, 35-54, 55+)

• Income: measured on 12-point scale from “$10,000 or less” to “$150,000 or more.”

• Education: measured on eight-point scale from “Less than High School” to “Professional
Degree (JD, MD).”

• Republican: binary equal to 1 if respondent self-identified as a Republican as opposed to a

Democrat or Independent.

• Trump: the question asks “whom did you vote for in the 2020 presidential election?” Mea-
sured as binary equal to 1 if responded Donald Trump. We ask this question post-treatment to
avoid concerns about differential attrition pre-treatment. We do not observe attrition anyway
based on this question.

4
4 Qualtrics
Qualtrics delivered 2,254 quality completes. They removed poor-quality responses and ensured
that each respondent was a U.S. citizen and passed a basic attention check at the start of the study.
The sample is representative of the U.S. population by Census benchmarks for age, gender, and
region. They did not collect partial responses. The median time to completion was 6 minutes.
They also implemented a speeding check – measured as one-half the median soft launch time –
which terminated those who were not responding thoughtfully.

5
5 Supporting Statistical Information

100

Study 1

Study 2
80
Salience

60

40

20
2019 2020 2021 2022 2023
Week
Figure A1: Salience of Semiconductors (Google Trends). This figure shows public interest from
Google searches in semiconductors, which as a topic includes related search terms like chips. Data
comes from Google trends. Vertical lines demarcate the timing of our two studies.

The Google trends plot in Figure A1 shows that semiconductors elicited public interest at the
time of our studies;2 interest in and concern with chips increased in early 2021 when downstream
industries, which produce goods for consumption such as laptops, cars, and household appliances,
suffered from production shortfalls and delivery lags as a result of chip shortages. These industries,
in turn, largely passed price increases resulting from such shortages onto consumers.

2
Notably, scholars have identified a link between Google trends and Gallup “most important problem" questions
(Mellon 2014). This suggests that Google trends are a good measure of issue salience.

6
 100

Study 1

Study 2
90
Salience

80

70

2019 2020 2021 2022 2023

Week
Figure A2: Salience of Steel. This figure shows public interest from Google searches in steel,
which as a topic includes related search terms. Data comes from Google trends.

7
Statistic N Mean St. Dev. Min Max
Reshoring 1895 3.89 0.91 1 5
Male 1895 0.48 0.50 0 1
Age 1895 46.28 16.82 18 94
Income 1895 5.91 3.36 1 12
Education 1895 3.72 1.54 1 8
Republican 1895 0.29 0.45 0 1
Trump voter 1895 0.34 0.47 0 1

Table A1: Descriptive Statistics (Survey 1).

8
 Statistic N Mean St. Dev. Min Max
Reshoring 2254 3.74 0.98 1 5
Male 2254 0.45 0.50 0 1
Age 2240 2.10 0.81 1 3
Income 2254 5.54 3.28 1 12
Education 2254 3.59 1.53 1 8
Republican 2254 0.32 0.47 0 1
Trump voter 2254 0.35 0.48 0 1

Table A2: Descriptive Statistics (Survey 2). Note that age is measured on a 1-5 scale correspond-
ing to age ranges.

9
 Distribution of DV

0.6
0.4
Density

0.2
0.0

1 2 3 4 5

Figure A3: Distribution of Dependent Variable (Survey 1).

10
 Distribution of DV
1.0
0.8
0.6
Density

0.4
0.2
0.0

1 2 3 4 5

N = 2254 Bandwidth = 0.1434

Figure A4: Distribution of Dependent Variable (Survey 2).

11
6 Discussion of Ethical and Human Subjects Principles
The human subjects research included in this paper complies the with Principles and Guidance
for Human Subjects Research outlined by the APSA and was evaluated by the Institutional Review
Board at the appropriate universities. The study was fielded with the survey firm Lucid and was
reviewed by the IRB at [Name Redacted] (Protocol #: [Number Redacted]) and was declared
exempt. The risks to subjects were evaluated to be minimal and the researchers took steps to ensure
that any potentially identifying information was protected and then redacted prior to making the
data available for analysis and replication. There were no conflicts of interest identified for the
researchers. The data for replication will be made available when the manuscript is published.
For these studies, respondents were initially asked to complete an electronic standard adult
consent form that informed them they were being asked to participate in a voluntary study that
had been evaluated by [Redacted University Name] institutional review board. The consent form
informed respondents they would be asked a variety of questions about their background, political
preferences, and thoughts on government policies, the estimated length of time, the contact infor-
mation for the investigator, and that the study was deemed to be of minimal risk. Respondents
could select “If you wish to participate, please click the ‘I Agree’ button and you will be taken to
the survey.” or “If you do not wish to participate in this study, please select ‘I Disagree’.” If the
latter was selected, the survey was terminated.
As noted above, respondents were recruited by the survey company Lucid. Respondents vol-
untarily choose to opt in to participate in the survey firm’s panel and are compensated based on
the terms of the survey vendor, which can include cash, gift cards, and loyalty reward points. The
respondents were all in the United States and respondents had free choice to opt in or out of the
survey firms’ panels. Respondents also had the opportunity to contact the researcher regarding
any concerns about compensation or the survey itself, but we did not receive any complaints or
concerns from participants. Our study does not engage in deception since each component of the
experiment provided factual information about the policies in question.
With regard to Principal 10 on the impact of the research on the political processes, we do
not believe there is any reason to believe that our studies would have had an impact on political
processes such as elections or policy creation. Respondents were only asked their opinion on
the subject of reshoring. We therefore do not see the survey as presenting any information to
respondents that would alter their political behavior or political processes.

References
Aronow, Peter Michael, Joshua Kalla, Lilla Orr and John Ternovski. 2020. “Evidence of Rising
Rates of Inattentiveness on Lucid in 2020.” SocArXIV Papers .
URL: https://osf.io/preprints/socarxiv/8sbe4/
Burleigh, Tyler, Ryan Kennedy and Scott Clifford. 2018. “How to screen out VPS and international
respondents using Qualtrics: A protocol.” Available at SSRN 3265459 .
Coppock, Alexander and Oliver A. McClellan. 2019. “Validating the Demographic, Political,
Psychological, and Experimental Results Obtained from a New Source of Online Survey Re-
spondents.” Research and Politics . Forthcoming.

12
Kim, Sung Eum and Yotam Margalit. 2021. “Tariffs As Electoral Weapons: The Political Geogra-
phy of the US-China Trade War.” International Organization 75(1):1–38.
Mellon, Jonathan. 2014. “Internet Search Data and Issue Salience: The Properties of Google
Trends as a Measure of Issue Salience.” Journal of Elections, Public Opinion, and Parties
24(1):45–72.
Peyton, Kyle, Gregory A. Huber and Alexander Coppock. 2020. “The Generalizability of Online
Experiments Conducted during the COVID-19 Pandemic.” SocArXIV Papers .
URL: https://osf.io/preprints/socarxiv/s45yg/

13