SSRN Id3997378

The Humanizing Effect of Market Interaction
Colin Harris† Andrew Myers‡ Adam Kaiser§
Abstract:
The quality and quantity of intergroup contact affects how outgroups are perceived. Positive
interaction tends to have a humanizing effect of moral inclusion. Negative interaction instead tends
towards dehumanization and moral exclusion. One avenue of intergroup contact that has been
empirically underexplored is interaction in a market. Do markets generate moral sympathy, or do
they allow us to ignore or deny the moral status of others? We create a measure of moral sentiment
that captures the frequency, valence, and type of moral language used about an outgroup. We
match our novel sentiment data to dyadic measures of market interaction to test if markets act as a
(de)humanizing force. We find a positive relationship between market interaction and the use of
(1) moral, (2) virtuous (but not vice), (3) bridging, and (4) bonding language to talk about a
contacted outgroup. Our results suggest market interaction has a humanizing effect.
JEL Codes: F14, F60, J15, P17, Z1

Keywords: Humanization, Dehumanization, Intergroup Contact, Market Interaction, Moral
Sentiment, Text Analysis
†
Harris14@stolaf.edu; Department of Economics, St. Olaf College.
‡
MyersA@stanford.edu; Department of Political Science, Stanford University.
§
AKaiser7@gmu.edu; Department of Economics, George Mason University.
Electronic copy available at: https://ssrn.com/abstract=3997378

1 Introduction
Market societies tend to be happier (Gropper et al. 2011; Gehring 2013), healthier (Stroup 2007),
and wealthier (Gwartney et al. 1999), with lower levels of income inequality (Apergis and Cooray
2015), infant mortality (Roberts and Olson 2013), and corruption (Goel and Nelson 2005), and
greater levels of social mobility (Corak 2013; Callais and Geloso 2021), education, and life
expectancy (Roberts and Olson 2013). Yet, for all but committed consequentialists, these robust
positive effects are a necessary but insufficient defense against the critics who argue markets are
morally corrupting.
The central moral critique sees markets as a dehumanizing force that either prevents us
from recognizing the moral status of others or encourages us to willfully ignore it. Jean-Jacques
Rousseau ([1754] 2006: 50), for example, argues that the establishment of a property-based market
society “suppressed the cries of natural compassion and the still feeble voice of justice, and filled
men with avarice, ambition, and vice.” Karl Marx ([1844] 2000: 28, 36, original emphasis) instead
emphasizes the corrupting effect of commodification and the resulting “devaluation of man”. A
market society not only produces “man as a commodity… it produces him… as a mentally and
physically dehumanized being”, estranged from his sense of self and from others.
Contemporary market critics follow both lines of argument. G.A. Cohen (2009: 77-78, 40)
argues that markets are “intrinsically repugnant” because they leverage motives of greed and fear
rather than genuine care and reciprocity. The prominence of these “repugnant motives” in a market
causes us to view others as either “sources of enrichment” or “threats to [our] success.” Michael
Sandel (2012: 3) instead attributes the moral failing of markets to their expansion “into spheres of
life where they don’t belong.” Accompanying the expansion is a tacit acceptance to treat the
objects of exchange as mere commodities, valued only according to their market price. But, as
Sandel (2012: 5) argues, many things in life are not “properly valued in this way” and to treat them
as such is to demean and degrade them.
In either account, markets result in our failure to value people “in the appropriate way—as
persons worthy of dignity and respect” (Sandel 2012: 13). Pro-market philosophers have taken
these moral criticisms head-on (see Brennan and Jaworski 2015), yet “the common concern of the
moral critics of markets (i.e. that markets are morally corrupting) is at root an empirical, rather
than a philosophical, claim” (Storr and Choi 2019: 12). When we interact with other people in a

market, do we come to view them as adversaries, mere commodities, or as equal persons worthy
of dignity and respect?
Two strands of literature come close to answering this question. The findings and data in
each are, however, limited for directly assessing the market (de)humanization hypothesis. The
literature on intergroup contact is concerned with the effect of any interaction between social
groups and as such has given less attention to interaction in a market. The literature on the morality
of markets is concerned with the overall moral character of a market society and as such has given
less attention to how market interaction affects our perception of others. We bridge these two
literatures by investigating how interacting with other people in a market affects our willingness
to consider their humanity.
Testing the market (de)humanization hypothesis requires both (1) a measure of moral
sentiment about different social groups and (2) a dyadic measure of market interaction that can be
matched to those groups. To measure interaction in a market, we use bilateral trade flows and
immigration counts by country of origin. To create an accompanying measure of moral sentiment,
we use text analysis on a corpus of New York Times articles from 1987-2007. Our sentiment
measure captures the frequency, valence, and type of moral language used about an outgroup.
We use these measures in fixed effects regression models to test the predictions that stem
from the philosophical perspectives on the (de)humanizing nature of markets. We find a positive
relationship between market interaction and the use of (1) moral, (2) virtuous (but not vice), (3)
bridging, and (4) bonding language to talk about a contacted outgroup. Our results suggest market
interaction has a humanizing effect.
The rest of the paper proceeds as follows. We first review the literature on
(de)humanization, intergroup contact, and the morality of markets. In doing so, we identify
existing limitations that help motivate and guide our empirical strategy (section 2). We then
explain our data and approach, including how we create our novel sentiment measure and how the
use of moral language maps onto the different forms of (de)humanization (section 3). We then
outline the different philosophical positions on the (de)humanizing nature of markets and identify
three distinct hypotheses (section 4). We connect these hypotheses to our data to generate the
distinct predictions of each position and then test these predictions and present our results (section
5). We conclude with avenues for future research (section 6).

2 (De)Humanization, Intergroup Contact, and Markets
To humanize is to recognize in others the qualities essential to being human. Which qualities are
deemed essential can vary based on experience and self-reflection (Gopnik and Wellman 1992;
Frith and Frith 2003; Wimmer and Perner 1983), but common essential qualities include cognitive
ability, secondary emotions, and moral agency (Haslam et al. 2008; Haslam and Loughnan 2014).
To not recognize or not acknowledge these qualities in other people is to dehumanize them.
Humanization often accompanies identity as people consistently attribute these qualities to those
in their ingroup but may fail to recognize or even actively deny the same qualities to members of
an outgroup (Gudykunst 1985; Izard 1960; Izuma and Adolphs 2013; Kinzler et al. 2009; Taft et
al. 2011).
There are two main ways an outgroup may be (de)humanized. The first is definitional and
only requires the attribution (denial) of essential human qualities to an outgroup. The second is
conditional on our tendency to attribute human qualities to those we identify with. To the extent
that this is the case, an outgroup may be (de)humanized by the creation (destruction) of a shared
identity and the expansion (contraction) of ingroup boundaries to include (exclude) those
previously identified with an outgroup (ingroup).
The form (de)humanization takes can also vary (see Haslam and Loughnan 2014; Kteily
and Landry 2022). For simplicity’s sake, we use “dehumanization” to refer to the act of attributing
negative qualities to an outgroup. Defined in this way, dehumanization can be seen as evoking a
sense of hatred or disdain for other people. “Humanization” then refers to the act of attributing
positive qualities to an outgroup and evokes a sense of sympathy or respect. We use the term
“infrahumanization” to refer to the act of not attributing positive or negative qualities to an
outgroup.1 Rather than hatred or sympathy, infrahumanization evokes a sense of objective
indifference.
1
The term infrahumanization comes from Leyens et al. (2000) to denote a less extreme form of dehumanization. The
way we use the term is also sometimes referred to as “moral exclusion” (Opotow 1990), “delegitimization” (Bar-Tal
1989), or “mechanistic dehumanization” (Haslam 2006). The way we use dehumanization is sometimes referred to as
“animalistic dehumanization” (Capozza et al. 2012; Goff et al. 2008) or “metaphor-based dehumanization” (Loughnan
et al. 2009).

The extent to which an outgroup is (de)humanized will depend in part on the quality and
quantity of contact between groups (intergroup contact).2 Positive interaction tends to have a
humanizing effect while negative interaction often results in dehumanization (Copozza et al.
2014). One major arena in which diverse groups are likely to interact is in a market. In fact, because
“most of the gains from trade lie outside homogeneous social groups”, markets actively incentivize
intergroup contact (Leeson 2008: 161). However, the existing literature is limited for investigating
this type of contact in two main ways.
The first is due to the lack of attention given to market interaction compared to other forms
of contact. The intergroup contact literature has explored variation in the mode of contact (e.g.,
face-to-face, extended, imagined, virtual) and group identity (e.g., national, political, racial, ethnic,
sexual, gender, religious, age), but far less attention has been given to the type of contact (e.g.,
cooperative, adversarial, planned, spontaneous, voluntary, forced) or the setting of where contact
takes place (e.g., educational, workplace, marketplace, online, sports, experimental laboratory).3
For example, relatively few studies focus on the effect of “superficial contact” or non-intimate
interaction at everyday places like a supermarket, café, or the workplace, which is a form of contact
more typical of our interactions in a market. And even for those that do, the measure of contact
rarely captures a direct measure of interaction related to exchange.4
The second limitation is due to the reliance on experimental or survey instruments which
are designed for purposes other than investigating interaction in a market. These methods work
well for what they’re designed for but often provide limited coverage, mismeasurement, and
questionable external validity when used for other purposes (see Guala 2005; Cersario 2022; Ponto
2
The original intergroup contact theory proposed by Allport et al. (1954) focused on prejudice, but the theory has
since been applied more generally. Pettigrew and Tropp (2006) and Pettigrew et al. (2011) provide a meta-analysis
and review of the main literature while Paluck et al. (2019) focus on the experimental results. Dovidio et al. (2017)
provide a review of the literature on intergroup contact and bias. For a review of the literature on the relationship
between intergroup contact and (de)humanization, see Copozza et al. (2014).
3
A recent exception in terms of type of contact is Lowe (2021) who compares cooperative and adversarial contact by
randomly assigning Indian men of different castes to heterogeneous- or homogeneous-caste cricket teams. However,
one cannot use his results as indicative of anything related to market interactions without first settling the debate about
whether markets promote cooperation or competition.
4
Thomsen and Rafiqi (2018), for example, use a survey measure (ESS) that allows them to capture both the quality
and quantity of superficial contact. However, the measure they use cannot differentiate between market interaction
and any other type of contact. Similarly, Harris and Valentine (2016) investigate intergroup contact in a workplace
environment but measure contact by the level of diversity amongst colleagues which does not capture interaction in
an act of exchange.

2015).5 For example, several studies on intergroup contact use the European Social Survey (ESS)
to assess anti-foreigner sentiment (Thomsen and Rafiqi 2018; Staerklé and Green 2018; Ziller
2022). The ESS is administered to around twenty European countries every other year and has
been since it started in 2002. However, the ESS’s measure for intergroup contact does not
differentiate between market interaction and general interaction. Further, even if we assume it
perfectly correlates to interaction in a market, there remains a severe coverage issue as the two
questions related to intergroup contact (quantity and quality) were only asked in one survey round
so far (Round 7 in 2014).
Nevertheless, it may be possible to assess the market (de)humanization hypothesis in an
indirect way given what is known about intergroup contact and (de)humanization. If markets
encourage positive interactions, we should expect intergroup contact in a market to have a
humanizing effect. If markets instead encourage negative interaction, we should expect markets to
have a dehumanizing effect. The relevant literature here is on the morality of markets. This
literature is concerned with assessing the overall moral character of markets, including the impact
of market institutions and market exchange on morally relevant outcomes (e.g., income, life
expectancy, inequality), moral behaviors (e.g., cooperation, corruption, charitable giving) and
moral values (e.g., materialism, trust, tolerance). The majority of this literature is philosophical in
nature and dates back to antiquity. The applied literature that attempts to empirically adjudicate
between the different philosophical perspectives is much more recent.
The experimental branch of the applied literature tends to focus on the effect of exchange
and extrinsic incentives on values and on social cooperation. The results on values are mixed. For
example, Falk and Szech (2013) find that the act of exchanging something can fundamentally
change how it is valued. In their study, nearly half of the participants were willing to forego a small
amount of money to prevent an act of harm against an innocent third party (a mouse). Yet only a
quarter were willing to forego the same amount of money to prevent the harm once exchange was
introduced. They conclude “markets erode moral values” (Falk and Szech 2013: 710). However,
by replicating the experiment and separating the treatments, Bartling et al. (2021: 27) suggest that
5
A survey instrument that does not work well for our purpose because it cannot differentiate between market
interaction and general interaction (e.g., ESS) may still be valuable for assessing the effect of any superficial contact
on anti-foreigner sentiment (e.g., Thomsen and Rafiqi 2018). Similarly, a survey instrument that does not work well
for our purpose because it cannot be matched to existing data on dyadic market interaction (e.g., WVS) may still be
valuable for assessing the moral character of market societies (e.g., Storr and Choi 2019).

this observed erosion in moral values is instead “caused by playing repeatedly and not by the
market institution.” Nevertheless, exchanging and pricing a good (commodification) may still alter
how we value it, and this may have moral implications. For example, Gneezy and Rustichini (2000)
find that pricing lateness alters it from being a shameful and guilt-ridden act to one that can be
purchased for convenience.6
The results on social cooperation are also mixed. There is consistent evidence that extrinsic
incentives crowd out prosocial motivations and reduce cooperation (Frey and Oberholzer-Gee
1997; Benabou and Tirole 2003, 2006; Vohs et al. 2006; Bowles and Hwang 2008; Reeson and
Tisdell 2010; Johnsen and Kvaløy 2016). However, there is also consistent evidence that people
rarely adhere to the purely self-interested strategy. Some level of altruistic and cooperative
behavior remains even with extrinsic rewards incentivizing against it (Roth et al., 1991; Fehr and
Gächter 2000; Gintis et al. 2003). Still, raise the stakes high enough and people will fail to meet
even their own standards of morality (see Schier et al. 2016).
Other experimental studies show exposure to and experience in markets can increase
prosocial behavior. Henrich et al. (2001), for example, find that market integration (“How much
do people rely on market exchange in their daily lives”) is associated with a greater level of
cooperation. Agneman and Chevrot-Bianco (forthcoming) also find a positive relationship
between market participation and cooperative behavior towards strangers. Even just being exposed
to words related to markets and trade can increase trust and trusting behavior (Al-Ubaydli et al.
2013). Whether these market experiences are viewed positively or negatively may also matter.
Matching closely to what is found in the intergroup contact literature, Choi and Storr (2020a) find
positive past experiences of market interaction can increase trust and reciprocity, but negative past
experiences do not. However, this effect is mitigated by the (im)personal nature of the exchange
and was not found in the impersonal exchange treatment.
The empirical branch of this literature has a harder time isolating the institutional features
of market exchange (e.g., pricing) and instead tends to focus on a bundled set of market
6
While not experimental in nature, Viviana Zelizer provides fascinating historical and sociological analysis on the
effects of pricing life insurance (Zelizer 1979), children (Zelizer 1994), and intimacy (Zelizer 2007). Zelizer’s (2010)
most recent book provides numerous other examples.

institutions.7 The central finding in this literature is that market societies are more moral societies,
or at least not any less moral than non-market societies. Storr and Choi (2019; see also Choi and
Storr 2020b) provide the broadest findings in this regard. Market societies perform better than non-
market societies on a range of different empirical measures with moral relevance, including
measures of values (e.g., less discrimination), behavior (e.g., less bribery), and material outcomes
(e.g., less income inequality).8 These results, they suggest, when “taken in whole, raise doubts that
markets are morally corrupting and, in fact, suggest that markets promote morality” (Storr and
Choi 2019: 274). Callais et al. (forthcoming) use similar (but fewer) morality measures with
matching methods and find no evidence that markets cause moral deterioration. People in market
societies also tend to give more to charity (Cai et al. 2022) and are less materialistic (Teague et al.
2020).
Like the intergroup contact literature, data in this literature are also limited for testing the
market (de)humanization hypothesis. For example, of the many variables Storr and Choi (2019)
consider, thirty-three measure moral values (as opposed to behaviors or outcomes with moral
relevance). However, all but four of these measures come from the World Values Survey (WVS)
which faces similar coverage issues as the ESS.9 Any empirical investigation that relies on the
WVS is limited to a very small panel, a large enough cross-section only in recent waves (the
approach taken by Storr and Choi [2019]), or time-series with a few predictable countries.10
The WVS measures do not work well for our purpose for another, more important reason:
they do not provide enough relational nuance to match moral sentiments to dyadic measures of
market interaction. The relational questions in the WVS (e.g., questions about wanting immigrants,
homosexuals, and people of a different race, language, or religion as neighbors) are about groups
for which direct measures of market interaction at the level needed (country-year) are rarely
available. As a result, the WVS may be used to show that a market society is more trusting or
7
Typically measured by the degree of economic freedom. The Fraser Institute’s Economic Freedom of the World
(EFW) index is standard in this literature. It measures the degree of economic freedom based on five broad areas: (1)
Size of Government, (2) Legal System and Property Rights, (3) Sound Money, (4) Freedom to Trade Internationally,
and (5) Regulation.
8
See Storr and Choi (2019: 260-269) for a complete list of the variables considered.
9
The WVS was first administered in 1981-1984 to only eleven countries. It was not until the sixth wave (2010-2014)
that sixty or more countries were surveyed. The WVS is also inconsistent with which questions are asked to each
country in each wave which further affects coverage.
10
Callais et al. (forthcoming) use matching methods to identify a causal relationship between market institutions,
market reform, and the WVS morality measures. However, they are only able to use seven of the morally relevant
survey questions due to issues of coverage.

accepting of outgroups in general, but this would not preclude the possibility that increasing trade
with an outgroup would increase the extent to which that group is dehumanized.
The existing literature can be used to answer some of the questions on the moral nature of
markets, but not all. How interaction in a market affects the willingness to respect others’ humanity
remains an open question, one that requires different data to answer. Our empirical strategy is
designed with this in mind.
3 Data
3.1 Measuring Moral Sentiment
We leverage a text-as-data approach to create a measure of moral sentiment that can be matched
to dyadic measures of market interaction, allowing us to better test the market (de)humanization
hypothesis.11 Our measure of moral sentiment captures the frequency, valence, and type of moral
language used about a nation in a corpus of New York Times (NYT) articles from 1987-2007. We
use nationality as the criteria for group identity for two reasons. First, nationality is a salient basis
for group identity (Schildkraut 2011) that often acts as a “superordinate identity” by mitigating
existing antagonisms between ethnic, religious, or political groups (Transue 2007; Levendusky
2018). Second, compared to many other identity groups, there are robust measures of market
interaction between nation-states.
We use newspaper articles because they provide an extensive and varied collection of text
compared to other corpora.12 This allows us to capture sentiment data about more countries than
would be possible if we used a less varied corpus. The breadth of topics (e.g., news, sports,
entertainment, financial) also makes it more likely a country is mentioned across multiple contexts
rather than in relation to any one issue. This helps ensure our sentiment measure captures a general
perception of a country rather than an opinion related to a specific issue. If we instead used
congressional speeches, for example, it is likely fewer countries would be discussed and those that
were discussed would be mentioned primarily in relation to a specific policy, including policies
related to our measures for market interaction (e.g., trade or immigration policy). We use the NYT
11
See Gentzkow et al. (2019) for an introduction to text-as-data methods for economic analysis.
12
Text analysis has been used on diverse set of corpora including political texts (Laver et al. 2003; Klemmensen et al.
2007), congressional speeches (Wang and Inbar 2021), newspapers (Kaneko et al. 2020), social media posts (Hutto
and Gilbert 2014; Hoover et al. 2020; Van Vilet 2021), and academic journals (Harris et al. 2022).

as it is among the top three highest-circulating newspapers in the U.S. and features in-depth
coverage of international issues (PEW Research Center 2021). The articles published in the NYT
between 1987 and 2007 represent the largest complete and accessible collection of newspaper
articles digitized for text-analysis.13
We use moral language rather than secondary emotions or other humanized qualities
because we are interested in a particular form of (de)humanization related to the perception of
moral agency. This form of (de)humanization closely relates to the moral concerns over markets,
even if other essential human qualities also matter for a holistic conception of humanity. The moral
language we use is based on the Moral Foundations Theory (MFT). MFT is a pluralist theory of
morality that identifies five foundational moral values related to care/harm, fairness/cheating,
loyalty/betrayal, authority/subversion, and sanctity/degradation (Haidt and Joseph 2004; Haidt
2007; Graham et al. 2009; Graham et al. 2013). The expression of these values and the degree of
their importance varies, but each is consistently found across cultures (Doğruyol et al. 2019). Using
a plurality of moral values for our investigation is important as looking at only one value (e.g.,
fairness) would bias results if what people really care about in a market is another (e.g., loyalty).
These five foundational concerns are further grouped into “binding” and “individualizing”
values. Binding values are defined in relation to a group identity and include loyalty/betrayal,
authority/subversion, and sanctity/degradation. Individualizing values are considered universal
because they are defined outside of a group context and include the values of care/harm and
fairness/cheating (Graham et al. 2009; Graham et al. 2013). Binding values tend to express a
greater concern for ingroup members, while individualizing values instead express a general
concern for others regardless of group identity (Nilsson et al. 2016, Van Leeuwen et al. 2014).
Borrowing from the social capital literature (e.g., Putnam 2000), we refer to language that
expresses binding values as bonding language and language that expresses individualizing values
as bridging language.
13
Stanford University only recently released full-text data for forty-one years (1980-2021) of New York Times articles
and forty-four years (1977-2021) of Washington Post articles. We did not have access to these data at the time of
writing.
10

The dictionary we use is the Extended Moral Foundations Dictionary (eMFD) produced by
Hopp et al. (2021).14 The eMFD classifies moral language along two dimensions: foundation and
valence. Valence denotes the positive-negative use of a word which is associated with virtues and
vices, respectively. Vice language (negative valence) includes words like “fearful”, “retribution”,
“betrayal”, “discrimination”, “flawed”, “cruel”, “hostility” and “rage”. Virtuous language
(positive valence) includes words like “loyal”, “celebrated”, “compassion”, “respect”, “fair”,
“deserve”, “credibility”, and “duty”. Figure 1 illustrates a subset of the eMFD broken down by
foundation and valence.15
To create our sentiment measure, we begin with the full corpus of NYT articles. This
corpus contains over 2.6 million articles with a cumulative sum of 1.14 billion words. We tag each
article according to which of 179 countries the article mentions using three forms of country
references: country name (e.g., “Poland”), country adjectival (e.g., “Polish”), and country
demonyms (e.g., “Poles”).16 This classification strategy allows us to capture most of the discourse
related to a country and its people in our data. Table A1 presents a complete list of reference
countries.
We compute the relative frequency of moral terms in each article by their foundation-
valence classification using Linguistic Inquiry and Word Count (LIWC) (see Tausczik and
Pennebaker 2010). To construct a panel dataset, we take the average relative frequency within a
tagged country and year. More formally, for article i in year t referencing country c and relative
frequency 𝜙, we define moral sentiment μ as:
The raw measure is then multiplied by 1,000,000 to make the coefficients visible in the regressions.
14
The eMFD builds on the Moral Foundations Dictionary (MFD) created by Graham and Haidt (2009) and provides
a variety of benefits compared to the original. The original dictionary was chosen a priori by a few people and contains
relatively few terms per foundation, many of which overlap (e.g., the care-vice category contains five variants of “kill”
which constitutes one-seventh of the terms in that category). In contrast, the eMFD was constructed by hundreds of
human coders annotating multiple corpora for moral content, including the NYT. It contains 3,270 moral words which
is more than ten times the number of words in the original dictionary.
15
The eMFD assigns foundation-specific probabilities to each word in the dictionary. However, for ease of
interpretation, we assign each word to a specific foundation according to its highest foundation probability.
16
An individual article may be tagged as referencing multiple countries. Approximately 30 percent of all articles
reference one or more countries in our sample.
11

3.2 Moral Language and (De)Humanization
Our measure of moral sentiment allows us to identify variation in the use of (1) total moral
language, (2) virtuous language, (3) vice language, (4) bonding language, and (5) bridging
language. The first three can be mapped onto the three different forms (de)humanization.17
The total use of moral language can be used to distinguish between infrahumanization and
(de)humanization. Using less moral language to discuss an outgroup indicates a view of relative
indifference or amorality and identifies infrahumanization. Using more moral language instead
indicates the moralization of others either to express disdain or admiration. To distinguish between
these, we can look at the valence of the language used. Using more virtuous language identifies
humanization while using more vice language identifies dehumanization.
Bonding and bridging language can be mapped onto the different ways to (de)humanize an
outgroup. Using more bonding language identifies (de)humanization by way of changing
ingroup/outgroup identity boundaries. Using more bridging language identifies (de)humanization
by way of changing how an outgroup is perceived. However, the valence is still needed to
determine if bonding or bridging language is used to dehumanize or humanize. There are four
possibilities: (1) dehumanizing bonding language (binding-vice), (2) dehumanizing bridging
language (individualizing-vice), (3) humanizing bonding language (binding-virtue), and (4)
humanizing bridging language (individualizing-virtue).
3.3 Measuring Market Interaction

In comparison to measuring moral sentiment, measuring market interaction is relatively
straightforward.18 We consider two dyadic measures of market interaction. The most direct
measure we use is the bilateral trade flow between the United States and each of the 179 reference
countries. Trade measures the sum of imports and exports in inflation-adjusted billions.19 As a
secondary measure of market interaction, we collect immigration statistics from the U.S.
17
Measuring (de)humanization by language use is common. The concept of infrahumanization, for example,
originated from having participants identify which secondary emotion terms they would attribute to members of their
ingroup and outgroup (Leyens et al. 2000; see also Leyens et al. 2007). The use of text analysis to measure
(de)humanization is more recent, but also common (see Stewart et al. 2011; Wojcieszak and Azrout 2016; Choi et al.
2020; Ousidhoum et al. 2019; Gallacher et al. 2021; Landry et al. 2021).
18
We cannot use economic freedom for this purpose as our measure of market interaction must be dyadic. We instead
use economic freedom as a control (see Table A5). However, it is not included in our main models (Tables 4-7) as it
reduces our sample size by two-thirds and remains insignificant across all specifications.
19
Our results also hold for imports and exports separately.
12

Immigration and Naturalization Service. Immigration measures the count of immigrants into the
U.S. by country of origin. Data is complete for 177 countries. Immigration is a less direct measure
of market interaction as it does not capture an explicit relationship of exchange. Immigration is
partially governed by the supply and demand conditions in the labor market and thus the principles
of exchange; however, Immigration may be better seen as a proxy for the quantity of intergroup
contact in a market setting.
3.4 Controls
While nationality can act as a superordinate identity, there are other identity categories that may
influence our perception of a nation. Religion, for example, may supersede nationality in some
contexts. As such, interaction with one country that is predominately e.g. protestant may influence
our perception of other protestant countries even if we rarely interact with these latter countries.
The same may be true for identities based on political, cultural, or ideological values, ethnicity, or
socioeconomic status (Ben-Ner et al. 2009). To account for this, we control for salient group
identities outside of nationality (e.g., Protestant, Islamic, Polity Score, GDP Per Capita, Years of
Schooling, Ethnic Fractionalization, FE Labor Part Rate [proxying for gender rights]).
Additionally, the content of an article likely affects the language used independent of how
a country is viewed. Stories of political upheaval, war, and violence are likely covered in negative
moral terms regardless of who is believed to be at moral fault. Similarly, oppressive regimes may
be discussed negatively even if the people are viewed favorably. The general material conditions
of a country may also be talked about in moral terms even if blame for these conditions cannot be
placed on the nation. We also control for variables likely to influence the content of an article a
country is mentioned in (e.g., MEPV, Polity Score, Gini, GDP Per Capita).
Lastly, because markets are not the only way we interact with people in different countries,
we control for a major alternative avenue of interaction (Internet Use).20 Table 1 provides brief
descriptions and data sources. Table 2 provides summary statistics.
3.5 Two-Way Fixed Effects
20
Geographical distance is another major avenue of interaction that is controlled for by fixed effects.
13

Both our sentiment and market interaction measures are relational and vary between pairs across
time. For this reason, we employ a country-level panel design with two-way fixed effects to
account for any time and country invariant unobservables likely to influence moral sentiment.
More specifically, we model:
where Yct is a measure of moral sentiment for country c in year t, 𝜏ct is a measure of market
interaction, Xit is a vector of controls, and 𝛼c and 𝛿t are country and year fixed effects, respectively.
In the next section, we explore the philosophical literature on the nature of markets to
identify the contrasting hypotheses relating markets and (de)humanization before moving on to
our empirical tests.
4 The (De)Humanizing Effect of Market Interaction

There are three main hypotheses on the nature of markets that identify plausible mechanisms by
which markets cause (de)humanization.
The Dehumanization Hypothesis: market interaction will result in dehumanization because

markets encourage us to benefit ourselves at the expense of others.
The Infrahumanization Hypothesis: market interaction will result in infrahumanization because

markets obfuscate a person’s true value.
The Humanization Hypothesis: market interaction will result in humanization because markets
encourage us to be other regarding.
We explore each below.
4.1 The Dehumanization Hypothesis

The argument that markets cause dehumanization centers on greed. Greed transforms cooperative
relationships into adversarial ones and encourages us to benefit ourselves at the expense of others,
including by denying others’ humanity if doing so is sufficiently beneficial. In this sense,
dehumanization in a market may result for instrumental reasons as it lowers the cost of acting
14

poorly towards others by making our harmful actions feel less morally repugnant.21 A less
innocuous version of this argument instead sees markets as creating a genuine hatred of other
people.
Either version sees markets as causing or leveraging greed which in turn motivates gain
without concern for others.22 Aristotle ([350 BC] 1999: 17), for example, considered retail trade
to be an unnatural “mode [of wealth seeking] by which men gain from one another.” Some trade
is necessary, but when exchange is transformed from a means of survival to an “art of producing
wealth”, we are encouraged to “increase [our] hoard of coin without limit” and come to view “all
things” in terms of how they can be used to promote this end. The result is that we reduce “every
quality or art into a means of getting wealth” and use “every faculty in a manner contrary to nature”
(Aristotle [350 BC] 1999: 15-16, emphasis added).
G.A. Cohen (2009: 39-40) provides a similar argument. He argues that by replacing
genuine other-regarding motivations with “cash rewards”, markets “habituated and inured” in us
a tendency to view others as either “sources of enrichment” or “threats to [our] success.” As a
result, we do “not care fundamentally, within market interaction, about how well or badly anyone
other than [ourselves] fares” (Cohen 2009: 44-45, original emphasis). Leveraging greed in this
way can generate beneficial outcomes and even encourage surface-level cooperation. However, as
Cohen (2009: 77) suggests, “we should never forget that greed and fear are repugnant motives.”
Because behavior in a market is only valued instrumentally, there is nothing preventing us from
dehumanizing others if it were in our interest to do so. Neither Cohen nor Aristotle see
dehumanization (attribution of negative qualities) as an inherent feature of market interaction.23
21
See Bandura et al. (1996) for a discussion on how dehumanization and other “mechanisms of moral disengagement”
allow us to convert “harmful acts to moral ones”.
22
There is not consistent agreement if markets cause or only leverage greed. Aristotle ([350 BC] 1999: 16) suggests
the “disposition in men” to “increase their money without limit” originates from the fact “their desires are unlimited”.
Similarly, Cohen (2009: 41) states, “Capitalism did not, of course, invent greed and fear; they are deep in human
nature. But… capitalism celebrates it.” Rosseau ([1754] 2006: 66, 50) instead views man as “naturally good”. Our
“[i]nsatiable ambition” and “thirst of raising [our] respective fortunes” is instead one of the first evils caused by the
creation of a property-based market society.
23
This claim is conditional on how we define dehumanization. Viewing people only in terms of their instrumental
value is a form of dehumanization that is closer to how we define infrahumanization as it does not require the
attribution of negative qualities. However, we do not associate infrahumanization with Aristotle or Cohen because we
believe there is a fundamental difference between acknowledging a person’s humanity and ignoring or denying it for
gain (‘instrumental dehumanization’) and not seeing a person as a person in the first place (‘infrahumanization’). It is
the difference between Cohen’s claim that markets will cause us to not care about other people and Marx’s claim that
markets will cause us to lose our sense of what it even means to be human.
15

Dehumanization is instead a plausible byproduct of greed as we come to care more for ourselves
than for others.
Jean-Jacques Rousseau, on the other hand, blames property and exchange directly for
corrupting our good nature and causing dehumanization.24 Rousseau ([1754] 2006: 66) viewed
market exchange as zero-sum “in which every man finds profit in the misfortunes of his
neighbour”. And since “we always gain more by hurting our neighbours than by doing them good”,
markets require we suppress our natural compassion and act with cruelty as they put us “under a
kind of necessity to ill-use all the persons of whom [we stand] in need”. Markets create “rivalry
and competition on the one hand, and conflicting interests on the other, together with a secret
desire on both of profiting at the expense of others.” As such, a market society “necessarily leads
men to hate each other in proportion as their interests clash” (Rousseau [1754] 2006: 67, 50, 66).
4.2 The Infrahumanization Hypothesis

One way to treat people as less than human is to actively deny or willfully ignore their status as
moral equals. Another way is to fail to recognize their moral status in the first place. We may
mistreat someone in a market not because of greed or malice, but because their true value was
obfuscated in some way. The argument that markets cause infrahumanization does not necessarily
deny the deleterious effects of greed, but the focus is instead on improper valuation resulting from
commodification and exchange.
Consider Marx’s ([1844] 2000: 28) argument on the connection between market
competition, commodification, and “the devaluation of man”.25 According to Marx, market
competition results in “the accumulation of capital in a few hands, and thus the restoration of
monopoly in a more terrible form”. The disparity between those with and those without property
transforms a worker into “the most wretched of commodities” who is at “the whim of the rich and
the capitalists” and “must sell himself and his humanity” to survive (Marx [1844] 2000: 28, 3-5).
Work in a market is “therefore not voluntary, but coerced; it is forced labor”. Being forced to labor
alienates a worker from what they produce and from the act of production itself. This alienation in
24
On the connection between property rights and markets, see Harris et al. (2020).
25
Modern arguments against commodification focus on the expansion of market principles to other social, political,
and physical spaces “where they don’t belong” (Sandel 2012: 3; see Lefebvre 2002; Gudeman 2016; Radin 1987,
1989; Anderson 1993; Sandel 2012; Zelizer 1979, 1994, 2007, 2010). We feature Marx’s argument here as his focus
is on the commodification of people (labor) rather than the commodification of other goods which would also be
improperly valued if exchanged in a market.
16

turn transforms “the very act of production” into an activity whereby a worker begins the process
of “estranging himself from himself” (Marx [1844] 2000: 30). In this sense, market “[p]roduction
does not simply produce man as a commodity… it produces him… as a mentally and physically
dehumanized being”, estranged from himself and even from “his human aspect” (Marx ([1844]
2000: 36, 32, original emphasis).
“An immediate consequence” of this “self-estrangement” is “the estrangement of man from
man” as when “man confronts himself, he confronts the other” (Marx ([1844] 2000: 31-32, original
emphasis). Infrahumanization is thus a natural result of the estrangement inherent in a market that
is “realized and expressed only in the relationship in which a man stands to other men” (Marx
[1844] 2000: 32). From this perspective, interaction in a market is strange and isolating, not
adversarial and tumultuous. The same market forces that cause us to fail to recognize our own
humanity cause us to fail to recognize the humanity in others too.
4.3 The Humanization Hypothesis

The argument that markets cause humanization centers on the idea that markets require other-
regarding behavior, regardless of how it is motivated.26 In this view, markets reward virtue, not
vice.
This position is commonly associated with the doux commerce theory attributed to
Montesquieu: “the natural effect of commerce is to lead to peace. Two nations that trade together
become mutually dependent… and all unions are based on mutual needs” (quoted in Hirschman
1977: 80). Markets temper the worst in us by creating a union of mutual accommodation. We rely
on others for our benefit just as they rely on us for theirs, and so we act accordingly. Our daily
bread may not come from the beneficence of others, but the baker nevertheless treats us
beneficently and expects the same in return. Even with purely self-interested motives, markets
encourage us to act in the public interest (Smith [1776] 1977), and transform our private vices
(e.g., greed) into public virtues (Mandeville [1714, 1732] 1988). The other-regarding behavior we
see in a market may not be motivated by genuine care, but that does not matter for the humanizing
effect of markets. Because market exchange relies on mutual dependency, success in a market
requires other virtues like temperance and prudence. And it is through these virtues that we come
26
Lavoie and Chamlee-Wright (2002) refer to the instrumental version of this argument as a “minimalist” defense of
the morality of markets.
17

to treat others with respect and dignity, even if we do not do so out of love or care (McCloskey
2006, 2010, 2016).27
Of course, we may also be genuine in our other-regarding behavior even if we did not set
out to be. Most of our interactions in a market appear superficial, but they do not stay that way for
long as “in the real world, repeated dealings, not one-shot games, predominate” (Storr 2010: 203).
A better conception of a market, Storr (2008) argues, is as a social space that habituates sociality
rather than animosity or indifference. Viewed in this way, markets are the “buttresses” of
community that “encourag[e] the development of social bonds” and “creat[e] favorable conditions
for feelings of” mutual sympathy (Storr 2009: 279-280). And it is through mutual sympathy that
we come to identify with and, in time, genuinely care for other people. As Adam Smith ([1759]
1982: 328) suggests, “the necessity or conveniency of mutual accommodation, very frequently
produces a friendship not unlike that which takes place among those who are born to live in the
same family. Colleagues in office, partners in trade, call one another brothers; and frequently feel
towards one another as if they really were so.” Other-regarding behavior may originate out of self-
interest, but it is through our repeated interactions in a market that we come to identify with,
genuinely care for, and humanize other people.
5 Testing the (De)Humanization Hypotheses

The three hypotheses outlined above can be combined with our sentiment measure to generate
empirical predictions as to the nature of markets.
5.1 Infrahumanization and (De)Humanization

The Infrahumanization Hypothesis suggests that markets will cause us to view other people with
objective indifference as we come to view them as mere commodities. As such, we will use less
moral language about the groups we interact with more in a market. A negative relationship
between market interaction and the overall use of moral language would evidence this hypothesis.
A positive relationship would instead show that market interaction is associated with a greater
27
McCloskey argues that markets also nurture the virtues of courage, justice, love, faith, and hope. Her work thus
makes an even stronger claim about our tendency to humanize those we interact with in a market, and to do so in a
genuine way.
18

moralization of others either in positive or negative terms. In order to tell which is the case, we
would need to look at the valence of the moral language that is used.
The Humanization Hypothesis suggests that markets will cause us to view other people as
equal persons worthy of dignity and respect as we come to rely on, sympathize with, or genuinely
care for them. As such, we will use more virtuous language about the groups we interact with more
in a market. A positive relationship between market interaction and virtuous language would
evidence this hypothesis. A negative relationship would evidence against this hypothesis, but it
alone would not provide evidence for the alternative hypotheses. A reduction in virtuous language
could result from either infrahumanization or dehumanization.
The Dehumanization Hypothesis suggests that markets will cause us to view other people
with disdain or even hatred as we are encouraged by greed to see them as adversaries and use them
instrumentally. As such, we will use more vice language about the groups we interact with more
in a market. A positive relationship between market interaction and vice language would evidence
this hypothesis. For the same reasons as before, a negative relationship would evidence against
this hypothesis but not provide evidence for the alternatives.
5.2 Methods of (De)Humanization

The method of (de)humanization may also be influenced by the nature of markets. Does interacting
in a market change how we perceive an outgroup, or does it change how we perceive identity and
the boundaries of our ingroup? To distinguish between the different methods of (de)humanization,
we can examine the use of moral language by valence and foundation.
Dehumanizing bonding language can sever identity connections by contrasting group
values and can be used to strengthen ingroup cohesion (Wenzel et al. 2010). In this sense, a positive
relationship between market interaction and dehumanizing bonding language may evidence an
instrumental form of dehumanization associated with Aristotle and Cohen whereby markets
encourage us to dehumanize others for the benefit of our ingroup. A positive relationship between
market interaction and dehumanizing bridging language would instead evidence a more genuine
form of dehumanization associated with Rousseau whereby markets create animosity between
groups and cause us to view others as adversaries.
Rather than severing social ties, markets may encourage us to develop social bonds through
the creation and recognition of shared values and identity. A positive relationship between market
19

interaction and humanizing bonding language may evidence a form of humanization associated
with Storr and Smith that identifies a genuine care for others.
Humanizing bridging language is more difficult to associate with a specific philosophical
perspective as the values that constitutes the category, fairness and care, relate to the concept of
humanization itself. Does humanization require a genuine care for others or are instrumental
motives sufficient provided we treat people with dignity and respect (i.e., fairly)? Cohen and
Mandeville may disagree. To account for this disagreement, we can disaggregate the language
category. A positive association between market interaction and care-virtue language would
evidence a more genuine form of humanization. However, if only fairness-virtue language is used,
a positive relationship between markets and humanizing bridging language may instead evidence
an instrumental form of humanization associated with Montesquieu and Mandeville.
5.4 Results
We now move to test these predictions. We begin our analysis with a baseline fixed-effects
regression between our market interaction measures and the sum of all moral language. The results
are presented in Table 3.28 In both models we observe a positive and statistically significant
relationship between market interaction and the use of moral language.
Since these results may be sensitive to a variety of confounding covariates, we re-estimate
the model with the controls discussed in section 3.3. The results are shown in Table 4 and are
robust to a variety of alternative control specifications (see Tables A3 and A4). Market interaction
has a significant positive effect on the total use of moral language, suggesting that market
interaction does not result in infrahumanization. To distinguish between humanization or
dehumanization, we look at the valence of moral language, presented in Table 5. Both Trade and
Immigration show a significant and positive relationship with virtuous language, but not vice
language. Together, these results indicate market interaction is associated with humanization rather
than infrahumanization or dehumanization.
28
Our results are presented with a restricted sample that removes both Mexico and Canada given their outlier nature
for our market interactions measures. Mexico on immigration is 10.6 standard deviations (SDs) above the mean.
Canada on trade is 15.2 SDs above. For comparison, China is 2.3 SDs above the mean on immigration and 4 SDs
above on trade. Table A6 shows the full, unrestricted sample results. Only the results for immigration are sensitive to
the inclusion of Mexico.
20

The question then of how we humanize remains. Do we extend sympathy or identity?
Tables 6 and 7 show the results for bridging and bonding language separated by valence. Trade is
significantly associated with both bonding and bridging humanization. Further, this result is not
solely driven by fairness concerns as Trade is positively and significantly associated with both
fairness and care language (see Table A6). Genuine or not, markets appear to encourage a greater
expression of care for others. However, Trade also shows a significant relationship with
dehumanizing bonding language related to authority/subversion values. In other words, the net
effect of Trade is a humanizing one (Table 4), but on some margins it may encourage
dehumanization for the purpose of ingroup cohesion. Immigration is also significantly associated
with both bonding and bridging humanization and shows no significant relationship to
dehumanizing language for any of the foundational moral values (see A7).
6 Conclusion
The world is becoming more interconnected due in large part to global trade. And while
globalization has generated significant material benefits, there are still major concerns over its
moral costs (Dunning 2003). As more and more of our interactions come to be governed by the
principles of market exchange, might we be losing part of our genuine human connection to others?
This is an empirical question that has largely been left to philosophical conjecture due to a lack of
data and methods to test it.
We provide a measure and method that allows such a test. Our results suggest that market
interaction has a humanizing effect of moral inclusion as we extend our consideration to others
and develop social bonds. As such, our results add to the empirical support for the idea that the
market is a humanizing moral space (Storr and Choi 2019). And while we situate our contribution
squarely in the intersection of the morality of markets and intergroup contact literatures, we
nonetheless think our methods may be useful for both literatures separately and for other questions
entirely.
Our sentiment measure only captures one form of (de)humanization related to the
perception of moral agency. This form of (de)humanization has the most direct connection to the
moral concerns over markets, but (de)humanization based on other criteria may also matter. Future
studies could employ a similar empirical approach but use different dictionaries to better capture
21

other forms of (de)humanization. Schweitzer and Waytz (2021), for example, recently released a
“Mind Perception Dictionary” that may provide a broader conception of (de)humanization related
to the recognition of other minds. Dictionaries related to prejudice, bias, or other implications of
the intergroup contact theory could also be used.
Additionally, our focus on market interaction was deliberate, but restrictive. We did not
consider measures of intergroup contact that had little connection to markets, and our use of
immigration was as a proxy for contact in a market rather than the primary focus. Those interested
in the broader intergroup contact theory may wish to use a similar approach as us but with an
explicit focus on immigration or another measure of contact entirely. Identity criteria beyond
nationality could also be used if there is no longer the need to match sentiment data to the dyadic
measures of market interaction.
Our need to match data also required that we aggregate at the country-year level. This
provides obvious benefits in terms of data availability, but it also removes interesting variation as
some interactions in a market will be positive and others negative. For example, even if the net
effect of trade (or immigration) is positive, some people will nonetheless lose. Disaggregating the
data to a more localized level could help address this issue, but it would require access to localized
opinions (e.g., local newspapers).
Lastly, while we frame and interpret our results using philosophical perspectives that
provide plausible mechanisms by which markets cause (de)humanization, we do not employ an
identification strategy that would allow us to rule out the possibility of reverse causation. Future
studies may wish to do so and will have more options available following the recent release of 40+
years of New York Times and Washington Post articles (see fn. 13).29 Nevertheless, our results at
least modestly suggest the worst of the market critique, that markets act as a dehumanizing force,
may not be the case.
29
As an example, free trade agreements may provide a possible avenue for causal investigation. However, of the
twenty agreements in force, only three (Canada, Mexico, and Jordan) provide more than three years of post-treatment
data given our current observation period (1987-2007). The new Stanford data (1977-2021) would allow all twenty
agreements to be used.
22

References
Agneman, G., Chevrot-Bianco, E. (Forthcoming). “Market Participation and Moral Decision-
Making: Experimental Evidence from Greenland.” The Economic Journal.
Allport, G. W., Clark, K., Pettigrew, T. (1954). The Nature of Prejudice. Reading: Addison-
Wesley Publishing Co.
Al-Ubaydli, O., Houser, D., Nye, J., Paganelli, M.P., Pan, X.S. (2013) “The Causal Effect of
Market Priming on Trust: An Experimental Investigation Using Randomized
Control.” PLoS ONE, 8(3): e55968.
Anderson, E. (1993). Value in Ethics and Economics. Cambridge: Harvard University Press.
Apergis, N., Cooray, A. (2015). “Economic Freedom and Income Inequality: Evidence from a
Panel of Global Economies- a Linear and a Non-Linear Long-Run Analysis.” The
Manchester School 85(1): 88–105.
Aristotle. ([350 BC] 1999). Politics. Translated by Benjamin Jowett. Kitchener: Batoche Books.
Bandura, A., Barbaranelli, C., Caprara, G. V., Pastorelli, C. (1996). “Mechanisms of moral
disengagement in the exercise of moral agency.” Journal of personality and social
psychology, 71(2), 364.
Bar-Tal, D. (1989). “Delegitimization: The extreme case of stereotyping and prejudice.” In D.

Bar-Tal, C.F. Grauman, A.W. Kruglankski, W. Stroebe (Eds.), Stereotyping and
Prejudice (pp. 169-182). New York: Springer.
Bartling, B., Fehr, E., Özdemir, Y. (2021). “Does Market Interaction Erode Moral Values?” The
Review of Economics and Statistics 1-32.
Benabou, R., Tirole, J. (2003). “Intrinsic and extrinsic incentives.” Review of Economic Studies,
70:489–520.
Benabou R., Tirole. J. (2006). “Incentives and prosocial behavior.” American Economic Review,
96:1652–1678.
Ben-Ner, A., McCall, B.P., Stephane, M., Wang, H. (2009) “Identity and in-group/out-group
differentiation in work and giving behaviors: Experimental evidence.” Journal of
Behavior and Organization 72(1): 153-170
Bowles, S., Hwang, S. (2008) "Social Preferences and Public Economics: Mechanism design
when social preferences depend on incentives." Economics Department Working Paper
Series, 31.
23

Brennan, J., Jaworski, P.M. (2015). Markets Without Limits: Moral Virtues and Commercial
Interests. New York: Routledge.
Cai, M., Caskey, G. W., Cowen, N., Murtazashvili, I., Murtazashvili, J. B., Salahodjaev, R.
(2022). Individualism, economic freedom, and charitable giving. Journal of Economic
Behavior & Organization, 200, 868-884.
Callais, J. T., Geloso, V. (2021). “Economic Freedom Promotes Upward Income Mobility.”
Economic Freedom of the World 189.
Callais, J. T., Harris, C., Borchard, B. (Forthcoming). “The Moral Costs of Markets: Testing the
Deterioration Hypothesis.” Journal of Behavior and Organization.
Capozza, D., Andrighetto, L., Di Bernardo, G. A., Falvo, R. (2012). “Does status affect
intergroup perceptions of humanity?” Group Processes & Intergroup Relations 15(3):
363-377.
Capozza, D., Falvo, R., Di Bernardo, G. A., Vezzali, L., Visintin, E. P. (2014). “Intergroup
contact as a strategy to improve humanness attributions: A review of studies.” TPM-
Testing, Psychometrics, Methodology in Applied Psychology 21(3): 349-362.
Cesario, J. (2022). “What can experimental studies of bias tell us about real-world group
disparities?” Behavioral and Brain Sciences, 45: E66.
Choi, G., Storr, V. (2020a). “Market Interactions, trust, and reciprocity”, PLoS One 15(5):
e0232704.
Choi, G., Storr, V. (2020b). “Measuring Markets and Morality”. Erasmus Journal for Philosophy
and Economics. 13(1): 42-60.
Choi, S., Liu, J. H., Csertő, I., Vincze, O., Fülöp, É., & Pólya, T. (2020). “Automated analysis of
narrative: NarrCat and the identification of infrahumanization bias within text.” Journal
of Language and Social Psychology, 39(2), 237-259.
Cohen, G. A. (2009). Why not Socialism? Princeton: Princeton University Press.
Corak, M. (2013). “Inequality from Generation to Generation: The United States in

Comparison.” In The Economics of Inequality, Poverty, and Discrimination in the 21st
Century, ed. R. Rycroft. Santa Barbara: ABC-CLIO, LLC.
Doğruyol, B., Alper, S., Yilmaz, O. (2019). “The five-factor model of the moral foundations
theory is stable across WEIRD and non-WEIRD cultures.” Personality and Individual
Differences, 151.
24

Dovidio, J. F., Love, A., Schellhaas, F. M., Hewstone, M. (2017). “Reducing intergroup bias
through intergroup contact: Twenty years of progress and future directions.” Group
Processes & Intergroup Relations, 20(5): 606-620.
Dunning, J. H. (2003). Making Globalisaton Good: The Moral Challenges of Global Capitalism.
New York: Oxford University Press.
Falk, A., Szech, N. (2013). “Morals and Markets.” Science 340(6133): 707–711.
Fehr, E., Gächter, S. (2000). "Fairness and Retaliation: The Economics of Reciprocity." Journal
of Economic Perspectives, 14 (3): 159-181.
Frey, B., Oberholzer-Gee, F. (1997). “The cost of price incentives: An empirical analysis of
motivation crowding-out.” American Economic Review, 87: 746–755.
Frith, U., Frith, C. D. (2003). “Development and neurophysiology of mentalizing”. Philosophical

transactions of the Royal Society of London. Series B, Biological sciences, 358(1431):
459–473.
Gallacher, J. D., Heerdink, M. W., Hewstone, M. (2021). “Online Engagement Between

Opposing Political Protest Groups via Social Media is Linked to Physical Violence of
Offline Encounters.” Social Media + Society, 7(1).
Gehring, K. (2013). “Who Benefits from Economic Freedom? Unraveling the Effect of
Economic Freedom on Subjective Well-Being.” World Development 50, 74–90.
Gentzkow, M., Kelly B., Taddy, M (2019). “Text as Data.” Journal of Economic Literature.
57(3): 535-574.
Gintis, H., Bowles, S., Boyd, R., Fehr, E. (2003). “Explaining altruistic behavior in
humans.” Evolution and human Behavior, 24(3): 153-172.
Gneezy, U., Rustichini, A. (2000). “A Fine is a Price.” The Journal of Legal Studies, 29(1): 1-17.
Goel, R. K., Nelson, M. A. (2005). “Economic freedom versus political freedom: cross‐country
influences on corruption.” Australian Economic Papers, 44(2), 121-133.
Goff, P. A., Eberhardt, J. L., Williams, M. J., Jackson, M. C. (2008). “Not yet human: implicit
knowledge, historical dehumanization, and contemporary consequences.” Journal of
Personality and Social Psychology 94(2): 292-306.
Gopnik, A., Wellman, H. M. (1992). “Why the child's theory of mind really is a theory”. Mind &
Language 7(1-2): 145–171.
25

Graham, J., Haidt, J. (2009). “The Moral Foundations Dictionary.”
https://moralfoundations.org/wp-
content/uploads/files/downloads/moral%20foundations%20dictionary.dic
Graham, J., Haidt, J., Nosek, B. A. (2009). “Liberals and conservatives rely on different sets of
moral foundations.” Journal of personality and social psychology, 96(5), 1029.
Graham, J., Haidt, J., Koleva, S., Motyl, M., Iyer, R., Wojcik, S., Ditto, P. H. (2013) “Moral
Foundations Theory: The Pragmatic Validity of Moral Pluralism.” Advances in
Experimental Social Psychology, 47, 55-130.
Gropper, D. M., Lawson, R. A., Thorne Jr, J. T. (2011). “Economic freedom and happiness.”
Cato Journal., 31, 237.
Guala, F., Mittone, L. (2005). Experiments in Economics: External Validity and the Robustness
of Phenomena. Journal of Economic Methodology, 12(4): 495-515.
Gudeman, S. (2016). Anthropology and Economy. Cambridge: Cambridge University Press.
Gudykunst, W. B. (1985). “A Model of Uncertainty Reduction in Intercultural

Encounters.” Journal of Language and Social Psychology, 4(2), 79–98.
Gwartney, J. D., Lawson, R. A., Holcombe, R. G. (1999). “Economic Freedom and the
Environment for Economic Growth.” Journal of Institutional and Theoretical Economic
155(4), 643–663.
Haidt, J. (2007). “The New Synthesis in Moral Psychology.” Science, 316(5827), 998-1002.
Haidt, J., Joseph, C. (2004). “Intuitive ethics: How innately prepared intuitions generate
culturally variable virtues.” Daedalus, 133(4), 55-66.
Harris, C., Cai, M., Murtazashvili, I., Murtazashvili, J. B. (2020). The Origins and Consequences
of Property Rights: Austrian, Public Choice, and Institutional Economics Perspectives.
Cambridge: Cambridge University Press.
Harris, C., Myers A., Carlen, S. Briol, C. (2022). “The Binding Force of Economics.” Advances
in Austrian Economics 26: 69-103.
Harris, C., Valentine, G. (2016). “Encountering difference in the workplace: Superficial contact,
underlying tensions and group rights.” Tijdschrift voor economische en sociale
geografie 107(5): 582-595.
Haslam, N. (2006). “Dehumanization: An integrative review.” Personality and Social

Psychology Review 10(3): 252-264.
26

Haslam, N., Loughnan, S. (2014). “Dehumanization and infrahumanization.” Annual Review of
Psychology, 65:399-423.
Haslam, N., Loughnan, S., Kashima, Y., & Bain, P. (2008). “Attributing and denying humanness
to others.” European Review of Social Psychology 19(1): 55-85.
Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., (2001), “In search of homo-
economicus: Behavioral experiments in 15 small-scale societies.” American Economic
Review, 91:73–78.
Hirschman, A. O. (1977). The Passions and the Interests: Political Arguments for Capitalism
Before its Triumph. Princeton: Princeton University Press.
Hoover, J., Portillo-Wightman, G., Yeh, L., Havaldar, S., Davani, A. M., Lin, Y., Kennedy, B.,
Atari, M., Kamel, Z., Mendlen, M., Moreno, G., Park, C., Chang, T. E., Chin, J., Leong,
C., Leung, J. Y., Mirinjian, A., & Dehghani, M. (2020). “Moral foundations twitter
corpus: A collection of 35k tweets annotated for moral sentiment.” Social Psychological
and Personality Science, 11(8), 1057-1071.
Hopp, F. R., Fisher, J. T., Cornell, D., Huskey, R., Weber, R. (2021). “The Extended Moral
Foundations Dictionary (eMFD): Development and Applications of a Crowd-Sourced
Approach to Extracting Moral Intuitions from Text.” Behavior Research Methods, 53(1),
232–246.
Hutto, C., Gilbert, E. (2014). “Vader: A parsimonious rule-based model for sentiment analysis of
social media text.” In Proceedings of the International AAAI Conference on Web and
Social Media, 8(1).
Izard, C. (1960). “Personality similarity and friendship”. The Journal of Abnormal and Social
Psychology, 61(1), 47–51
Izuma, K., & Adolphs, R. (2013). “Social manipulation of preference in the human
brain.” Neuron, 78(3), 563-573.
Johnsen, A. A., Kvaløy, O. (2016). “Does Strategic Kindness Crowd Out Prosocial Behavior?”
Journal of Economic Behavior & Organization, 132(Part A): 1-11.
Kaneko, T., Asano, T., Miwa H. (2020). “Estimating Ideal points of Newspapers from Editorial
Texts.” The International Journal of Press/Politics, 26(3): 719-742.
Kinzler, K. D., Shutts, K., Dejesus, J., Spelke, E. S. (2009). “Accent trumps race in guiding
children's social preferences.” Social cognition, 27(4):623–634.
Klemmensen, R., Hobolt, S. B., Hansen, M. E. (2007). “Estimating Policy Positions Using
Political Texts: An Evaluation of the Wordscores Approach”. Electoral Studies 26(4):
746-755.
27

Kteily, N.S., & Landry, A.P. (2022). “Dehumanization: Trends, insights, and challenges.” Trends
in Cognitive Sciences, 26(3), 222–240.
Landry, A. P., Ihm, E., Kwit, S., & Schooler, J. W. (2021). “Metadehumanization erodes
democratic norms during the 2020 presidential election.” Analyses of Social Issues and
Public Policy, 21(1), 51-63.
Laver, M., Benoit, K., Garry, J. (2003). “Extracting policy positions from political texts using
words as data.” American Political Science Review, 97(2), 311-331.
Lavoie, D., Chamlee-Wright, E. (2000). Culture and Enterprise: The Development,

Representation and Morality of Business. London/New York: Routledge.
Leeson, P. T. (2008). “Social Distance and Self‐Enforcing Exchange.” The Journal of Legal
Studies, 37(1):161–188.
Lefebvre, H. (2002). Critique of Everyday Life. Translated by John Moore. London: Verso.
Levendusky, M. (2018). “Americans, Not Partisans: Can Priming American National Identity
Reduce Affective Polarization?” The Journal of Politics, 80(1).
Leyens, J. P., Paladino, P. M., Rodriguez-Torres, R., Vaes, J., et al. (2000). “The Emotional Side
of Prejudice: The Attribution of Secondary Emotions to Ingroups and Outgroups.”
Personality and Social Psychology Review, 4(2):186–197.
Leyens, J. P., Demoulin, S., Vaes, J., Gaunt, R., et al. (2007). “Infra‐humanization: The wall of
group differences.” Social Issues and Policy Review, 1(1):139–172.
Loughnan, S., Haslam, N., Kashima, Y. (2009). “Understanding the relationship between
attribute-based and metaphor-based dehumanization.” Group Processes & Intergroup
Relations 12(6): 747-762.
Lowe, M. (2021). “Types of contact: A field experiment on collaborative and adversarial caste
integration.” American Economic Review 111(6): 1807-1844.
Mandeville, B. ([1714, 1732] 1988.) The Fable of the Bees or Private Vices, Publick Benefits.
With Commentary by F.B. Kaye. Indianapolis: Liberty Fund.
Marx, K. ([1844] 2000). Economic & Philosophic Manuscripts of 1844. Marxists Internet
Archive. Retrieved November 8, 2021, from
https://www.marxists.org/archive/marx/works/download/pdf/Economic-Philosophic-
Manuscripts-1844.pdf
McCloskey, D.N. (2006). The Bourgeois Virtues: Ethics for an Age of Commerce. Chicago:
University of Chicago Press.
28

McCloskey, D.N. (2010). Bourgeois Dignity: Why Economics Can't Explain the Modern World.
Chicago: University of Chicago Press.
McCloskey, D.N. (2016). Bourgeois Equality: How Ideas, Not Capital or Institutions, Enriched
the World. Chicago: University of Chicago Press.
Nilsson, A., Erlandsson, A., Västfjäll, D. (2016). “The congruency between moral foundations
and intentions to donate, self-reported donations, and actual donations to charity.”
Journal of Research in Personality, 65, 22-29.
Opotow, S. (1990). “Moral exclusion and injustice: An introduction.” Journal of Social Issues,
46(1): 1-20.
Ousidhoum, N., Lin, Z., Zhang, H., Song, Y., Yeung, D. Y. (2019). “Multilingual and multi-
aspect hate speech analysis.” arXiv preprint arXiv:1908.11049.
Paluck, E. L., Green, S. A., Green, D. P. (2019). “The contact hypothesis re-
evaluated.” Behavioural Public Policy 3(2): 129-158.
Pettigrew, T. F., Tropp, L. R. (2006). “A meta-analytic test of intergroup contact

theory.” Journal of Personality and Social Psychology 90(5): 751-783.
Pettigrew, T. F., Tropp, L. R., Wagner, U., Christ, O. (2011). “Recent advances in intergroup
contact theory.” International Journal of Intercultural Relations 35(3): 271-280.
Ponto, J. (2015). “Understanding and Evaluating Survey Research.” Journal of the Advanced
Practitioner in Oncology, 6(2):168-71.
Putnam, R. D. (2000). Bowling alone: The collapse and revival of American community. New
York: Simon and Schuster.
Radin M.J. (1987). “Market-inalienability.” Harvard Law Review 100(8): 1849–1937.
Radin M.J. (1989) “Justice and the market domain.” Nomos, 31, In: Chapman J.W. and Pennock
J.R. (eds.), Markets and Justice. New York: New York University Press, pp. 165–197.
Reeson, A., Tisdell, J. (2010). “The market instinct: The demise of social preferences for self-
Interest.” Environmental and Resource Economics, 47: 439–453.
Roberts, J. M., Olson, R. (2013). How Economic Freedom Promotes Better Health Care,
Education, and Environmental Quality. Washington, DC: The Heritage Foundation.
Roth, A. E., Prasnikar, V., Okuno-Fujiwara, M., Zamir, S. (1991). “Bargaining and market
behavior in Jerusalem, Ljubljana, Pittsburgh, and Tokyo: An experimental study.” The
American Economic Review 81(5): 1068-1095.
29

Rousseau, J. J. ([1754] 2006). Discourse on Inequality. Overland Park: Digireads Publishing
Sandel, M.J. (2012). What Money can’t Buy: The Moral Limits of Markets. Basingstoke:
Macmillan.
Schier, U. K., Ockenfels, A., Hofmann, W. (2016). “Moral values and increasing stakes in a
dictator game.” Journal of Economic Psychology, 56:107–115.
Schildkraut, D. J. (2011). National identity in the United States. In Handbook of identity theory
and research (pp. 845-865). Springer, New York, NY.
Schweitzer, S., Waytz, A. (2021). “Language as a window into mind perception: How mental
state language differentiates body and mind, human and nonhuman, and the self from
others.” Journal of Experimental Psychology 150(8): 1642-1672.
Smith, A. ([1759] 1982.) The Theory of Moral Sentiments. Indianapolis: Liberty Fund.
Smith, A. ([1776] 1977.). An Inquiry into the Nature and Causes of the Wealth of Nations.
Chicago: Chicago University Press.
Staerklé, C., Green, E. G. (2018). “Right‐wing populism as a social representation: A

comparison across four European countries.” Journal of Community & Applied Social
Psychology 28(6): 430-445.
Stewart, C. O., Pitts, M. J., & Osborne, H. (2011). “Mediated intergroup conflict: The discursive
construction of “illegal immigrants” in a regional US newspaper.” Journal of language
and social psychology, 30(1), 8-27.
Storr, V.H. (2008). “The market as a social space: On the meaningful extraeconomic
conversations that can occur in markets.” Review of Austrian Economics 21: 135-150.
Storr, V.H. (2009). “Why the market? Markets as social and moral spaces.” Journal of Markets
& Morality 12(2): 277-296.
Storr, V.H. (2010). “The Social Construction of the Market.” Society 47: 200-206.
Storr, V.H., Choi, G.S. (2019). Do Markets Corrupt Our Morals? London: Palgrave Macmillan.
Stroup, M. D. (2007). “Economic Freedom, Democracy, and the Quality of Life.” World
Development 35(1): 52-66.
Taft, M., Kacanas, D., Huen, W., Chan, R. (2011). “An empirical demonstration of contrastive
rhetoric: Preference for rhetorical structure depends on one's first language.” Intercultural
Pragmatics 8(4): 503-516.
30

Tausczik, Y., Pennebaker, J. (2010). “The Psychological Meaning of Words: LIWC and
Computerized Text Analysis Methods,” Journal of Language and Social Psychology,”
29(1): 24-54.
Teague, M.V., Storr, V.H., Fike, R. (2020). “Economic Freedom and Materialism: An Empirical
Analysis.” Constitutional Political Economy, 31: 1-44.
The Pew Research Center (2021). “The Newspapers Fact Sheet.”

https://www.pewresearch.org/journalism/fact-sheet/newspapers/
Thomsen, J. P. F., Rafiqi, A. (2018). “When does superficial intergroup contact reduce anti-
foreigner sentiment? Negative contact as an essential condition.” International Journal of
Comparative Sociology, 59(1), 25–43.
Transue, J. (2007). “Identity Salience, Identity Acceptance, and Racial Policy Attitudes.”
American Journal of Political Science, 51(1):78–91.
Van Leeuwen, F., Koenig, B. L., Graham, J., & Park, J. H. (2014). “Moral concerns across the
United States: Associations with life-history variables, pathogen prevalence,
urbanization, cognitive ability, and social class.” Evolution and Human Behavior, 35(6),
464-471.
Van Vilet, L., (2021). “Moral Expressions in 280 Characters or Less: An Analysis of Politician
Tweets Following the 2016 Brexit Referendum Vote,” Frontiers in Big Data: 4.
Vohs, K. D., Mead, N. L., Goode, M. R. (2006). “The psychological consequences of

money.” Science 314(5802): 1154-1156.
Wang, S. Y. N., Inbar, Y. (2021). “Moral-language use by US political elites.” Psychological

Science, 32(1), 14-26.
Wenzel, M., Mummendey, A., Waldzus, S. (2010). “Superordinate Identities and Intergroup
Conflict: The Intergroup Projection Model.” European Review of Social Psychology 18:1
331-372.
Wimmer, H., Perner, J. (1983). “Beliefs about beliefs: Representation and constraining function
of wrong beliefs in young children's understanding of deception.” Cognition, 13(1): 103–
128.
Wojcieszak, M., Azrout, R. (2016). “I Saw You in the News: Mediated and Direct Intergroup
Contact Improve Outgroup Attitudes.” Journal of Communication, 66:1032–60.
Zelizer, V.A. (1979). Morals and Markets: The Development of Life Insurance in the United
States. New York: Columbia University Press.
31

Zelizer, V.A. (1994). Pricing the priceless child: The changing social value of children.
Princeton: Princeton University Press.
Zelizer, V.A. (2007). The Purchase of Intimacy. Princeton: Princeton University Press.
Zelizer, V.A. (2010). Economic Lives: How Culture Shapes the Economy. Princeton: Princeton
University Press.
Ziller, C. (2022). “Established and excluded? Immigrants’ economic progress, attitudes toward
immigrants, and the conditioning role of egalitarianism and intergroup contact.” Political
Studies 70(1): 236-256.
32

The Humanizing Effect of Market Interaction - Figures and
Tables
Table of Contents
1 Descriptives 2
1.1 Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Table 1 - Variable Descriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Table 2 - Summary Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Formulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 Main Paper 4
Figure 1 - Extended Moral Foundations Dictionary Words . . . . . . . . . . . . . . . . . 4
Table 3 - Baseline Models: Trade and Immigration . . . . . . . . . . . . . . . . . . . . 5
Table 4 - Full Models: Trade and Immigration . . . . . . . . . . . . . . . . . . . . . . 6
Table 5 - Vice/Virtue Valence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Table 6 - Bonding and Bridging by Valence - Trade . . . . . . . . . . . . . . . . . . . . 8
Table 7 - Bonding and Bridging by Valence - Immigration . . . . . . . . . . . . . . . . 8
3 Appendix 9
Table A1 - Reference Country List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Table A2 - Full Models with Outliers: Trade and Immigration . . . . . . . . . . . . . . 10
Table A3 - Controls Robustness Check - Trade . . . . . . . . . . . . . . . . . . . . . . . 11
Table A4 - Controls Robustness Check - Immigration . . . . . . . . . . . . . . . . . . . . 12
Table A5 - Fraser Freedom Robustness Check . . . . . . . . . . . . . . . . . . . . . . . . 13
Table A6 - Bonding and Bridging - Full - Trade . . . . . . . . . . . . . . . . . . . . . . . 14
Table A7 - Bonding and Bridging - Full - Immigration . . . . . . . . . . . . . . . . . . . 14
33
1 DESCRIPTIVES
1 Descriptives
1.1 Variables
Table 1: Variable Descriptions
Variable Description Source

Dependent Variables
Moral Language Country-year average of sentiment scores for Authors’ calculations.
(multiple) NYT articles. Based on the Extended Moral
Foundation Dictionary.
Independent Variables
Trade Sum of US export and imports to/from reference Correlates of War.
country.
Immigrants Count of immigrants to the US from reference US Department of Jus-
country. tice, Immigration and
Naturalization Service.
Controls
MEPV Sum of the magnitudes of civil and interstate Center for Systemic
violence and war. Peace.
Polity Score Standard Polity score ranging from 0 (hereditary Polity 5. Center for Sys-
monarchy) to 20 (consolidated democracy). temic Peace.
Years of Schooling Average number of years of schooling. Barrow and Lee (2013).
Protestant Percent of population that is protestant. Correlates of War.
Islamic Percent of population that is muslim. Correlates of War.
Gini Gini coefficient. Higher values of the Gini coef- World Bank.
ficient indicates greater wealth inequality.
GDP Per Capita GDP per capita in 2010 USD. World Bank.
EFW Measure of economic freedom. Fraser Institute.
Ethnic Fractionalization Index corresponding to the probability that two Drazanova (2019)
individuals are from different ethnic groups.
Internet Use Perfect of population that uses the internet World Bank
FE Labor Part Rate Female labor participation rate. Word Bank.
34
1.2 Formulas 1 DESCRIPTIVES
Table 2: Summary Statistics
Statistic N Mean SD Min P(25) P(75) Max

Moral Language 3,722 258,187.300 23,212.550 165,393.800 240,761.200 276,185.900 347,600.000
Trade 3,383 10,184.1 38,343.3 0.0 76.5 4,205.6 544,015.1
Immigration 3,136 5,765.8 26,263.3 0.0 143.0 3,921.7 946,167.0
MEPV 3,158 0.8 1.8 0.0 0.0 0.0 14.0
Polity Score 3,083 2.2 6.9 −10.0 −5.0 9.0 10.0
Average Years of Schooling 3,200 6.7 1.4 0.8 5.7 7.0 12.7
Percent Protestant 3,200 0.1 0.1 0.0 0.1 0.1 0.9
Percent Islam 3,200 0.1 0.2 0.0 0.0 0.0 1.0
Gini 3,717 41.9 10.1 20.7 33.7 49.5 65.8
GDP Per Capita 3,331 11,715.9 19,035.4 178.8 1,227.1 12,642.4 170,534.7
EFW 1,206 6.5 1.1 2.4 5.8 7.4 8.7
Ethnolinguistic Fractionalization 2,997 0.4 0.2 0.0 0.1 0.6 0.8
Internet Use 3,035 7.8 16.3 0.0 0.00 5.9 90.6
Female Labor Participation Rate 2,925 49.6 17.1 8.0 38.5 60.6 90.7
1.2 Formulas
I
1X
µct = ϕcti
n i=1
Yct = β0 + β1 τct + βi Xit + αc + δt + ϵct
35
2 MAIN PAPER
2 Main Paper
Figure 1: Extended Moral Foundations Dictionary Words
Note: Larger words indicates correlate to words that received higher probabilities
within its eMFD foundation.
36
2 MAIN PAPER
Table 3: Baseline Models: Trade and Immigration
Moral Language
(1) (2)
Trade 0.029**
(0.01)
Immigration 0.154***
(0.06)
Observations 3341 3098
R2 0.111 0.117
Country Fixed Effects Yes Yes
Year Fixed Effects Yes Yes
∗
p<.10; ∗∗ p<.05; ∗∗∗ p<.01. Standard errors in parentheses are
clustered at the country level.
37
2 MAIN PAPER
Table 4: Full Models: Trade and Immigration
Moral Language
(1) (2)
Trade 0.025***
(0.01)
Immigration 0.162**
(0.08)
MEPV 948.418 838.731
(749.51) (718.04)
Polity2 -183.838 -180.718
(173.17) (175.65)
Gini -43.845 -32.027
(113.09) (113.98)
GDP per Capita 0.207 0.241
(0.17) (0.15)
Ethnic Fractionalization 50360.356 46475.269
(31744.93) (33727.50)
Protestant 9409.592 -271.332
(14010.33) (16897.02)
Islam -6219.912 -8691.083
(39655.10) (39689.85)
Internet Use -37.797 -44.605
(43.71) (44.04)
Years of Schooling 904.466 956.801
(1104.70) (1244.22)
FE Labor Part Rate -205.568 -188.444
(166.49) (168.82)
R2 0.141 0.142
∗
p<.10; ∗∗ p<.05; ∗∗∗ p<.01. Standard errors in parentheses
are clustered at the country level.
38
2 MAIN PAPER
Table 5: Vice/Virtue Valence
Trade Immigration
(1) (2) (3) (4)
Virtue Vice Virtue Vice
Trade 0.013*** 0.012
(0.00) (0.01)
Immigration 0.092*** 0.070
(0.02) (0.07)
Observations 1343 1343 1327 1327
R2 0.080 0.140 0.086 0.140
Country Fixed Effects Yes Yes Yes Yes
Year Fixed Effects Yes Yes Yes Yes
Full Controls Yes Yes Yes Yes
∗
p<.10; ∗∗ p<.05; ∗∗∗ p<.01. Standard errors in
parentheses are clustered at the country level.
Control coefficients are truncated for brevity.
39
2 MAIN PAPER
Table 6: Bonding and Bridging by Valence - Trade
Bridging Bonding
(1) (2) (3) (4)
Trade 0.008*** 0.003 0.004* 0.009**
(0.00) (0.00) (0.00) (0.00)
Observations 1343 1343 1343 1343
R2 0.048 0.104 0.123 0.176
Table 7: Bonding and Bridging by Valence - Immigration
Bridging Bonding
(1) (2) (3) (4)
Immigration 0.037** 0.029 0.055*** 0.041
(0.02) (0.04) (0.02) (0.04)
Observations 1327 1327 1327 1327
R2 0.040 0.099 0.127 0.172
40
3 APPENDIX
3 Appendix
Table A1: Reference Country List
Afghanistan Comoros Iran Monaco Slovenia

Albania Costa Rica Iraq Mongolia Solomon Islands
Algeria Croatia Ireland Morocco Somalia
Andorra Cuba Israel Mozambique South Africa
Angola Cyprus Italy Myanmar South Korea
Argentina Czech Republic Ivory Coast Namibia Spain
Armenia Denmark Jamaica Nauru Sri Lanka
Australia Djibouti Japan Nepal Sudan
Austria Dominican Jordan Netherlands Suriname
Republic
Azerbaijan East Timor Kazakhstan New Zealand Swaziland
Bahamas Ecuador Kenya Nicaragua Sweden
Bahrain Egypt Kiribati Niger Switzerland
Bangladesh El Salvador Kosovo Nigeria Syria
Barbados Equatorial Kuwait North Korea Taiwan
Guinea
Belarus Eritrea Kyrgyzstan Norway Tajikistan
Belgium Estonia Laos Oman Tanzania
Belize Ethiopia Latvia Pakistan Thailand
Benin Federated States Lebanon Palau Togo
of Micronesia
Bhutan Fiji Lesotho Panama Tonga
Bolivia Finland Liberia Papua New Tunisia
Guinea
Bosnia and France Libya Paraguay Turkey
Herzegovina
Botswana Gabon Liechtenstein Peru Turkmenistan
Brazil Gambia Lithuania Philippines Tuvalu
Brunei Georgia Luxembourg Poland Uganda
Bulgaria Germany Macedonia Portugal Ukraine
Burkina Faso Ghana Madagascar Qatar United Arab
Emirates
Burundi Greece Malawi Romania United Kingdom
Cambodia Grenada Malaysia Russia Uruguay
Cameroon Guatemala Maldives Rwanda Uzbekistan
Canada Guyana Mali Samoa Vanuatu
Cape Verde Haiti Malta San Marino Venezuela
Central African Honduras Marshall Islands Saudi Arabia Vietnam
Republic
Chad Hungary Mauritania Senegal Yemen
Chile Iceland Mauritius Seychelles Zambia
China India Mexico Sierra Leone Zimbabwe
Colombia Indonesia Moldova Singapore
41
3 APPENDIX
Table A2: Full Models with Outliers: Trade and Immigration
Moral Language
(1) (2)
Trade 0.024***
(0.01)
Immigration -0.013**
(0.01)
MEPV 955.817 888.699
(709.91) (701.81)
Polity2 -186.816 -194.485
(171.12) (175.58)
Gini -39.299 -12.425
(111.65) (114.28)
(0.15) (0.15)
Ethnic Fractionalization 48975.047 44843.590
(30686.67) (33383.42)
Protestant 9952.472 2841.530
(13805.11) (16311.91)
Islam -4715.911 -7817.303
(39483.55) (39525.78)
Internet Use -32.302 -21.169
(39.99) (43.22)
(1074.16) (1219.74)
(158.88) (167.77)
R2 0.140 0.137
∗
p<.10; ∗∗ p<.05; ∗∗∗
p<.01. Standard errors in parentheses are clustered at the
country level.
42
Table A3: Controls Robustness Check - Trade
Moral Language
(1) (2) (3) (4) (5) (6) (7) (8)
Trade 0.032* 0.030 0.033*** 0.033*** 0.028** 0.032*** 0.032*** 0.029***
(0.02) (0.02) (0.01) (0.01) (0.01) (0.01) (0.01) (0.01)

MEPV 1738.366*** 1789.070*** 801.619 788.714 802.328 1136.732* 1110.416* 954.887
(456.56) (506.43) (730.14) (731.71) (765.83) (621.65) (623.02) (755.16)
Polity2 -84.249 -139.687 -134.751 -118.979 -168.435 -188.370 -161.804
(125.60) (164.34) (168.61) (170.95) (177.10) (177.46) (175.03)
Gini -111.183 -126.537 -126.476 -169.644 -169.829 -36.137
(129.97) (130.28) (132.03) (136.49) (135.99) (115.34)
GDP per Capita 0.070 0.095 0.025 0.127 0.134
(0.10) (0.12) (0.14) (0.17) (0.16)
43
Ethnic Fractionalization 16240.671 2660.901 298.301 48010.917

(29251.14) (32924.99) (33956.95) (30913.47)
Protestant 530.340 -2495.482 7321.865
(13777.63) (14381.36) (13955.73)
Islam 20820.740 13286.033 -4522.028
(36830.48) (35903.50) (39193.19)
Internet Use -36.619 -38.560
(43.18) (44.78)
Years of Schooling 902.038
(1095.24)
Observations 3069 2996 1705 1687 1585 1509 1485 1343
R2 0.131 0.132 0.112 0.113 0.117 0.107 0.110 0.140
Country Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes
Year Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes
∗ ∗∗ ∗∗∗
p<.10; p<.05; p<.01. Standard errors in parentheses are clustered at the country level.
Table A4: Controls Robustness Check - Immigration
Moral Language
(1) (2) (3) (4) (5) (6) (7) (8)
Immigration 0.151*** 0.160*** 0.205*** 0.214*** 0.194** 0.184** 0.183** 0.172**
(0.05) (0.05) (0.06) (0.06) (0.08) (0.08) (0.08) (0.08)

MEPV 1520.010*** 1622.153*** 930.705 913.283 949.234 1024.016* 993.587 833.184
(450.65) (491.89) (642.05) (643.31) (670.82) (601.97) (602.38) (722.05)
Polity2 -92.516 -132.019 -126.565 -117.261 -171.044 -195.291 -159.807
(143.29) (165.68) (170.43) (173.99) (176.85) (176.41) (175.80)
Gini -120.042 -136.439 -133.582 -160.921 -158.684 -22.917
(133.45) (133.65) (134.52) (136.10) (135.60) (115.93)
GDP per Capita 0.082 0.120 0.041 0.197 0.177
(0.10) (0.12) (0.13) (0.14) (0.14)
44
Ethnic Fractionalization 9185.238 3898.180 3449.430 44672.529

(33148.71) (34613.84) (36100.14) (32768.30)
Protestant -10288.338 -14133.267 -2823.221
(16195.75) (16890.23) (16538.44)
Islam 12629.812 4669.751 -7165.897
(36321.24) (35334.64) (39320.32)
Internet Use -53.405 -45.994
(41.74) (44.72)
Years of Schooling 1001.224
(1231.31)
Observations 2768 2703 1643 1625 1525 1491 1467 1327
R2 0.127 0.128 0.114 0.115 0.118 0.109 0.113 0.141
Country Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes
Year Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes
∗ ∗∗ ∗∗∗
Table A5: Economic Freedom Robustness Check
Moral Language
(1) (2)
Trade 0.030***
(0.01)
Immigration -0.003
(0.10)
MEPV 1187.620 1102.126
(718.75) (705.83)
Polity2 -350.620* -367.384*
(191.06) (192.02)
Gini 127.908 149.057
(153.65) (154.68)
(0.21) (0.22)
Ethnic Fractionalization 80064.309*** 90697.276***
(30154.13) (34041.01)
Protestant 21307.017 22110.559
(20234.35) (20053.35)
Islam 45501.406 49775.445
(52168.18) (52506.92)
Internet Use -19.367 -28.229
(50.87) (52.98)
(1364.90) (1382.00)
(207.11) (208.71)
EFW 1720.408 1745.952
(1419.11) (1454.57)
R2 0.203 0.201
∗ ∗∗ ∗∗∗
45
Table A6: Bonding and Bridging - Full - Trade
Virtue Vice
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Authority Care Fairness Loyalty Sanctity Authority Care Fairness Loyalty Sanctity
Trade 0.002 0.004*** 0.005*** 0.003** -0.000 0.007** 0.001 0.002 0.002 0.000
(0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00) (0.00)
Observations 1343 1343 1343 1343 1343 1343 1343 1343 1343 1343
R2 0.213 0.099 0.063 0.129 0.166 0.169 0.131 0.081 0.192 0.101
Country Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Year Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Full Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Table A7: Bonding and Bridging - Full - Immigration
Virtue Vice
46
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Authority Care Fairness Loyalty Sanctity Authority Care Fairness Loyalty Sanctity
Immigration 0.028** 0.014* 0.023** 0.017 0.010** 0.042 0.006 0.024 0.002 -0.002
(0.01) (0.01) (0.01) (0.01) (0.00) (0.03) (0.02) (0.02) (0.01) (0.01)
Observations 1327 1327 1327 1327 1327 1327 1327 1327 1327 1327
R2 0.216 0.082 0.066 0.136 0.169 0.165 0.117 0.081 0.194 0.109
Country Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Year Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Full Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
3 APPENDIX

SSRN Id3997378

Uploaded by

Copyright:

Available Formats

SSRN Id3997378

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SSRN Id3997378

Uploaded by

Copyright:

Available Formats

The Humanizing Effect of Market Interaction

Colin Harris† Andrew Myers‡ Adam Kaiser§

JEL Codes: F14, F60, J15, P17, Z1

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

3.3 Measuring Market Interaction

Electronic copy available at: https://ssrn.com/abstract=3997378

3.5 Two-Way Fixed Effects

Electronic copy available at: https://ssrn.com/abstract=3997378

4 The (De)Humanizing Effect of Market Interaction

The Dehumanization Hypothesis: market interaction will result in dehumanization because

The Infrahumanization Hypothesis: market interaction will result in infrahumanization because

We explore each below.

4.1 The Dehumanization Hypothesis

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

4.2 The Infrahumanization Hypothesis

Electronic copy available at: https://ssrn.com/abstract=3997378

4.3 The Humanization Hypothesis

Electronic copy available at: https://ssrn.com/abstract=3997378

5 Testing the (De)Humanization Hypotheses

5.1 Infrahumanization and (De)Humanization

Electronic copy available at: https://ssrn.com/abstract=3997378

5.2 Methods of (De)Humanization

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Bar-Tal, D. (1989). “Delegitimization: The extreme case of stereotyping and prejudice.” In D.

Electronic copy available at: https://ssrn.com/abstract=3997378

Cohen, G. A. (2009). Why not Socialism? Princeton: Princeton University Press.

Corak, M. (2013). “Inequality from Generation to Generation: The United States in

Electronic copy available at: https://ssrn.com/abstract=3997378

Frith, U., Frith, C. D. (2003). “Development and neurophysiology of mentalizing”. Philosophical

Gallacher, J. D., Heerdink, M. W., Hewstone, M. (2021). “Online Engagement Between

Electronic copy available at: https://ssrn.com/abstract=3997378

Gudeman, S. (2016). Anthropology and Economy. Cambridge: Cambridge University Press.

Gudykunst, W. B. (1985). “A Model of Uncertainty Reduction in Intercultural

Haslam, N. (2006). “Dehumanization: An integrative review.” Personality and Social

Electronic copy available at: https://ssrn.com/abstract=3997378

Electronic copy available at: https://ssrn.com/abstract=3997378

Lavoie, D., Chamlee-Wright, E. (2000). Culture and Enterprise: The Development,

Electronic copy available at: https://ssrn.com/abstract=3997378

Pettigrew, T. F., Tropp, L. R. (2006). “A meta-analytic test of intergroup contact

Radin M.J. (1987). “Market-inalienability.” Harvard Law Review 100(8): 1849–1937.

Electronic copy available at: https://ssrn.com/abstract=3997378

Staerklé, C., Green, E. G. (2018). “Right‐wing populism as a social representation: A

Electronic copy available at: https://ssrn.com/abstract=3997378

The Pew Research Center (2021). “The Newspapers Fact Sheet.”