Lay attitudes to trade with low-wage countries

Judgment and Decision Making, Vol. 3, No. 4, April 2008, pp. 335–343
Lay attitudes to trade with low-wage countries
Simon Kemp?
Department of Psychology
University of Canterbury
Abstract
Three studies presented scenarios to lay people to investigate their willingness to restrict imports. Greater restriction
was preferred when similar goods were made at home, when the owners of the foreign businesses made very good
pro?ts, and, less consistently, when the goods came from a low wage country. Particular reluctance to import from a
low-wage country did not vary with whether a home ?rm was likely to lose business or the level of understanding of
comparative advantage, but was related to the pro?ts made by foreign business owners. The results show that lay people
views are based on concern for people in other countries as well as in their own.
Keywords: trade policy, attitudes, comparative advantage, lay economics.
1 Introduction
e.g., Corden, 1974; Davidson et al., 2006; Gomes, 2003;
Kemp, 2007; Mayda & Rodrik, 2005; Mayer, 1984; Ol-
Do people have a particular reluctance to their country
son, 1982; Scheve & Slaughter, 2001).
importing goods from low-wage countries? If so, what
The focus of the present paper is narrower: Do peo-
might be the basis for this reluctance? Both these ques-
ple in developed countries particularly dislike importing
tions have been raised within the ?eld of international
goods from countries with low wages and low labour
economics (e.g., Krugman & Obstfeld, 2000), but have
costs? And if so, why?
not to my knowledge been the subject of empirical inves-
Before reviewing some reasons why people might not
tigation. This paper investigates them.
want to import from low-wage countries, it is worthwhile
To set the issue in context, it is worth ?rst noting that
to note that there at least two reasons for expecting pre-
lay people generally are not convinced of the bene?ts of
cisely the opposite result. Firstly, and most obviously,
free international trade, and recent survey results indi-
goods from such countries are likely to be cheaper. Tak-
cate that a majority of people in many western countries
ing a strictly sel?sh perspective, the vast majority of peo-
would prefer to see more rather than fewer restrictions
ple bene?t from being able to buy, for example, shirts,
on imports than exist at present (e.g., Mayda & Rodrick,
dairy products, or cars that are cheaper but of comparable
2005; Hainmueller & Hiscox, 2006). In this respect, the
quality to those made in one’s own country. Secondly,
general public is at odds with the economics profession,
taking a more utilitarian perspective, poorly paid work-
samples of which have produced substantial majorities
ers in a foreign country could be seen as more in need
in favour of freeing up trade (e.g. Alston et al., 1992;
of one’s custom than higher-paid workers at home or in a
Frey et al., 1984). The difference in views may well go
rich foreign country.
back to the nineteenth century, and there have been sev-
On the other hand, there are several possible rea-
eral suggestions as to why the difference might arise (see,
sons why people might oppose importing from low-wage
countries. Three of these reasons are investigated here:
?The research was supported by a grant from the Psychology De-
partment of the University of Canterbury. I am grateful for stimulat-
that people fear the business and job losses that result
ing interactions with Jonathan Baron and Friedel Bolle. I am also very
from local higher wage workers having to compete with
grateful to Friedel for his help in translating the questionnaire of Study
low wage workers; that the dislike stems from failure to
2, to Anika Köhler for help in obtaining a sample from eastern Germany,
understand Ricardo’s principle of comparative advantage;
and to the Alexander von Humboldt Foundation for supporting me in
Germany. Some of this work was presented at the 32nd Conference
and that people fear that low-paid foreign workers are
of the International Association for Research in Economic Psychology
being exploited. These three reasons were investigated
at Ljubljana, Slovenia in September, 2007, where a number of partici-
because they are related to three “misconceptions about
pants made constructive comments. Finally, I am happy to acknowledge
comparative advantage” discussed by Krugman and Ob-
insightful and helpful suggestions from two anonymous reviewers. Ad-
dress Simon Kemp, Psychology Department, university of Canterbury,
stfeld (2000, pp. 23–26). A fourth reason, that people be-
Christchurch, New Zealand. Email: Simon.Kemp@canterbury.ac.nz.
lieve that goods imported from low-wage countries might
335

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
336
be of lower quality, was not investigated, but instead the
understanding of it tended to be more protectionist in out-
effects of such a belief were controlled for in the studies
look. Krugman and Obstfeld (2000, p. 23) suggest that a
that follow.
common myth — “Free trade is bene?cial only if your
Perhaps the most obvious reason to resist importing
country is strong enough to stand up to competition” —
from foreign low-wage countries is the fear that one’s
arises because of this misunderstanding. The authors go
compatriots — or perhaps even oneself — might lose
on to point out that the principle is important because a
their businesses or livelihoods as a result of this compe-
country which does not have an absolute advantage in
tition. This is by no means an unreasonable fear. Al-
producing a good over some other country might still
though one can debate precisely how much unemploy-
have a comparative advantage. This comes about because
ment arises from this cause rather than from technologi-
the country’s overall lower productivity is then re?ected
cal advance, there is no doubt that people can lose their
in lower wages, enabling it to sell the good for a lower
jobs or their businesses. It is also well-known that such
cost in the other country. However, because the lower
losses can be enduring and have serious consequences
cost is linked to the lower wages, importing such goods
for people’s well-being (Irwin, 2002; Kletzer, 1998; Lu-
might be seen as unfair by people in the other country.
cas et al., 2004; Tre?er, 2001). Moreover, there is al-
Finally, it is possible that people might oppose imports
ready good empirical evidence that people’s opposition
from countries with low labour costs because they believe
to importing goods increases if similar goods are avail-
that workers in these countries are exploited by pro?teer-
able from producers in one’s own country or if the con-
ing business owners who pay low wages (Krugman &
sequences for employment at home are pointed out, al-
Obstfeld, 2000). In this case, the opposition would arise
though this previous research has not investigated the ef-
from altruistic motives on behalf of the foreign workers.
fect of the wage levels in the country supplying the im-
Of course, whether or not the exploited foreign workers
ports (Baron & Kemp, 2004; Hiscox, 2006). It is impor-
really bene?t from such altruistic protectionism is debate-
tant to note that this opposition arises not only from those
able. As Krugman and Obstfeld (2000, p. 24) point out:
whose livelihoods or businesses are personally affected,
“What is the alternative?”
but also from others who are concerned for the welfare of
Three scenario studies investigating these issues are re-
producers and workers in their own country. Such con-
ported below. The ?rst two studies feature research de-
cern could be considered as altruism for one’s compatri-
signs in which three different components or factors in
ots or as parochialism, and connects to a body of previ-
the scenarios are systematically varied. The three factors
ous research investigating situations in which individuals
are: whether the goods under consideration for impor-
may act in the interests of an in-group but not in their own
tation come from a low or high wage country; whether
interests or those of a wider out-group (e.g., Baron, 2001;
similar goods are already made at home and home busi-
Schwartz-Shea & Simmons, 1991; see Kemp, 2007, for
nesses might suffer from the competition; and the level
discussion of trade implications). Nonetheless, although
of pro?t made by the owners of the foreign businesses.
previous work leads us to expect opposition to imports
After reading each scenario, respondents are asked for
that compete with home businesses, it is not clear that
their preferences about restricting the imports. It is ob-
it should particularly matter if the imports originate in a
viously interesting to know whether the restriction pref-
low-wage country rather than one that pays higher wages
erence varies with the other country’s wages, but inter-
but can still produce the goods more cheaply.
active effects are also important for evaluating why such
One of the most important economic arguments that
a preference might arise. If there is a particularly strong
greater welfare results from freer trade derives from Ri-
preference for restricting imports when similar goods are
cardo’s (1817/1971) principle of comparative advantage.
made at home and the imported goods come from a low-
Very brie?y, the principle suggests that where goods
wage country, this is a strong indication that the resis-
should be most ef?ciently produced depends on the ratio
tance to importing from a low-wage country is related to
of the costs of production (comparative advantage) rather
fear of competition and unemployment at home. Simi-
than on the absolute costs (absolute advantage). A corol-
larly, if people want to restrict imports from a low-wage
lary is that a country need not be the most ef?cient pro-
country because they fear exploitation, we would expect
ducer of any commodity in order to bene?t from trade.
resistance to importing from a country where the busi-
Krugman (1994) points out that some people might
ness owners make very large pro?ts to be stronger when
be generally hostile to trade because they misunderstand
the imports come from a low-wage country.
economic arguments for it, such as Ricardo’s principle
Study 3 used a somewhat different method to examine
of comparative advantage. Ricardo’s principle is neither
the relationship between people’s restriction preference
trivial nor obvious (Samuelson, 1972), and Baron and
and a test of their understanding of the principle of com-
Kemp (2004) found both that the principle was poorly
parative advantage (taken from Baron & Kemp, 2004).
understood by lay people, and that people with a lower
The ?rst and third studies employed small samples from

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
337
New Zealand; the second a small sample from eastern
2.1.2 Respondents
Germany. Some speci?c predictions are made before the
last two studies.
The questionnaire was completed by 18 students and 47
non-student members of the general public. The students
were recruited during a scheduled class in psychology,
2 Study 1
the general public by asking a number of paid student
interviewers to recruit up to six non-students from people
2.1 Method
they knew around Christchurch, New Zealand. Overall,
28 respondents were male. The median age of the sample
2.1.1 Questionnaire
was in the range 25–34 years, and ranged from 15–24
Respondents were asked to complete a questionnaire on
years (27 respondents) to over 65 (2 respondents). For
trade. Each questionnaire presented the respondent with
some analyses reported below, respondents were divided
eight scenarios about the importation of consumer goods,
into two age groups, 34 or younger (45 respondents) and
and asked how they felt about restricting the importation
35 or older (20 respondents).
of the goods on a nine-point scale from 1 (Import with
no restrictions or tariffs) through 5 (Import some goods
2.2 Results
with tariffs) to 9 (Allow no goods to be imported at all).
This scale is referred to below as the restriction measure.
Averaged over the eight scenarios, the restriction score
They were also asked whether such imports would bene?t
(scaled from 1, no restriction, to 9, complete restriction)
the people of your country on the whole, and whether the
was 4.44 (SD = 1.48). Two respondents always preferred
imports bene?t the people of the other country. These
completely unrestricted trade (all restriction scores were
two questions were answered on 5-point scales from 1
1), and one always entered 5. The average bene?t to one’s
(Certainly not) through 3 (Not sure) to 5 (Certainly yes).
own country was 3.37 (SD = 0.64), that to the other coun-
The eight scenarios varied the two constructs in a 2 X 2
try 3.58 (SD = 0.58). (Bene?t scores were scaled from 1,
X 2 within-subjects design. The constructs were whether
certainly not, to 5, certainly yes.)
the same goods were produced in New Zealand or not;
The key results were tested in three analyses of vari-
whether the goods were produced in a country that pays
ance, all featuring a 2 (goods made in NZ versus not
higher or lower wages to its workers than New Zealand;
made in NZ) by 2 (high or low-wage exporting country)
and whether the owners of the businesses made “some
by 2 (very good versus some pro?ts made by the business
but not very large pro?ts” or “very good pro?ts”. In
owners) within-subjects design.
all cases respondents were told that the workers did not
There was greater willingness to restrict imports if that
own the businesses. When the goods were produced in
type of good was made in New Zealand (Restriction Av-
New Zealand, respondents were also informed: “If peo-
erage = 5.30) than if it was not (Average = 3.58; F(1,61)
ple buy the foreign goods a ?rm in New Zealand will
= 71.5, p < .001). Respondents were keener to restrict
lose business”. For the Goods produced in New Zealand
imports from low-wage (Average 5.07) than high-wage
and Higher wage condition, respondents were informed
(Average = 3.81) countries (F(1, 61) = 49.1, p < .001).
that “this type of [imported] goods is still cheaper and of
More restriction was preferred when the business own-
equally good quality to that produced in New Zealand.”
ers made very good (Average = 4.79) rather than some
The order of the eight scenarios was systematically
pro?t(Average = 4.09; F(1, 61) = 29.9, p < .001). As
varied, and each scenario was on a separate page. An
shown in Table 1, there was a particular preference for re-
example of a scenario (produced in New Zealand; higher
stricting imports from low-wage countries when the busi-
wages; very good pro?ts) with bold type as on the ques-
ness owners made very good pro?ts (F(1, 61) = 17.1, p
tionnaire read:
< .001). There were no other statistically signi?cant (p <
Goods of this type are produced in New
.05) interactions.
Zealand. If people buy the foreign goods a
The difference between the restriction scores for im-
?rm in New Zealand will lose business.
porting from the high and low wage countries was cal-
culated for each pairing of the other factors (e.g., made
The goods are produced in a country that pays
in New Zealand; some pro?ts) for each respondent. This
higher wages to its workers than New Zealand,
analysis showed that 16 (of the 65) respondents consis-
but this type of goods is still cheaper and of
tently favoured more restriction on importing from poor
equally good quality to that produced in New
countries for all four combinations of the other factors; 3
Zealand.
were consistently indifferent; and none consistently pre-
The workers do not own the businesses. The
ferred to import from the low-wage country. For the low
owners make very good pro?ts.
wage-condition only, a restriction score difference be-

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
338
ported above were repeated ?rstly with the addition of
Table 1: Some results from Study 1. Average restriction
the sample factor only, secondly with the addition of the
scores and bene?t to the other country ratings for import-
sex factor only, and ?nally with the addition of the age
ing from high or low wage countries where the business
factor only. These nine extra analyses uncovered a to-
owners make some or very good pro?ts.
tal of three statistically signi?cant effects of the back-
Low wage
High wage
ground variables or their interactions with the scenario
variables.1 Crucially, there were no statistically signi?-
Restriction scores
cant interactions of the background variables with the re-
Very good pro?ts
5.62
3.96
sults reported earlier, showing that these ?ndings are not
Some pro?ts
4.52
3.66
moderated by these background variables.
Bene?t to the other country
Very good pro?ts
2.98
3.98
3 Study 2
Some pro?ts
3.41
3.96
It may be questioned whether the results of Study 1 hold
Note. Restriction scores range from 1 (import with no
outside New Zealand. Study 2 was a replication of Study
restrictions or tariffs) to 9 (allow no goods to be imported
1 carried out in eastern Germany. The area was chosen
at all). Bene?t to the other country ratings range from 1
partly for reasons of convenience, but also because there
(Certainly not) to 5 (Certainly yes).
are several important social and economic differences to
New Zealand. Firstly, up until 1990, it was a separate
country with a planned, socialist economy. Secondly, un-
tween that for very good and some pro?ts was also cal-
like New Zealand, it currently forms part of a major world
culated both for when the goods were made in NZ and
trading block (the European Union). Thirdly, at the time
when they were not. Thirty respondents gave consistent
of conducting Study 1, the of?cial unemployment rate in
answers over these two “where made” conditions: 19
both New Zealand and the Christchurch area was well
respondents advocated more restriction with very good
under 5 percent. By contrast, at the time of Study 2, the
pro?ts, 10 advocated the same amount of restriction, and
unemployment rate throughout the eastern states of Ger-
1 advocated less restriction with very good pro?ts. Taken
many was around 15 percent. An obvious expectation
together these results show some degree of consistency
was that the higher unemployment rate might lead people
within the respondents but also a good deal of individual
to be more protectionist.
variation that was not related to the manipulated factors.
The rated bene?t to one’s own country was signi?-
cantly greater when the goods were made at home (Aver-
3.1 Method
age bene?t = 3.89) than when they were not (Average =
The questionnaire and recruitment methods were gener-
2.85; F(1, 60) = 69.1, p < .001). There were no other sta-
ally as similar as possible to those used in Study 1. The
tistically signi?cant (p < .05) main or interactive effects.
questionnaire from that study was translated into German
The rated bene?t to the other country was slightly but
by a native German speaker and then checked against the
signi?cantly greater when the goods were not made in
original version by a native speaker of English who spoke
NZ (Average bene?t = 3.65) than when they were (Av-
?uent German. As in the previous experiment, the sam-
erage bene?t = 3.52; F(1, 61) = 4.23, p = .044). The
ple contained a mixture of students and the general pub-
rated bene?t was also higher when the wages were high
lic. A requirement for both students and general public
(Average = 3.97) rather than low (Average = 3.19; F(1,
was that they were east German. Both the students and
61) = 41.7, p < .001) and when some (Average = 3.69)
the general publics were recruited by paid interviewers,
rather than very good (Average = 3.48) pro?ts were made
whose instructions were translations of those given to the
by the business owners (F(1, 61) = 9.1, p = .004). As
New Zealand interviewers of Study 1. The ?nal sample
shown in Table 1, the effects of pro?t were stronger when
contained 25 student and 50 general public respondents.
importing from the low-wage country (F(1, 61) = 16.1,
p < .001). The bene?t to the other country was seen as
1When sample was added, there was a three way interactive effect
particularly low when the other country’s business own-
of where the goods were made, the level of pro?t and the sample on
the perceived bene?t to the other country (F(1, 60) = 4.51, p = .038),
ers made very good pro?ts but paid low wages, a result
with the students showing a more pronounced interactive effect than the
mirroring that found for the restriction preference.
general public. Women were overall more willing to restrict imports
Analyses of variance were also conducted with three
(Average = 4.83) than men (Average = 3.89; F(1, 60) = 6.62, p = .013).
background variables: sample, respondent sex and re-
There was a signi?cant interactive effect of age and where the goods
were made, with older people seeing less bene?t to New Zealand of
spondent age (dichotomised as younger or older). Be-
importing goods when similar were made at home (F(1, 59) = 4.54, p =
cause these three variables intercorrelate, the analyses re-
.037).

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
339
There were 35 male and 40 female respondents. Overall,
the median age was in the band 35–44 years with a range
Table 2: Some results from Study 2. Average restriction
from under 25 to over 65 years of age. Thirty-four re-
scores and bene?t to one’s own country ratings for im-
spondents were aged 34 or younger, 41 were 35 or older.
porting from countries where the business owners make
some or very good pro?ts, as a function of the wage levels
in the other country and (restriction scores only) whether
3.2 Results
the goods were made in Germany or not.
Over the eight scenarios, the restriction score averaged
Owner pro?ts:
Very good
Some
4.70 (SD = 1.74). Four respondents always entered the
Restriction scores
same restriction score of 1. Rated bene?t to one’s own
country averaged 3.35 (SD = 0.89), that to the other coun-
High wages
4.72
4.55
try 3.91 (SD = 0.80).
Low wages
5.21
4.31
There was greater willingness to restrict goods if that
Made in Germany
5.62
5.42
type was made in Germany (Average = 5.52) than if it was
Not made in Germany
4.31
3.44
not (Average = 3.87; F(1,72) = 55.3, p < .001). Respon-
Bene?t to one’s own country
dents more concerned to restrict imports where the own-
ers made very good (Average = 4.97) rather than some
High wages
3.49
3.41
pro?ts (Average = 4.43; F(1, 72) = 11.42, p = .001). As
Low wages
3.13
3.39
shown in Table 2, there was a particular reluctance to im-
Note. Restriction scores range from 1 (import with no
port from low-wage countries when the business owners
restrictions or tariffs) to 9 (allow no goods to be imported
made very good pro?ts, a similar result to Study 1 (F(1,
at all). Bene?t to one’s own country ratings range from 1
72) = 11.1, p = .001). Table 2 also illustrates that, perhaps
(Certainly not) to 5 (Certainly yes).
surprisingly, respondents were more sensitive to pro?ts
when similar goods were not made locally than when they
were (F(1, 72) = 9.2, p = .003). There were no other sta-
tistically signi?cant (p < .05) interactions, and the main
business owners make very good pro?ts (see Table 2; F(1,
effect of wages was also not signi?cant.
73) = 6.42, p = .013).
The difference between the restriction scores for im-
The perceived bene?t to the other country was signi?-
porting from the high and low wage countries again was
cantly greater when the goods were not made in Germany
calculated for each pairing of the other factors and each
(Average bene?t = 4.03) than when they were (Average
respondent. Three (of the 75) respondents consistently
bene?t = 3.79; F(1, 73) = 12.2, p = .001). Again, the
favoured more restriction on importing from poor coun-
bene?t was perceived to be higher when the wages were
tries for all four combinations of the other factors; 8 were
high (Average = 4.05) than when they were low (Average
consistently indifferent; and 4 consistently preferred to
= 3.77; F(1, 73) = 8.3, p = .005). However, there was no
import from the low-wage country. As for Study 1, for
signi?cant main effect of pro?t level, and nor were there
the low wage-condition only, a restriction score differ-
any statistically signi?cant (p < .05) interactions.
ence between very good and some pro?ts was calculated
As for Study 1, these analyses were repeated with the
for both the made in Germany and not made in Germany
successive addition of sample, sex and age factors. None
conditions. Forty-four respondents answered consistently
of these factors nor their interactions produced any signif-
over the two conditions: 23 respondents advocated more
icant (p < .05) results on the restriction measure or on the
restriction with very good pro?ts, 60 favored the same
bene?t to one’s own country ratings, but there were some
amount of restriction, and 5 favored less restriction with
signi?cant effects on the ratings of bene?t to the other
very good pro?ts. Again, the results indicate some con-
country. Students saw higher levels of pro?t as more ben-
sistency within the respondents, but a number of respon-
e?cial to the other country than the general public (F(1,
dents displayed idiosyncratic interaction effects.
72) = 7.29, p = .009). Younger people also saw rela-
The rated bene?t to one’s own country was signi?-
tively more bene?t to the other country when higher prof-
cantly greater when the goods were not made at home
its were made by the foreign business owners (F(1, 72) =
(Average bene?t = 3.83) than when they were (Average
8.50, p = .005). Both these effects are consistent with
= 2.88; F(1, 60) = 69.1, p < .001). There was also more
other research indicating a more pro-pro?t, pro-capitalist
bene?t perceived when the goods were made in a high
attitude amongst younger rather than older east Germans
(Average bene?t = 3.45) than in a low wage country (Av-
(Landier et al., 2008). Finally, there was a signi?cant ten-
erage bene?t = 3.26; F(1, 73) = 4.38, p = .040). Finally,
dency for older rather than younger respondents to see
the bene?t to us was perceived as particularly low when
more bene?t for the other country when the goods are not
the imports come from a low wage country where the
made in the home country (F(1, 72) = 5.20, p = .026).

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
340
Again, the background variables did not interact with the
eight items used by Baron and Kemp (2004). All the
?ndings that were the main focus of the study.
items confronted the respondent with a decision about
Some results were compared directly with those from
the ef?cient allocation of production of units that had two
Study 1. Analysis of variance on the combined restric-
components: a computer made up of a processor and pe-
tion scores with country of sample, whether the goods
ripherals or a motorcycle with frame and engine. In this
were made in the home country, the wage level of the ex-
experiment, the two components of the motorcycle were
porting country and the pro?ts made by the owners as in-
made by different branches of a ?rm in one country; the
dependent variables showed no signi?cant (p < .05) main
components of the computer were made in different coun-
effect of country of sample or of the interaction of the
tries. There was a hard and an easy version of each sce-
country with any of the other independent variables, ex-
nario. An example is given below:
cept for wage level (F(1, 133) = 26.0, p < .001). The New
“Two branches of a ?rm, both within one country,
Zealand sample was relatively more unwilling to import
make parts for a motorbike. Branch A can make en-
from a low-wage country. Note especially the lack of in-
gines at a cost of $100 each and frames at a cost of $80
teraction with whether the goods were made in the home
each. Branch B can make engines at a cost of $180 each
country or not. The relatively higher east German un-
[the easy version had $80 each] and frames at a cost of
employment level did not lead to respondents from that
$100 each. A third branch puts the engines and frames
region being especially reluctant to import goods similar
together.
to those made at home.
“Each branch has a limited number of skilled workers
and no possibility of recruiting more. Nor can the work-
4 Study 3
ers be moved between the branches. However, workers
can easily transfer from making engines to frames or vice
Earlier work (Baron & Kemp, 2004) found that people
versa. The motorbikes sell well. “At present, A and
who wanted to restrict imports also tended to have lower
B make equal numbers of motorbikes, and each branch
understanding of Ricardo’s principle of comparative ad-
makes an equal number of frames and engines.
vantage. However, this earlier research did not investigate
“Think about how you allocate work to get the most
whether understanding of the principle is related to par-
ef?cient production.”
ticular unwillingness to import from low-wage countries.
Two speci?c predictions are investigated here. Firstly, if
Respondents then allocated production of both frames
misunderstanding of the principle is related to a partic-
and engines on the (separate) scales: All to Branch A;
ular reluctance to import from low-wage countries, one
Most to Branch A, the rest to Branch B; Equal to both;
would expect that the negative correlation between re-
Most to Branch B, the rest to Branch A; All to Branch B.
striction scores and understanding of the principle found
As in Baron and Kemp (2004), each item was scored
in this earlier research would be moderated by the wage
so that allocating all the production of one component to
level of the country supplying the imports: There should
the producer with greatest comparative advantage scored
be a stronger negative correlation when low-wage rather
one point, “most” received half a point, equal allocation
than high-wage countries are under consideration. Sec-
received 0 points, allocation of “most” to the other pro-
ondly, one would expect individuals who are particularly
ducer lost half a point, and allocation of all to the other a
keen to restrict imports from low-wage countries to have
full point. The ?nal total scores could thus range between
a more limited understanding of the principle.
+8 and -8. (Recall that there are two components for each
question.)
4.1 Method
4.1.1 Questionnaire
Respondents completed a questionnaire consisting of two
4.1.2 Respondents
major parts. The ?rst part was a cut-down version of the
questionnaire used in Study 1. The restriction measure
Sixty-eight non-student members of the New Zealand
was the sole dependent variable and only four scenarios
general public, 28 of them male, completed question-
were presented, which varied whether similar goods were
naires. Thirty-six respondents were between 15 and 24
made in New Zealand or not and whether the goods were
years old, 14 between 25 and 34, 3 between 35 and 44,
imported from a country paying lower or higher wages.
10 between 45 and 54, 3 between 55 and 64, and the re-
In the second part of the questionnaire respondents
mainder were 65 or over. They were recruited by paid
were tested on their understanding of the principle of
student assistants using similar procedures to the previ-
comparative advantage. This test contained four of the
ous experiments.

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
341
4.2 Results
ple’s reluctance to import from high wage countries was
-0.30 (p < .05). Thus, similar correlations were obtained
Analysis of the restriction scores showed greater reluc-
regardless of whether importing from low or high wage
tance to import from the low than the high wage country
countries was in question.
(Average Low wage rating = 5.2; Average High wage rat-
The second analysis ?rst assessed each individual’s
ing = 4.2; F(1,65) = 19.7, p < .001). Similarly the respon-
differential reluctance to import from low versus high
dents were signi?cantly more reluctant to import when
wage countries (i.e. rating of reluctance to import from
the goods were made in New Zealand (Average Rating
low wage countries minus rating of reluctance to import
made in NZ = 5.6; Average Rating not made in NZ =
from high wage countries). This variable was then in turn
3.8; F(1, 65) = 57.3, p < .001). There was no signi?-
correlated with the Comparative Advantage Test scores,
cant interaction between the two factors (F(1, 65) = .70,
but the result was found not to be statistically signi?cant
ns). Essentially, these results replicate those found for the
(r = -.05). Taken together, these two analyses indicate
same factors in Study 1. Taken over the four scenarios,
that greater reluctance to import from low wage countries
the restriction score averaged 4.7 (SD = 1.6).
is not simply explained by the respondents’ frequent mis-
Three respondents had a constant restriction score of 1,
understanding of the principle of comparative advantage.
and one always recorded a score of 5. Twenty-seven re-
spondents favored more restriction when importing from
low-wage countries for both levels of the made in NZ
General discussion
factor; 10 were consistently indifferent, and 7 favored
more restriction when importing from high wage coun-
The New Zealand studies (1 and 3) found respondents
tries. There was a signi?cant tendency for those who fa-
were overall more willing to restrict imports from low
vored more restriction on low-wage imports (as opposed
than from high wage countries. These ?ndings taken to-
to indifference or less restriction) when goods were made
gether with the interactive effect of wage and pro?t level
in NZ to also favor more restriction of low-wage imports
found with the German sample of Study 2, shows that,
when they were not made in New Zealand (?2[df=1] =
indeed, people are sometimes keener to restrict imports
19.0, p < .001). A similar result (?2[df=1] = 16.8, p <
from lower than from high wage countries.
.001) was found for those favoring less restriction on low-
All three studies found the respondents to be more will-
wage imports (as opposed to either indifference or more
ing to restrict imports when similar goods are made in
restriction). These analyses suggest the existence of a
one’s own country, and, unsurprisingly, there was greater
group of people who are particularly sympathetic to im-
perceived bene?t to one’s own country when no similar
porting from low-wage countries, and a group of people
goods are made at home. Moreover, as the differences
who are opposed to such imports.
between the means show, the effect of whether similar
The Comparative Advantage Test score average was
goods are made at home on the restriction measure is
3.0 (SD = 1.6; range -6 to 8). This average implies that
large. However, all three studies indicate that willing-
Ricardo’s principle was generally not well understood by
ness to restrict goods that might cause a home country
the sample. Effectively the score can be interpreted as im-
?rm to lose business is not particularly related to whether
plying that the average respondent often chose to make
the goods come from a low or a high-wage source. If this
the component where there was an absolute advantage
were the case we would have obtained interactions be-
or tried to distribute production equally, but did not of-
tween the effects of source country wage and whether the
ten choose to follow the logic of comparative advantage.
goods are made at home. Not one study showed such
(Similar levels of understanding were reported by Baron
an interaction, and the conclusion is thus that the two
and Kemp, 2004.) Those with more understanding of the
variables have independent effects. Thus, the reason for
principle tended to be more in favour of importing gener-
particular reluctance to import from low-wage countries
ally (Pearson r between restriction and Comparative Ad-
should be sought elsewhere.
vantage Test = -0.34, p < .05) a result that again replicates
Previous research has shown a low level of lay under-
a similar ?nding by Baron and Kemp (2004).
standing of the principle of comparative advantage, and
The key question for the present study is whether lack
that lower levels of understanding are associated with
of understanding of the principle of comparative advan-
greater protectionism (Baron & Kemp, 2004; Krugman,
tage explains why many people are particularly reluctant
1994). These results were also found here. However,
to import from low wage countries. People’s reluctance
Study 3 goes beyond this previous work by showing that
to import from low wage countries (i.e. averaging the
lower levels of misunderstanding are not associated with
rating over the two low wage scenarios) Pearson corre-
a particular reluctance to import from low-wage coun-
lated -0.29 (p < .05) with the Comparative Advantage
tries. Thus, this reluctance is not caused by misunder-
Test score. However, the equivalent correlation with peo-
standing of the principle.

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
342
A number of results indicate that people do not wish to
economists (e.g., Alston et al., 1992; Frey et al., 1984).
support exploitation in foreign countries, and that this re-
It is also worth noting that not all respondents preferred
sistance to foreign exploitation is connected with greater
to import from high-wage countries. In Studies 1 and
reluctance to import from low-wage countries. In Stud-
2, only a few respondents consistently favoured imports
ies 1 and 2, the respondents, whether Germans or New
from high over low wage countries or vice versa over all
Zealanders, favoured more restriction when the foreign
four combinations of the other scenario variables. In-
business owners made very good rather than some prof-
deed, in Study 3 a few respondents displayed a consistent
its. The studies also produced signi?cant interactions of
preference for importing from low-wage countries. Per-
the effects of wages and pro?ts: Both New Zealand and
haps, as suggested in the introduction, a few people did
German respondents were particularly reluctant to import
believe that workers in poor countries had more need to
from low-wage sources when the business owners made
sell their products abroad. In Studies 1 and 2 a number
very good pro?ts. The implication is that reluctance to
of respondents in the low-wage conditions did not con-
import from low-wage countries is in?uenced by people’s
sistently favour more restriction from one or other of the
perceptions of exploitation in such countries.
pro?t conditions over the two “where made” conditions,
At least for the New Zealand respondents of Study 1,
although there was a general tendency to restrict more
the perceived bene?t to one’s own country is affected by
from countries where the owners made very good pro?ts.
whether similar goods are made in one’s own country,
The overall lack of consistency may be a consequence of
while perceived bene?ts to the other country are mainly
the respondents not always entering the study with well
affected by the other country’s wage and pro?t levels.
thought out views about trade, and thinking through some
However, restriction scores were often affected by all
of the issues while actually completing the questionnaire.
three variables. Putting together these results suggests
It is also likely that when people think about import-
that people’s willingness to restrict imports depends not
ing from low-wage countries they consider a variety of
only on what is perceived as good for one’s own coun-
factors beyond those investigated here, for example, en-
try but also on what is perceived as good for the foreign
vironmental concern or unwillingness to support oppres-
country. Thus, attitudes to importing have at least some
sive foreign governments. The wording of the scenarios
component of international altruism, although whether
attempted to exclude one such factor — differences in
this international altruism is reasonably based is very
quality of the goods — but it is possible that this factor,
questionable. In many cases workers in low-wage coun-
as well as the others, still played some role in the respon-
tries may not have better alternatives.
dents’ thinking. On the other hand, it is not easy to see
Perhaps the most interesting ?nding from the research
how such factors could have affected the crucial results
— that there is particular reluctance to import from low-
found here, such as the interactive effect of pro?ts and
wage countries when the business owners make very
wages.
good pro?ts — was found for both the New Zealand
Taking a broad view, the present results show that the
and the German samples. However, the other results are
attitudes of ordinary people towards trade policy are not
somewhat different. There are several possible explana-
solely determined by their own personal pro?t and loss,
tions for the differences. Germany has a different trad-
or even solely by the pro?ts and losses of their compa-
ing pattern to New Zealand, and imports from different
triots. Their views are also in?uenced by their percep-
low-wage countries with different conditions. The results
tions of conditions in foreign countries. Speci?cally, the
from the German sample also indicate differences among
present results suggest that ordinary people are often re-
east Germans according to their experience of the former
sistant to importing goods from foreign businesses whose
communist regime.
owners make very good pro?ts, and that this resistance
The background variables investigated here (student
is particularly strong when the businesses are located in
versus non-student, age and sex) did not moderate either
low-wage countries.
the non-interaction between whether similar goods were
Lay opinions are not generally decisive in determin-
made at home and foreign wage levels or the interaction
ing government trade policies, but they do in?uence them
between foreign wage and pro?t levels. However, it is
(e.g., Davidson et al., 2006; Kemp, 2007). Indeed, their
possible that other variables might, or that the ?ndings
resistance may be an important obstacle to liberalising
will not hold in all countries.
trade generally and with low-wage countries in particu-
The results of the present studies feature a good deal
lar. The results presented here suggest that overcoming
of individual variation. Not only did all the studies fea-
such resistance may not always be best done by appealing
ture large standard deviations in the restriction and other
to people’s self-interest. Practically, the research shows
measures, but also every study contained a few respon-
that people’s resistance can be appreciably reduced if they
dents who supported unrestricted importation over all the
are assured that the owners of the businesses in the low-
different scenarios, as does the majority of professional
wage countries make some rather than very large prof-

---

Judgment and Decision Making, Vol. 3, No. 4, April 2008
Trade with low-wage countries
343
its. However, giving such an assurance may not always
Kemp, S. (2007). Psychology and opposition to free
be straightforward: In fact, investors in low-wage coun-
trade. World Trade Review, 6, 25–44.
tries often do require a higher rate of return because such
Kletzer, L. G. (1998). Job displacement. Journal of Eco-
countries are perceived as riskier and less creditworthy
nomic Perspectives, 12, 115–136.
(e.g., Depken et al., 2007; Ul Haque et al., 1998). Thus,
Krugman,
P. (1994).
Ricardo’s dif?cult idea.
the application of the present research is not completely
Downloaded
from
http://web.mit.edu/
straightforward.
krugman/www/ricardo.htm on June 20, 2006.
Krugman, P. R., & Obstfeld, M. (2000). International
economics: Trade and policy, 5th ed. Reading, MA:
References
Addison-Wesley.
Landier,
A.,
Thesmar,
D.,
&
Thoenig,
Alston, R. M., Kearl, J. R., & Vaughan, M. B. (1992).
M.
(2008).
Investigating
the
origins
of
Is there a consensus among economists in the 1990’s?
capitalism-aversion.
Downloaded
from
American Economic Review, 82, 203–209.
http://www.unige.ch/ses/ecopo/staff/
Baron, J. (2001). Confusion of group interest and self-
thoenig/EP3.pdf on 14 March, 2008.
interest in parochial cooperation on behalf of a group.
Lucas, R. E., Clark, A. E., Georgellis, Y., & Diener, E.
Journal of Con?ict Resolution, 45, 283–296.
(2004). Unemployment alters the set point for life sat-
Baron, J., & Kemp, S. (2004). Support for trade restric-
isfaction. Psychological Science, 15, 8–13.
tions, attitudes, and understanding of comparative ad-
Mayda, A. M., & Rodrik, D. (2005). Why are some peo-
vantage. Journal of Economic Psychology, 25, 565–
ple (and countries) more protectionist than others? Eu-
580.
ropean Economic Review, 49, 1393–1430.
Corden, W. M. (1974). Trade policy and economic wel-
Mayer, W. (1984). Endogenous tariff formation. Ameri-
fare. Oxford: Clarendon.
can Economic Review, 74, 970–985.
Davidson, C., Matusz, S., & Nelson, D. (2006). Fairness
Olson, M. (1982). The rise and decline of nations: Eco-
and the political economy of trade. The World Econ-
nomic growth, stag?ation, and social rigidities. New
omy, 29, 989–1004.
Haven: Yale University Press.
Depken, C. A., LaFountain, C., & Butters, R. (2007).
Ricardo, D. (1971). On the principles of political econ-
Corruption and Creditworthiness:
Evidence from
omy, and taxation. Harmondsworth: Penguin. (Origi-
Sovereign Credit Ratings.
Available at SSRN:
nally published in 1817.)
http://ssrn.com/abstract=899414.
Samuelson, P. A. (1972). The way of an economist.
Frey, B. S., Pommerehne, W. W., Schneider, F., &
Reprinted in (R. C. Merton, ed.) The collected papers
Gilbert, G. (1984). Consensus and dissention among
of Paul A. Samuelson. Cambridge: Cambridge MIT
economists: An empirical inquiry. American Eco-
Press.
nomic Review, 74, 986–994.
Scheve, K. F., & Slaughter, M. J. (2001). What deter-
Gomes, L. (2003). The economics and ideology of free
mines individual trade-policy preferences? Journal of
trade: A historical review. Cheltenham, UK: Edward
International Economics, 54, 267–292.
Elgar.
Schwartz-Shea, P., & Simmons, R. T. (1991). Egoism,
Hainmueller, J., & Hiscox, M. J. (2006). Learning to love
parochialism, and universalism. Rationality and Soci-
globalization: Education and individual attitudes to-
ety, 3, 106–132.
ward international trade. International Organization,
Tre?er, D. (2001). The long and short of the Canada-U.S.
60, 469–498.
Free Trade Agreement. National Bureau of Economic
Hiscox, M. J. (2006). Through a glass and darkly: Atti-
Research Working Paper 8293.
tudes toward international trade and the curious effects
Ul Haque, N., Nelson M. C., & Mathieson, D. J.
of issue framing. International Organization, 60, 755–
(1998). The relative importance of political and eco-
780.
nomic variables in creditworthiness ratings.
IMF
Irwin, D. A. (2002). Free trade under ?re. Second Edi-
Working Paper No.
98/46 Available at SSRN:
tion. Princeton: Princeton University Press.
http://ssrn.com/abstract=882300

---