Sie sind auf Seite 1von 24

American Association for Public Opinion Research

Effects of Internet Commerce on Social Trust


Author(s): Diana C. Mutz
Source: The Public Opinion Quarterly, Vol. 73, No. 3 (Fall, 2009), pp. 439-461
Published by: Oxford University Press on behalf of the American Association for Public
Opinion Research
Stable URL: http://www.jstor.org/stable/40467612
Accessed: 26-01-2017 12:35 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

American Association for Public Opinion Research, Oxford University Press are
collaborating with JSTOR to digitize, preserve and extend access to The Public Opinion Quarterly

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Public Opinion Quarterly, Vol. 73, No. 3, Fall 2009, pp. 439-461

EFFECTS OF INTERNET COMMERCE ON SOCIAL


TRUST

DIANA C. MUTZ

Abstract As of the early 21st century, one of the most popular uses of
the internet is for online shopping. In this study I examine how online
purchasing affects levels of generalized social trust, a quality widely be-
lieved to be central to the health and well-being of contemporary societies.
Drawing on two original studies, including an experiment embedded in a
representative national survey, and a hybrid laboratory/field experiment,
I find consistent evidence that positive e-commerce experiences promote
generalized social trust. I discuss the implications of these findings for the
role of business in helping to maintain attitudes supportive of democracy.

E-commerce has been expanding at an impressive rate; as of 2001 well over


100 million Americans had purchased a product online (Intermarket Group
2001). As of 2007, 51.1 percent of adult internet users reported that they
bought products and services online (Direct Marketing Association 2007), and
total online sales are expected to hit 259 billion dollars (Forrester Research
2007). At the same time, the increasing popularity of the internet has prompted
concerns about the impact this new technology has on social life, and on levels
of social trust in particular.
In this study I hypothesize that participation in e-commerce stimulates higher
levels of social trust. When people purchase goods from others who are not
known in a face-to-face context, and those others turn out to be trustworthy,
this raises people's baseline expectation about the trustworthiness of others. If
people increasingly experience successful commercial interactions via internet
with people in faraway places, then increasing e-commerce will have positive
implications for generalized social trust.

diana c. mutz is the Samuel A. Stouffer Professor of Political Science and Communication at the
University of Pennsylvania, 3620 Walnut Street, Philadelphia, PA 19104, USA. The author would
like to thank the online car dealer whose timely delivery of a vehicle purchased sight unseen led
the author to higher levels of social trust, and to bigger and better online acquisitions. The author
would also like to thank Shiloh Krieger and Danielle Dougherty for assisting in the collection of
data for Study 2, and the Institute for the Study of Citizens and Politics for funding portions of this
study. Address correspondence to Diana C. Mutz; e-mail: Mutz@sas.upenn.edu.

doi:10.1093/poq/nfp042 Advance Access publication August 16, 2009


The Author 2009.
Published by Oxford University Press on behalf of the American Association for Public Opinion Research.
All rights reserved. For permissions, please e-mail: joumals.permissions@oxfordjournals.org

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
440 Mutz

I conduct t
methodolog
and the gen
iments with
an importan

Social Tr
Social trust
(e.g., Putna
for econom
which are e
to pay a th
data, econo
of w people
capita outp
emphasis in
200 1 ). Soci
time and sp
1979; Gidd
The kind o
generalized
personally.
amount of r
Interactions
But when a
empirical ba
workers, cl
social trust
trust is for
will trust e
on this def
the skids f
more likely
default assu
generally co
and Miller
allowing fo
Likewise, f
of e-comme

1 . Social trus
another, but sc
social trust, w

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 441

of social trust go along with earlier and more vigorous adoption of e-comm
As an article in The Economist (2002) summarized, "If you like surfing t
it is probably because you believe people are basically good." In an aggr
countries high in social trust are also high in internet penetration (Huan
2003). Likewise, individuals who use the internet and buy things online
to be higher in social trust than those who do not (Pierce and Lovrich 2
But because of the observational nature of these studies, it is diffic
rule out alternative explanations for these associations. Some experim
evidence has corroborated the causal link in one direction - that trust facilit
participation in e-commerce (e.g., Mutz 2005), but this still leaves ope
possibility of a circular relationship. In other words, although more tr
people may, as a consequence, be more likely to participate in e-commer
does not eliminate the possibility that successful participation in e-com
also increases levels of generalized social trust.
Without questioning the causal importance of social trust in bringing
economic engagement of various kinds, this study examines whether t
verse causal direction also occurs; that is, does online purchasing incr
levels of generalized social trust for those who participate in successful
transactions? I turn next to theory and evidence supporting this possibilit

Why Causal Effects from E-Commerce?


To date, little is known about what makes some people and some socie
more trusting than others. Empirical evidence suggests that telephon
transportation infrastructure raise trust, most likely by decreasing the
to which people in disparate geographic locations feel far apart (Knac
Zak 2003). Even though people do not personally know all or even ma
the people they can communicate with in other locations, trust incre
"effective social distance declines" (Knack and Zak 2003, p. 103). As pe
experience a wider radius of face-to-face contacts, encompassing memb
other groups, ethnicities, and so forth, those outside of the realm of face
contact no longer seem as threatening. Consistent with this theory, intern
appears to encourage "weak ties" in particular, that is, a greater num
nonintimate relationships with people who are dissimilar from the user
2000; Robinson et al. 2000; Hampton and Wellman 2003). Likewise, thr
the internet, people do business with people who are likely to be quite dif
from themselves (e.g., Etzioni and Etzioni 1997).
As with other forms of communication, perceptions of social distanc
decline as a result of internet-based exchanges and the many commercia
actions it encourages with distant others. But we know little if anythin
how e-commerce participation in particular affects social trust. In rese
date, trust has come up primarily in the marketing literature as a que
how to cultivate potential consumers' trust in a website in order to en

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
442 Mutz

purchasing
not the cen
ultimately
No one dou
line, partic
quality of t
promised, a
compromis
voluntarily
locations.
Whereas the days of strictly face-to-face commerce allowed one to pick and
choose with whom to do business based on personal relationships, supra-local
commerce requires a greater leap of faith because a person can no longer march
down to Main Street if he or she is unhappy with a purchase. Low expectations,
combined with positive outcomes, should combine to make e-commerce a
positive force for social trust if people can be induced to take part. When
negative experiences with e-commerce occur, they should have precisely the
opposite effect and lower generalized social trust.
Are expectations really so different for online versus "bricks and mortar"
businesses? Do the conditions outlined above genuinely exist in the contempo-
rary American culture (low expectations among nonusers and generally positive
results)? When a representative sample of Americans was asked parallel ques-
tions about businesses in their local area and online businesses, over 60 percent
of respondents reported that online businesses would be "very likely" or "some-
what likely" to try to cheat them, whereas only 21 percent say the same of local
businesses. Moreover, if businesses did try to cheat a consumer, 46 percent
of respondents thought local businesses would be "very likely" or "somewhat
likely" to get away with it, compared to 74 percent of people who thought the
same of e-businesses. Clearly, e-businesses are viewed as less trustworthy than
bricks and mortar ones, and as less accountable as well. Moreover, people who
have never tried buying online are less trusting than those who have2 (see also
Fogg et al. 2002; Horrigan and Rainie 2002).
For enhanced levels of social trust to result, these low expectations must
combine with positive outcomes. Are online shoppers generally satisfied with
their experiences? According to consumer research studies, overall customer
satisfaction with online retail shopping is extremely high (see Austin 2008).
A representative survey done at the same time these experimental data were
collected likewise confirmed that more than 80 percent of those who have
purchased online reported positive or very positive experiences, as opposed to
less than 5 percent who report very or somewhat negative experiences.3

2. These figures are from the survey using a representative national sample that accompanied
Study 1 (n = 1,326), and they are similar to estimates from other representative studies.
3. These figures are from Study 1, and they are similar to or slightly lower than satisfaction levels
reported by the others cited.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 443

Overall, then, internet commerce should have a net positive effect o


eralized social trust. Because risk-taking is central to how people gain
internet transactions have the potential to foster greater social trust whe
are successful transactions. Importantly, initially low expectations (low tru
e-commerce, combined with a favorable outcome, produces a greater B
shift than a transaction with high expectations (high trust) and the same
able outcome.
Interestingly, the transition from face-to-face commerce to purchasing by
cataloge probably required a similar leap of faith that only slowly came about
through a similar circular process. Historians have documented the initial lack
of trust in distant, unseen cataloge companies such as Montgomery-Ward, as
well as the delight and surprise experienced by those who were induced to try
this new shopping method, and found these remote vendors to be trustworthy
(see Boorstin 1974; Arena 1996). Such interactions - those perceived as high
risk- are precisely the kind with the potential to increase generalized social
trust.

Micro-Level Mechanisms

At the micro-level, there are multiple mechanisms explaining why online buy-
ing should lead to enhanced social trust. For the sake of example, assume that
a first-time buyer is relatively cautious, and unlikely to take risks, particularly
of the monetary kind. This person is nonetheless induced to buy online - for
reasons of lower prices, better selection, or perhaps products that simply are
unavailable locally. Although a high level of trust in humankind was not her
reason for choosing to buy online rather than from a bricks and mortar business,
having sent the money into the ether, she now must rationalize her own behav-
ior: if she is the kind of person who would do this, buying online is probably
not so risky as it seemed. In other words, if people can be induced to engage
in a counter-attitudinal behavior for reasons other than already high levels of
trust, their attitudes are likely to follow, so long as the inducement is not too
heavy handed (e.g., Festinger and Carlsmith 1959; Bern 1967), and particularly
when the risk-taking is ultimately rewarded. Thus people may shift their atti-
tudes toward the trustworthiness of unknown others so that these attitudes are
consistent with their demonstrated willingness to purchase online.
Another possibility is that positive or negative internet purchase experiences
simply serve as salient exemplars influencing social judgment, and thus social
trust. Social judgment depends upon exemplars that are readily available in
people's minds. When asked if people are generally trustworthy, it seems natural
to be influenced by recent salient exemplars, in this case based on internet
purchases that worked out surprisingly well or poorly. Although exemplification
effects are well documented as influences on attitudes toward many specific
groups (see, e.g., Bodenhausen et al. 1995; Zillman and Brosius 2000), their

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
444 Mutz

impact on j
But to the e
turns out t
positive exe
Interesting
known entit
first-time o
a coupon an
But then -
with their
buying exp
increases. P
ideal contex
nonetheless
increase.
Below I test the hypothesis that successful online transactions will increase
social trust using two experimental studies designed for this purpose. I begin
using an experiment designed to manipulate people's intent to buy online. Study
1 has the advantage of a large sample size, and of relying on an experimental
design embedded in a representative national survey. Although the generaliz-
ability of this sample population is quite advantageous, it unfortunately limits
the strength of the experimental treatment and thus the generalizability of these
findings to actual online purchasing experiences.
In Study 2, 1 employ a combination laboratory/field experiment in which I
induce those who have never before made an online purchase before to do so for
the first time. I systematically alter the quality of their purchase experience, and
then evaluate its effects on subjects' levels of generalized social trust relative
to a control group. To the extent that good business turns out to also be good
for social trust, these studies can go a long way toward establishing cause and
effect evidence of this mechanism.

Study 1: The Survey Experiment


In the survey experiment, a national probability sample of just over 780 respon-
dents was randomly assigned to one of three different versions of a question-
naire, so that each version was administered to roughly 260 people. Knowledge
Networks of Menlo Park, CA, conducted the survey interviews with their
representative panel via either personal computer or WebTV (for homes without

4. In the representative sample used in Study 1 , the most common reasons cited by those who had
already participated in e-commerce were lower prices (39 percent), convenience (37 percent), and
curiosity to try shopping a new way (28 percent). Interestingly, only 19 percent said they were
motivated to shop online for the first time because it was a website for an e-business that was
already a well-known, reliable business offline as well.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 445

internet access).5 This aspect of the sample eliminated potential selecti


effects; all respondents had internet access, though they had varying degrees o
experience and comfort using it.6
Ideally it would be possible to randomly assign respondents to differen
online purchasing experiences, but this is obviously impractical in the conte
of a survey experiment. Instead, I manipulated respondents' perceptions of wha
it would be like to purchase via the internet in order to evaluate the implicatio
of successful and unsuccessful e-transactions for levels of generalized soc
trust. Three experimental conditions included one in which buying via intern
was encouraged, one in which buying was discouraged, and a third, contr
condition. My goal was to manipulate intent to purchase online, even thoug
actual purchases had not yet been made. To do so, I encouraged respondents
think exclusively about the reasons they would want to make purchases onlin
or about the reasons that they would not. In order to avoid confounding, no
of the treatments mentioned anything dealing directly with trust in intern
commerce or trust in people more generally.
Those in the Positive Internet Purchase condition were given a series of
questions that included information on the fact that "a record number of peop
did some or all of their holiday shopping on the web this year, and surve
show that the overwhelming majority of them found shopping online to b
very enjoyable experience." The treatment thus urged people to think abou
variety of positive aspects of internet shopping, including convenience, low
prices, easier comparison shopping, and better product selection. Responden
in this condition were also given information from a recent survey suggesti

5. The overall response rate, including all stages involving potential attrition and nonresponse, t
into account (1) the panel recruitment response rate (44.2 percent); (2) the household profile r
(65.3 percent); (3) the household retention rate (40.5 percent); and (4) the survey completion r
(92.7 percent). Thus, the overall response rate combining all four figures above was 10.8 perce
(see "s," AAPOR 2008, pp. 39-40). The response rate excluding the impact of panel attrition, th
is, treating nonresponse bias as ignorable from this source, was 26.8 percent. Nonresponse bia
due to panel attrition has been estimated at 1 percentage point over 323 social, political, value
and consumer behavior measures. For additional papers addressing data quality using Knowled
Networks' Internet panel methodology, see Krosnick and Chang 2001; Wiebe, Eyerman, and Lo
2001; and Chatt and Dennis 2003.
6. Although this respondent panel constitutes a representative probability sample, unlike a rand
sample that might be recruited and interviewed by telephone, these respondents either already
access to the internet via personal computer at home or work, or they were given some access
the internet via WebTV, which was provided to them in exchange for their survey participatio
Thus, all of these respondents could, if they so desired, access the internet in order to m
consumer purchases. The usual roadblocks to shopping online should be reduced in this sample
so participation in online commerce is more purely a matter of personal choice rather than acce
Because WebTV is far more difficult to use for purposes of entering text, however, the Web
respondents are not regular internet users in the same sense as those with personal computers.
second advantage to using this particular pool of respondents is that additional information ab
them is known as part of a previous interview. All respondents had completed a computer us
profile within the past two months, thus providing information on their previous online experienc
and purchasing habits.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
446 Mutz

that "more
online shop
Respondent
received a stimulus that was the reverse of the one described above. Instead of
hearing about all of the advantages of shopping via internet, they were instead
reminded of the drawbacks of purchasing in this fashion. These included ship-
ping charges, the costliness of returning things that did not fit, slow delivery
times, difficulty in determining the details or quality of a given product without
seeing it in person, technical problems when submitting orders, and the temp-
tation to make impulse purchases when all it takes to purchase is a click of
the mouse. The treatment did not address fraud or issues related to trust in any
way. No mention was made of problems with e-commerce that involve trust,
such as credit card theft or misuse of private information.7
Generalized social trust, the key dependent variable, was assessed using a
battery of six questions. These included the four conventional, most commonly
used social trust items across the social sciences, as well as two additional
variations of them that were added to produce a more reliable index (see
Appendix B). To assess whether experimental manipulations of intent to buy
online were successful, four items in the posttest addressed the individual's
likelihood of purchasing online during the coming year. These questions were
combined into an index producing a Cronbach's alpha of .82 (see Appendix B
for details).

Results of Study 1
Consistent with previous studies, these data also show that those who have
already bought online are significantly higher in trust than those who have

7. Based on information already known about the respondents from an internet use battery filled
out two months prior, those randomly assigned to the Positive Internet Purchase condition who
were known to have had a positive experience buying online were asked if they had had such an
experience. The point of doing so was to prime their positive thoughts about internet purchasing,
knowing that the answer would be yes. If a respondent was known not to have had such an
experience, he or she was asked if anyone he/she knew had had a positive experience buying things
online. Finally, to urge respondents in the Positive Internet Purchase condition to think even more
concretely about buying via internet, each was asked to list two items they might buy via internet
in the next few months, if they were to buy something via internet.
If they were assigned to the Negative Internet Purchase Condition, respondents who were already
known to have had a negative experience shopping online from a survey two months prior were
asked if they had ever had "a negative experience of some kind when buying things on the internet."
If they had had no experience shopping online or no negative experience, they were asked instead
if anyone they knew had had a negative experience buying things online. Finally, they were asked
to reflect on what they had heard about the reasons people dislike shopping online. As with the
Positive Internet Purchase condition, the point of this series of questions was not so much to obtain
answers (since these were already known) as to give respondents cause to mentally rehearse the
negative aspects of internet shopping if they were assigned to the negative treatment.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 447

never bought online (t = 3.42, p < .01).8 Of course, with survey data
cannot tell whether they are more likely to have bought online as a result
being more trusting, or whether they are more trusting as a result of success
online purchasing experiences. For purposes of examining this central quest
we turn to the experimental analyses.
To compare means across the three experimental groups, I conducted a o
way analysis of variance with planned contrasts between the mean of the cont
group and that of each of the experimental groups. To increase the efficien
of the model, I included a single covariate that was strongly predictive of
dependent variable, but uncorrelated with the treatments, in order to red
the within-group variance (see Franklin 1991). In the model testing mean
differences in intent to purchase online in the coming year, I included p
online purchasing experience in the model as a covariate. Likewise, in the
model testing mean differences in social trust, I included interpersonal tr
(see Appendix B for details).
To summarize, my expectations were that (1) the positive and negative tre
ments would alter intent to purchase online in the future, and, more importan
that (2) stimulating people's desire to buy things online would cause resp
dents in the Positive Intent to Buy condition to be more likely to expres
generalized trust in others, and those in the Negative Intent to Buy condi
should be less likely to express generalized trust than the Control condition
As a manipulation check, intent to purchase online was examined acros
the three randomly assigned experimental conditions. As shown in figure
intent to purchase online was successfully manipulated by these treatme
(F = 14.85, p < .001). The means are in the expected directions with
least intent to purchase online in the Negative Internet Purchase condition,
greatest in the Positive Internet Purchase condition, and the Control condit
in between. Planned contrasts between each of the experimental conditio
and the control confirmed that the Negative Internet Purchase condition
significantly lower in Intent to Purchase than in the Control condition (p < .0
and the Positive Internet Purchase condition was significantly greater in In
to Purchase Online than in the Control condition (p < .01). The manipulatio
clearly established differing levels of intent to participate in e-commerce.
I next compared the mean levels of generalized trust by experimental con
tions. As shown in figure 2, consistent with my hypothesis, the Positive Inter
Purchase condition produced the highest level of social trust, and the Nega
Internet Purchase condition produced the lowest level of social trust. An a
ysis of variance confirmed that these three means were significantly differ
from one another (F = 3.96, p < .05), although the findings appear to be dri
primarily by the Positive condition, whose mean was significantly differe
from the Control condition (p < .05), while the Negative condition was not

8. The standardized mean among those who had never purchased online was -.05, whereas it
.06 for those who had already purchased online.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
448 Mutz

0.10 -i

0.05 - .

I 0.00

O Positive Internet Control _____ Condition Negative Internet


<d _o 05 - Purchase _____ Purchase

I _o.1o
I --15
- -0.20

n oc

Figure 1. M
tent to Pur
NOTE. - An
rience dem
experimen
condition w
positive con
(SD = .87), w

0.10 -j

0.08

0.06

I 0.04

| 0.02

I ooo

5 Positive Internet Purchase Control Condition Negative Internet Purchase

i 5 -002 Positive Internet Purchase Control __| Condition Negative m Internet


-0.04
-0.06 ^^^"
4)08

Figure 2. Ef
(Study 1).
NOTE. - Analysis of variance controlling for interpersonal trust demonstrated
significant differences in generalized trust by experimental conditions (F =
3.96, p < .05, n = 781). Contrasts with the Control condition were significant
for the positive condition (p < .05) but not for the negative condition. The grand
mean across all conditions was -.01 (SD = .72), with index scores ranging
from -2.13 to 1.35.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 449

Further analyses confirmed that the manipulation of intent to buy o


was responsible for the differences observed in the dependent variable
experimental conditions. When the manipulation check variable for int
buy online was included in the model, it completely accounted for the sig
impact of the experimental manipulation on generalized social trust, confi
that it was manipulation of intent to buy that altered social trust.
Thus the results of Study 1 are highly consistent with the theory that ec
transactions online may have the potential to change levels of generalize
trust. Nonetheless, the survey-experimental method substantially cons
the level of realism in this study because subjects were not able to actually
chase anything online, although the manipulation check verified that their
to purchase online changed significantly. In addition, given the nature
sample- all were already online to some extent- this study may underes
the extent to which successful e-commerce transactions shape attitudes
generalized others. The fact that changing people's perceptions of e-com
transactions has implications for levels of social trust is highly suggesti
it does not speak directly to whether actual purchase experience matter
the world of e-commerce increase (or decrease) levels of social trust by
consumers positive (or negative) purchase experiences?

Study 2: The Hybrid Laboratory/Field Experiment


Ideally one would draw conclusions from a more true-to-life scenario,
ably one that also afforded experimental control. The second study was de
toward this end. The goal was to experimentally manipulate individuals
first experience with an e-commerce transaction in as naturalistic a
as possible. The very first such positive or negative experience should
particularly clear impact on perceptions of the trustworthiness of other
the lack of other experiences with e-commerce. To evaluate the implica
successful and unsuccessful internet purchases on levels of generalized
trust, I designed a single factor, three-condition experiment similar to St
except that the study began in a laboratory where the experimental tre
was initiated, and was completed weeks later when the same participan
contacted by telephone in their homes. This repeated-measures/within-s
design made powerful analyses possible even with a relatively small sam
respondents.9 If anything, small samples should bias results toward a l
statistical significance.

9. The difference in the statistical power of a within-subjects design relative to a between


design of the same sample size is striking because of the elimination of within-group v
in the denominator of the F statistic. Each respondent is compared to his or her own pe
mean at a previous point in time, rather than to an aggregate mean. Because within-gro
is eliminated, much smaller differences in means are statistically significant when usin
powerful within-subject designs.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
450 Mutz

Potential p
showed up t
this study w
nity groups
for the unr
as well as o
before. To
thing via in
possibility
experiences
After comp
a token of a
CD of their
of co three
shopping ex
were given
which one
the unrelat
and negativ
coupon goo
Participant
they could
In reality,
this study.1
any other
music, sear
check out,
required, on
when an or
research ass

10. This exper


screen.

11. At the time these data were collected, roughly 33 percent of the U.S. public had e
purchase online. Of households with computer and access to the internet, 53 percent ha
online purchase. This study was conducted from June 21, 04 to March 14, 05.
12. In addition, to satisfy human subject requirements, respondents had to indicate wh
the initial informed consent form that they would not mind being contacted again at s
point, and provided a phone number.
13. They were not told until after the study was over that this gift was part of a stud
completely separate from the initial one. In this way, I avoided calling undue atten
experimental treatments and subsequent measures.
14. In order to facilitate delivery of this large inventory directly from the lab, the U
bookstore allowed us to purchase a small number of each CD and then return the ones
not chosen by any of our participants.
15. The company was identified as a Canadian enterprise to ensure that whatever effect
trust would be seen as generalizable beyond the borders of the United States.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 451

in the positive condition received their free gift in the mail promptly (arou
days later), along with a friendly "Thanks for your Business" receipt fe
the company name and logo. Those in the negative condition instead rec
CD that appeared to be broken and blank instead of the CD they had req
The accompanying receipt instructed customers to contact the manufa
rather than the company about problems or defects in the product, but di
offer any means for doing so.
Because measures of social trust were available for these participants
the laboratory study done before the unrelated study that was used as t
for subject recruitment, it was possible to use follow-up phone calls to
ment a repeated measures experimental design to assess the impact of
purchase experiences on pre- to posttest changes in social trust. All part
in this study received a follow-up phone call roughly two to four week
participating in the experiment. The interviewers identified themselves
ing from Ohio State University, a different university from the location
laboratory study.16 During the follow-up phone call, participants were
the same battery of social trust questions used in the pretest questionn
the unrelated study. As a manipulation check, at the very end of the s
interview, respondents were asked about whether they had recently rec
CD in the mail (for those in the treatment conditions) or from a lab as
(for those in the Control condition), in appreciation of their participati
recent University of Pennsylvania study. All respondents confirmed re
the CDs, whether in person or by mail.
The complex, multistage nature of this study made it difficult to obtain
numbers of participants. Only those who had never purchased online
were eligible to begin with, and then only those who volunteered their
number for follow-up on the initial study could potentially be recon
All together, 25 participants qualified for the study, and 22 of them
successfully recontacted by telephone.17 On the one hand, this small
size stacked the deck against statistically significant findings. If an ef
apparent despite the limited sample size, it would need to be an extre
large and robust effect if this were a between-subjects design. On th
hand, the powerful nature of the within-subjects design used here was
of identifying effects that would require a much larger sample in a be
subjects design.18

16. The Center for Survey Research at Ohio State University conducted the posttest tele
interviews based on phone numbers provided by the investigator.
17. The subjects were roughly evenly divided by conditions, with ns of 7, 7, and 8. Beca
were screened into the study only if they had never before purchased online, the sample w
in education and income than the population at large (mean age 36; median education com
vocation school; median income $30,000, 65 percent female).
18. Moreover, whether between-subjects or within-subjects in design, smaller experimen
ples provide more conservative tests of the hypothesis.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
452 Mutz

Results o
InStudy 2,
ences in eit
would cause
likely to ex
Purchase co
statistical s
of variance
the key dep
there a is s
of social tr
experiment
change in t
Purchase co
Control con
In order to
trust) and t
a mixed-mo
being comp
differences
assignment
ferencesbe
for extensive control variables.
The analysis of variance demonstrated no main effects on trust from pre to
post (F = .48, p > .20), and no significant main effects by the experimental
condition (F = .47, p > .30). However, as predicted, there was a significant
interaction between pre- to post-trust and experimental treatments, suggesting
that the change occurred differentially by the experimental condition (F = 2.77,
p < .05).
As shown in figure 3, the differences by condition are quite logical. Respon-
dents in the Control condition changed the least in their extent of social trust
(from 1.48 in the pretest to 1.52 in the posttest). Respondents in the Positive
Internet Purchase condition became more trusting (from 1.44 in the pretest
to 1.56 in the posttest) and those in the Negative Internet Purchase condition
showed the largest shift of all, in the direction of less generalized social trust
than the control (from 1.71 in the pretest, down to 1.57 in the posttest). Those
who encountered a negative experience online became less trusting as a result,
and those who had had a positive one became more trusting.

Discussion

In interpreting the real world significance of these findings, it is worth remem-


bering that these respondents were exposed to only a single incident of online

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 453

0.15

0.10 i
"55

jo ~ 0.05 : rr^.BomM.
g . o.oo

U =? Positi
I I -005 - |HH[
uj o -0.10 ^^^^H

O -0-15

-0.20

Figure 3. Effects of Internet Purchase Experience on Pretest-Posttest C


in Generalized Social Trust (Study 2).
Note. - Repeated measures analysis of variance was used to test the hy
esis of differential pre- to posttest change in trust based on the experim
condition. There was no significant pre- to posttest change across all condi
(F = .48, p = .50, n = 22). But as hypothesized, there was a significant in
tion between the experimental condition and pre- to posttest change (F
p < .05, n = 22), indicating differential pre- to posttest change in trust
on experimental treatments.

buying or the lack thereof. In one sense, this is a surprisingly weak st


to produce significant effects, particularly with a small sample. The fact t
single positive or negative online experience registered significant effect
this period of time is impressive, and suggests a highly robust impact.
effects of the treatment had to compete with other experiences and infor
that might reach these respondents in the course of everyday life durin
time period, in order to produce a signal strong enough to register as statis
significant.
At the same time, because these were first-time buyers, they were probably
more likely to respond to the information than would a person for whom this
was just one of many online purchase experiences. For a first-timer buyer, the
effect on social trust is undiluted by other online experiences. Trust increases
when a person takes an action that is perceived as potentially risky, and it works
out beneficially. Once online buying is widely accepted and no longer perceived
as unusually risky, participation in e-commerce is unlikely to encourage higher
levels of social trust.19

19. Indeed, if one looks at other impersonal forms of commerce that are no longer perceived as
risky, such as buying via catalogs, the survey data in Study 1 demonstrates that there is no longer
a relationship between generalized social trust and participation in this form of commerce based
on measures in the survey experiment. In the beginning, however, catalog companies had to go to
great lengths to court the trust of potential customers (see Boorstin 1974).

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
454 Mutz

Over time,
pattern. In
relatively f
form of ec
informative
States, mos
be reinforc
these effec
make new p
as a whole c
catalog com
In this sens
commerce
should be a
medium.
In the past, the assumption has been that social trust was related to e-
commerce participation solely because social trust was a prerequisite for being
willing to participate in the internet economy. However, the results of this study
suggest that participation in e-commerce that is initiated for other reasons, can
also lead to higher levels of generalized trust. Because most of the literature
on this topic comes from cross-sectional data, most previous findings are also
consistent with the reverse possibility observed here: that participation in e-
commerce fosters higher levels of generalized trust.
But what about studies suggesting that internet use erodes social trust? In
the case of e-commerce, the argument would be that if a person no longer
has to visit the local store, and interact with other human beings in order to
obtain the products and services that he or she needs, then social relationships
and social trust will suffer as a result. However, importantly, the loss in these
examples is of face-to-face contact and social interaction, and not necessarily
of generalized social trust. Studies suggesting that internet use may substitute
for face-to-face contact have not found the anticipated negative relationship
(e.g., cf. Cole 2000; Robinson et al. 2000; Nie and Erbring 2002; Nie, Hillygus
and Erbring 2002). Moreover, these studies are of internet use more generally,
and not of participation in e-commerce in particular.
The evidence in this study combines with that of other researchers to strongly
suggest a reciprocal virtuous circle; when e-commerce works as it should,
buying online helps foster trust in those one does not know and has no reason
to trust. But high levels of social trust - most likely from other causes - also
facilitate participation in e-commerce; if one believes in the basic goodness of
others, then it will seem less risky to buy online. To the extent that reciprocal
causation is involved in this relationship, previous studies of this cross-sectional
relationship have probably overestimated how much social trust influences
online buying due to the endogeneity of cause and effect (see Huang et al.
2003).

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 455

It is worth noting that based on this theory, while successful internet tr


tions may have the capacity to foster a virtuous circle of social trust and
commerce, negative experiences should do just the opposite. One becom
trusting as a result of negative purchase experiences. Fortunately, most en
ters with generalized others via e-commerce tend to be positive experi
but the potential for negative influence is equally present.
Perhaps more importantly, these results suggest that there are alter
avenues to increasing social trust, ones in which the American busines
munity may play an important role. And at the risk of sounding glib, i
that rather than joining a bowling league or the garden club, perhaps
cans can stay home and shop. By engaging in economic transactions wit
we do not know and probably will never meet, we enhance our faith
general goodness of others, at least to the extent that those experiences
predominantly positive ones.
Of course, businesses may act reliably because it is in their long-te
economic self-interest to do so, rather than out of a sense of moral co
regarding social trust. Their questionable motives do not appear to m
Profitable business practices are also good for society as a whole.
perceive their treatment as customers to be information about the deg
trustworthiness they can expect from distant, impersonal others. Thu
business practices have important ramifications for the long-term well-be
societies, not just the companies' bottom lines.
Social scientists interested in the internet's impact on social trust ha
cused on the more obviously "social" kinds of internet activities, such a
or participation in chat rooms, at the expense of the most obvious pu
for which the internet is commonly used: to buy things. In reality, the in
is a heavily commercial medium. Most of the top 100 most frequently
websites are dedicated to commerce and entertainment (see Gandy 2002
leading some scholars to conclude that the internet is "coming to resem
realm of shopping, play, entertainment, and little else" (Barber, Matts
Peterson 1997, p. 38). For some scholars, the overwhelmingly comme
nature of the internet suggests that its prosocial promise remains unfu
But this perception overlooks the unintended consequences of e-comme
participating in e-commerce, consumers gain impressions about the tr
thiness of people beyond the realm of their personal acquaintances.
Of course, e-commerce experiences are only one of many forms of in
tion that people may use to make judgments about social trust. In the agg
social trust in the United States has changed very little during the five
when it has been tracked, and responses to these kinds of items have
shown to be highly context sensitive (see Smith 1997). If people are fir
minded of crime or divorce rates, for example, it is not surprising i
levels of trust are expressed. Overall, this evidence lends support to th
that salient exemplars may exercise significant influence on social tru
ments. To the extent that e-commerce experiences serve as pleasant ex

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
456 Mutz

of the trust
forces may
overly optim
internet alon

Appendix
of Intent to Purchase

POSITIVE INTERNET PURCHASE TREATMENT

A record number of people did some or all of their holiday shopping o


web this year, and surveys show that the overwhelming majority of them
shopping online to be a very enjoyable experience. People who were ne
the idea of buying things via the internet noted how much more conveni
was than driving to their local stores, and how much lower the prices we
the very same products online. Lower prices, combined with many comp
offering free shipping, made shopping via internet a real bargain. They
noted that it was far easier comparison shopping on the internet. Shopper
appreciated the better selection of products offered online relative to their
stores.

Amazingly, surveys show that more than 99 percent of internet shoppers report
that they are happy with their online shopping experiences and said they would
do even more of their shopping online in the future. Have you read or heard
anything about this recently? (Yes/No)

Rs receive either Question A or B:

A. [IF KNOWN TO HAVE "Somewhat positive" or "Very positive" experiences


shopping online]: Have you ever had a positive experience buying things on
the internet?

B. [IF NO EXPERIENCE OR NO KNOWN POSITIVE EXPERIENCE]: Has


anyone you know had a positive experience buying things on the internet?

Think for a minute about what you have heard other people say about why
they like shopping on the internet - this could be information from people you
know or from newspapers, magazines, or television. Do people like purchasing
online because of

Yes No

The prompt and polite service


Avoiding crowds
Open 24 hours a day

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 457

Better prices
No need for transportation to stores
Better selection of products
Easier comparison shopping
Less time-consuming

If you were to buy something on the internet in the next few months, what kind
of product is it most likely to be? Please name at least two possible items.

NEGATIVE INTERNET PURCHASE TREATMENT

As you might have heard, buying things through the internet has not b
popular as quickly as many had predicted. Shipping charges often make
chasing online more expensive than the same product purchased throu
local merchant, especially if it must be returned. In addition, delivery
can be slow, particularly at busy times of year. Some internet shoppers
complained about inferior products being sold on the internet, or about
culty in determining the details of a given product without seeing it in p
Others have noted technical problems when submitting orders, and ext
that are often not shown until you are just about to make a final purchase.
shoppers say there is a greater danger of making impulse purchases when
takes to purchase is a click of the mouse. Still others point out that there i
much you can do if you are not happy with your online purchase. For thes
other reasons, many shoppers prefer to continue shopping at their local
Have you read or heard anything about these problems recently? (Yes/N

Ask Rs either Question A or B:

A. [IF KNOWN TO HAVE "Somewhat negative" or "Very negative" expe


ences shopping online]: Have you ever had a negative experience of some
when buying things on the internet?

B. [IF NO EXPERIENCE OR NO KNOWN NEGATIVE EXPERIENCE]:


Has anyone you know had a negative experience when buying things on the
internet?

FOR ALL IN THIS CONDITION: Think for a minute about what you have
heard people say about why they dislike shopping on the internet - this could
be information from people you know or from newspapers, magazines, or
television. What is it people dislike about shopping online? Do people dislike
purchasing online because of

Yes No

Slow and unresponsive customer service


Unknown companies which may or may not be reliable

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
458 Mutz

Higher pri
Notenough
Technical d
Companies

Appendix
Survey-E
PREVIOUS
buy product
clothes, flo
equipment,
counting ev
a week, onc
months, on

INTERPER
speaking, h
(2)How oft
you think y
only some

HONESTY OF LOCAL VERSUS ONLINE BUSINESSES: Think for a minute


about (the businesses in your local area/businesses that sell products or ser-
vices on the Internet). How likely is it that a person buying something
from (a local business/an internet business) would have one of those (e-
businesses/businesses) try to cheat him or her in some way? (very likely to
very unlikely).

ACCOUNTABILITY OF LOCAL VERSUS ONLINE BUSINESSES: If a


person were being cheated by one of the businesses (in your local area/operati
on the internet), how likely is it that the (local business/internet business) wo
get away with it? (very likely to very unlikely).

GENERALIZED SOCIAL TRUST: Mean of six standardized items, includ


(1) would you say that most of the time people try to be helpful, or that they
mostly just looking out for themselves? (2) Do you think most people wo
try to take advantage of you if they got a chance, or would they try to
fair? (3) Generally speaking, would you say that most people can be trus
or that you can't be too careful in dealing with people? (4) Most people i
this society are trustworthy (strongly agree to strongly disagree); (5) W
they face temptations, people are not very honest (strongly agree to stron
disagree); (6) Which statement comes closer to your views, even if neither
exactly right? Human nature is basically bad, and you can't be too careful

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 459

your dealings with people, or human nature is basically good, and peop
be trusted. Cronbach's alpha = .81.

INTENT TO PURCHASE VIA INTERNET: Mean of four standardized


including ( 1 ) during the next year, how likely are you to buy a product or
through the internet? (DEFINITELY WILL buy something online, PRO
WILL buy something online, 50-50 CHANCE of buying something onl
PROBABLY WILL NOT buy anything online, DEFINITELY WILL NO
anything online); (2) relative to this past year, do you think you will mak
of your purchases through the internet this coming year, fewer of your p
through the internet or about the same number as this past year? (definite
purchases online, probably more purchases online, about the same num
last year, probably fewer purchases online, definitely fewer purchases
(3) in the next 12 months, about how much money do you think you wi
on purchases made through the internet? (none, less than $25, $25-49,
$75-99, $100-249, $250-499, $500-999, $1000 or more); (4) on average,
often are you likely to make purchases through the internet in the comin
(once a week, once every two weeks, once a month, once every three m
once every six months, once a year, less than once a year, never). Cron
alpha = .82.

References

AAPOR. 2008. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for
Surveys. 4th ed. Revised 2008, available at http://www.aapor.org/uploads/Standard_Definitions_
04_08_Final.pdf(accessed 6/1/2009).
Arena, Joe. 1996. "Framing an Ideology of Information: Retail Credit and the Mass Media, 1910-
30." Media, Culture and Society 18:423^5.
Austin, Marcus. 2008. "E-Retail Customer Satisfaction Remains High." www.internetretailing.
net/news.
Barber, Benjamin, Kevin Mattson, and John Peterson. 1997. "The State of Electronically Enhanced
Democracy: A Report of the Walt Whitman Center." New Brunswick, NJ.
Bern, Daryl J. 1967. "Self-Perception: An Alternative Interpretation of Cognitive Dissonance
Phenomena." Psychological Review 74:183-200.
Bodenhausen, Galen V., Schwarz Norbert, Herbert Bless, and Michaela Waenke. 1995. "Effects of
a Typical Exemplars on Racial Beliefs: Enlightened Racism or Generalized Appraisals". Journal
of Experimental Social Psychology 31:48-63.
Boorstin, Daniel J. 1974. The Americans: The Democratic Experience. New York: Vintage.
Boslego, Jordan. 2005. "Engineering Social Trust: What Can Communities and Institutions Do?"
International Health 27(1): Available at http://hir.harvard.edu/articles/13 19/.
Chatt, Cindy, and J. Michael Dennis. 2003. "Data Collection Mode Effects Controlling for Sample
Origins in a Panel Survey: Telephone versus Internet." Paper presented at the Annual Meeting
of the Midwest Chapter of the American Association for Public Opinion Research, Chicago, IL,
USA, November 2 1-22.
Cole, Jeffrey I. 2000. "Surveying the Digital Future." UCLA Center for Communication Policy,
University of California-Los Angeles.
Direct Marketing Association. 2007. DMA 2007 Statistical Fact Book. New York: Direct Marketing
Association.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
460 Mutz

Economist. 20
Etzioni, Amitai, and Orren Etzioni. 1997. "Communities: Virtual versus Real." Science
277(5324):295.
Festinger, Leon, and James M. Carlsmith. 1959. "Cognitive Consequences of Forced Compliance."
Journal of Abnormal and Social Psychology 58:203-10.
Fogg, B. J., Cathy Soohoo, David Danielson, Leslie Marable, Julianne Stanford, and Ellen R.
Tauber. 2002. "How Do People Evaluate a Web Site's Credibility? Results from a Large Study."
Stanford University Persuasive Technology Lab, Palo Alto, CA.
Forrester Research. 2007. "The State of Retailing Online 2007." The 10th Annual Shop.org Study.
Franklin, Charles. 199 1 . "Efficient Estimation in Experiments." The Political Methodologist 4: 1-3.
Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York:
New York Free Press.
Gandy, Oscar. 2002. "The Real Digital Divide: Citizens versus Consumers." In The Handbook of
New Media, eds. L. A. Lievrouw and Sonia Livingstone, pp. 448-60. London: Sage.
Giddens, Anthony. 1990. The Consequences of Modernity. Cambridge, UK: Polity.
Hampton, Keith, and Barry Wellman. 2003. "Neighboring in Netville: How the Internet Supports
Community and Social Capital in a Wired Suburb." City and Community 2:277-3 1 1 .
Horrigan, John, and Lee Rainie 2002. "Counting on the Internet." Pew Internet and American Life
Project.
Huang, Hai, Claudia Keser, Jonathan Leland, and Jason Shachat. 2003. "Trust, the Internet and the
Digital Divide." IBM Systems Journal 42(3):507-18.
Intermarket Group. 2001. The Internet Commerce Briefing: Three Volume Set. Business to Con-
sumer E-Commerce Report. New York: Intermarket Group.
Jackman, Robert W, and Ross A. Miller. 1998. "Social Capital and Politics." Annual Review of
Political Science 1:47-73.

Knack, Stephen, and Philip Keefer. 1997. "Does Social Capital Have an Economic Payoff? A
Cross-Country Investigation." Quarterly Journal of Economics 1 12:1251-88.
Knack, Stephen, and Paul J. Zak. 2003. "Building Trust: Public Policy, Interpersonal Trust, and
Economic Development." Supreme Court Economic Review 10:91-107.
Krosnick, Jon A., and LinChiat Chang. 2001. "A Comparison of the Random Digit Dialing Tele-
phone Survey Methodology with Internet Survey Methodology as Implemented by Knowledge
Networks and Harris Interactive." Paper presented at the Annual Meeting of the American As-
sociation for Public Opinion Research, Montreal, Quebec, Canada, May 17-20. Available at
httpV/www.knowledgenetworks.com/ganp/docs/OSUpaper.pdf.
Luhmann, Niklas. 1979. Trust and Power. Chichester, UK: Wiley.
Mutz, Diana C. 2005. "Social Trust and E-Commerce: Experimental Evidence for the Effects of
Social Trust on Individuals' Economic Behavior." Public Opinion Quarterly 69:1-23.
Nie, Norman H., and Lutz Erbring. 2002. "Internet and Society: A Preliminary Report." IT and
Society l(l):275-83.
Nie, Norman H., D. Sunshine Hillygus, and Lutz Erbring. 2002. "Internet Use, Interpersonal
Relations and Sociability: Findings From a Detailed Time Diary Study." In The Internet and
Everyday Life, eds. Barry Wellman and Caroline Haythornthwaite. Oxford, UK and New York:
Basil Blackwell.
Pierce, John C, and Nicholas P. Lovrich. 2003. "Internet Technology Transfer and Social Capital:
Aggregate and Individual Relationships in American Cities." Comparative Technology Transfer
and Society 1:49-7 '1.
Putnam, Robert D. (with) Robert Leonardi, and Raffaella Y. Nanetti. 1993. Making Democracy
Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press.
Raney, Rebecca Fairley. 2000. "Study Finds Internet of Social Benefit to Users." New York Times
(llMay):D7.
Robinson, John P., Meyer Ketsnbaum, Alan Neustadtl, and Anthony Alvarez. 2000. "Mass Media
Use and Social Life Among Internet Users." Social Science Computer Review 18:490-501.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms
Effects of Internet Commerce on Social Trust 461

Rothstein, Bo, and Eric M. Uslaner. 2005. "All for All: Equality, Corruption and Social Tru
World Politics 58:41-72.
Smith, Tom W. 1997. "Factors Relating to Misanthropy in Contemporary American Society."
Social Science Research 26: 170-96.
Sobel, Joel. 2002. "Can We Trust Social Capital?" Journal of Economic Literature 40(2): 139-54.
Warren, Mark E. 1999. Democracy and Trust. Cambridge: Cambridge University Press.
Wiebe, Elizabeth, Joe Eyerman, and John D. Loft. 2001. "Evaluating Non-Response in a Web-
Enabled Survey on Health and Aging." Paper presented at the Annual Meeting of the American
Association for Public Opinion Research, Montreal, Quebec, Canada, May 17-20.
Zak, Paul J. 2007. "The Neuroeconomics of Trust." In Renaissance in Behavioral Economics, ed.
Roger Frantz. New York: Routledge.
Zak, Paul J., and Stephen Knack. 2001 . "Trust and Growth." The Economic Journal 11 1 : 295-32 1 .
Zillman, Dolf, and Hans-Bernd Brosius. 2000. Exemplification in Communication. New York:
Lawrence Erlbaum.

This content downloaded from 193.231.1.90 on Thu, 26 Jan 2017 12:35:13 UTC
All use subject to http://about.jstor.org/terms

Das könnte Ihnen auch gefallen