Professional Documents
Culture Documents
2859
This preference for natural things influences not just hypothetical choices,
but also real-world decisions. In a study conducted by Baron, Holzman, and
Schulkin (1998), obstetricians and gynecologists who showed preference for
naturally occurring over identical man-made hormones in a hypothetical
choice also held more negative attitudes toward hormone replacement
therapy—an “unnatural” intervention whose overall benefits outweigh risks,
according to available medical evidence at the time of the study. Similarly,
preference for natural medicine versus a chemically identical synthesized
medicine was negatively associated with attitudes and behaviors toward
the influenza vaccine (Dibonaventura & Chapman, 2008), another overall
beneficial, but unnatural product.
Why do people prefer natural entities? Two distinct bases for such “natu-
ralness preference” were offered by Rozin and colleagues (Rozin, 2005;
Rozin et al., 2004): instrumental reasons (due to functional superiority of
natural products, such as effectiveness, safety, taste, or environmental ben-
efits) and ideational reasons (due to moral or aesthetic superiority of the
natural: that natural is prior to human intervention, or simply feels “right”).
The authors proposed that a substantial proportion of natural preference has
ideational bases. In their experiments (Rozin et al., 2004), ideational reasons
were inferred if participants preferred a natural product over its unnatural
alternative even though the options were specified to be identical in cost,
taste, healthfulness, and chemical composition (thus presumably excluding
instrumental reasons to prefer the natural). Indeed, the majority of partici-
pants did show a persisting naturalness preference under such conditions,
which was interpreted as evidence for ideational reasons behind the natural-
ness preference.
The above studies clearly demonstrate a pervasive naturalness preference.
However, the conclusion that the majority of participants must therefore
prefer natural for ideational rather than instrumental reasons is questionable.
One alternative possibility was that participants did not believe that the two
options were truly chemically identical (also acknowledged by Rozin et al.,
2004). If so, it is conceivable that participants may prefer the natural option
because of some inferred instrumental advantage of the natural.
Few studies have investigated this issue. In the only such study (Spranca,
1992, cited by Rozin et al., 2004), two drinking waters, one from a natural
spring and another made by adding/deducting minerals from distilled water,
were stated to be chemically identical, and the majority of participants
believed the statement. Belief in chemical identity was also unrelated to
participants’ preferences. Subsequent studies on the naturalness preference
(e.g., Rozin et al., 2004) largely assumed that participants accepted the iden-
tity of alternatives if told so. However, Spranca’s (1992) findings could be
restricted to the water scenario: Chemical identity may be easier to believe
WHY DO PEOPLE LIKE NATURAL? 2861
about waters than about more complex products. Indeed, water may be the
least complex substance about which natural and unnatural alternatives are
both available. Thus, investigators who assume participants believed that
natural and unnatural options in other studies were indeed identical as
described may have oversimplified the reasons behind the naturalness
preference.
We suspect that people may be reluctant to believe natural and unnatural
products can be identical in terms of their instrumental utility. Several studies
(Rozin, 2005, 2006; Rozin, Fischler, & Shields-Argeles, 2009) revealed that
people’s perception of the naturalness of an entity is more influenced by its
processing history than by its content, and perceived naturalness is reduced
by contact with unnatural entities and by chemical transformation, especially
with additives. Because any processing or contact with other entities may
increase potential risk through contagion, it is likely that concerns about risk
could exert an important influence on people’s preference for natural prod-
ucts. In fact, advertisements for natural merchandise usually emphasize their
avoidance of potentially risky processing and chemical additives.
Instead of assuming ideational reasons as the main source of naturalness
preference, we propose that ideational and instrumental reasons are both
important, and closely connected with each other, with the possibility of either
category of reasons originating from the other. Specifically, we want to stress
the possibility that instrumental reasons form prior to ideational reasons. The
idea that natural items are healthier, tastier, more effective, or better for the
environment may have permeated people’s belief systems to a great degree,
leading to a heuristic that “natural is better,” and thus, a preference for natural
options even when there is no evidence that they are better than unnatural
alternatives. It may appear as if such preference is purely ideational, given the
absence of real instrumental advantage in certain situations, while the true
reason may lie in the assumed instrumental benefits of natural products.
Although purely ideational reasons may exist in naturalness preference, we
hypothesize that part of the ostensible “ideational” naturalness preference is
associated with beliefs in instrumental benefits of natural products.
To test such hypotheses, we need to clarify several questions: First, do
people believe that natural and unnatural products are identical in hypotheti-
cal scenarios? Second, is suspicion about claimed identity related to what was
previously inferred to be ideationally based naturalness preference? Third,
are stronger beliefs that natural is instrumentally superior correlated with the
seemingly ideationally based naturalness preference? Studies 1 and 2 answer
the first two questions using hypothetical scenarios about specific natural and
unnatural entities. Study 3 addresses the last question by examining general
beliefs about instrumental advantages of natural products, and Study 4
shows the potential financial impact of the naturalness preference.
2862 LI AND CHAPMAN
Study 1
Methods
made it explicit that all other aspects of the two cities were equal: “The two
cities are very similar in safety, living expenses, culture, etc. Even the envi-
ronmental status of the two cities is identical: There is no difference in the
quality of air, water, soil, etc. And people living in the two cities are similar
in their level of health and enjoyment of their environment.” Importantly,
we equated the potential risk of the health hazards by stating that “The
pollen of these trees causes no other health problems” in City A, and “The
chemical residual causes no other health problems, because there are only
small traces of it” in City B.
Preference between the natural and unnatural alternatives were measured
following the description of each scenario, on a 7-point scale from strongly
prefer option X (natural vitamin C/city with natural DHA, rating of 3) to
strongly prefer option Y (unnatural vitamin C/city with unnatural DHA,
rating of -3), with 0 indicating no preference. Subsequently, participants
were reminded of the two statements stating “no other health problems” in
the descriptions of City A and City B, respectively, and for each statement,
the participants were asked “How confident were you about this statement
when you indicated your preference between the two cities?” Responses were
recorded on a 10-point scale from 1 “not confident at all” to 10 “very
confident.” The order of the two confidence questions was counterbalanced
among participants.
Results
Table 1
Preference for the Natural Option and Its Correlation with Beliefs in
Statements About Identity and Risk of the Natural and Unnatural Alternatives
SD = 2.75) compared with that about the tree pollen (M = 7.51, SD = 2.11),
t(446) = 15.63, p < .001, d = .74, suggesting greater perceived risk for an
unnatural compared with a natural health hazard. We also found that greater
difference in confidence between the two statements (in the direction of
lower confidence about chemical residues) correlated with preference for the
city where allergen was from the tree pollen, r = .39, p < .001 (see Table 1).
Frequency of confidence ratings revealed that 58.8% of the participants
indicated greater confidence in the “no-other-risk” statement about the tree
pollen, 28.6% of the participants showed equal confidence in the two state-
ments, and only 12.5% of the participants indicated greater confidence in the
same statement about the chemical residual. Among the 128 participants
who showed equal confidence in the two statements, preference rating
was still significantly above the indifference point, although the extent of
preference dropped by almost half from that among all participants, M = .70,
SD = 1.69, t(127) = 4.67, p < .001, d = .41. Preference rating frequency
indicates that about half (n = 65) of these participants who indicated
equal confidence preferred the natural option (preference above zero), 25%
(n = 32) showed no preference, and 24.2% (n = 31) preferred the unnatural
option.
Discussion
Study 2
The results in Study 1 lend support to our hypothesis that people are
reluctant to believe that natural and unnatural alternatives are identical, even
when they are told so, and such disbelief contributes to naturalness prefer-
ence. To complement the findings about risk beliefs in the loss domain, Study
2 used a modified version of the vitamin scenario adapted from Study 1.
Instead of implying chemical identity, we explicitly stated that two brands of
vitamin C pills (one from natural source and one manufactured) were chemi-
cally identical. Thus, we eliminated confounds from carriers or other ingre-
dients. To increase validity, we included two measures of confidence in
claimed identity of the options, one about chemical identity and one about
risk, and used different response formats in these two measures. The study
was conducted among a national sample. We expect the participants to show
skepticism in the statement that the natural and unnatural vitamin pills are
chemically identical, as in Study 1, and that such suspicion would be corre-
lated with preference for the natural vitamin C.
Methods
Results
Consistent with the findings from Study 1, preference for the natural
vitamin C was significantly above zero, the no-preference point, M = 1.93,
SD = 1.59, t(498) = 27.04, p < .001, d = 1.21 (see Table 1). Among all 499
participants, 383 (or 76.8%) showed a preference for the natural vitamin C.
WHY DO PEOPLE LIKE NATURAL? 2867
Also consistent with Study 1, despite scientific test results mentioned in the
scenario, the majority of the participants (65%) responded no to the question
asking their confidence about the chemical identity of natural and unnatural
vitamin C pills. The participants were also less confident that the synthesized
vitamin C pills posed no risk (M = 5.33, SD = 2.50) than they were about
natural vitamin C pills (M = 7.53, SD = 2.14), t(498) = 17.12, p < .001,
d = .77, indicating greater perceived risk in synthesized vitamin C pills. In
addition, as shown in Table 1, preference for the natural vitamin pills corre-
lated with suspicion about chemical identity, as well as with lower confidence
about the no-risk statement for unnatural versus natural vitamin C pills,
r = .24 and r = .45, respectively, ps < .001.
We then limited the analysis to the 93 participants who responded yes in
the question about belief in claimed chemical identity and also indicated
equal confidence in the no-risk statements for the two brands of vitamins,
and thus presumably accepted the identity of the two options. Mean natu-
ralness preference among these participants was significantly above zero,
although reduced to less than half of that among all participants, M = .87,
SD = 1.65, t(92) = 27.04, p < .001, d = .53. When these participants were cat-
egorized based on whether their preference rating was above, equal to, or
below zero, the indifference point, however, only 40.9% (n = 38) preferred
natural vitamin C pills, while 51.6% (n = 48) showed no preference, and 7.5%
(n = 7) preferred unnatural vitamin C pills.
Discussion
Study 2 replicated the results from Study 1 in the general population and
in the gain domain. People preferred vitamin C pills made from natural
ingredients to those synthesized in the lab. Consistent with Study 1, most
participants in Study 2 did not believe that the natural and synthesized pills
were chemically identical, although the scenario stated scientific tests as
evidence that they were, and the participants suspected of greater risk for
synthesized pills. Both types of suspicion were correlated with preference for
the natural vitamin, again suggesting that what would usually have been
categorized as ideational-based naturalness preference—because the condi-
tions in the scenario indicate no instrumental advantage in choosing the
natural and the participants were assumed to accept the conditions in the
scenario—is at least partly attributable to beliefs about instrumental aspects
of the products.
Study 2 also showed that these suspicious thoughts do not fully explain
people’s preference for the natural: The mean preference rating among
people who thought the natural and synthesized vitamin C pills were chemi-
cally identical and posed equal risks was still significantly above zero, indi-
2868 LI AND CHAPMAN
cating a preference for natural over the unnatural vitamins (albeit somewhat
weaker than mean preference among all participants). However, consistent
with Study 1, we only found a small fraction of participants who believed in
the identity of natural and unnatural options in chemical composition and
risk and still demonstrated a naturalness preference (n = 38, or 10% of the 383
participants who showed naturalness preference).
Taken together, Studies 1 and 2 suggest that a substantial proportion of
naturalness preference may be due to instrumental reasons, whereas purely
ideational-based preference may be inferred among only a small proportion
of the participants. We do not exclude, however, the possibility that one type
of reason originates from the other.
One common limitation of Studies 1 and 2 is that the hypothetical sce-
narios concerned specific items, such as vitamin C pills and allergens. Given
that beliefs in specific natural and unnatural entities may vary, people’s
beliefs about natural products in general need to be investigated. Another
limitation in Studies 1 and 2, as well as in previous research (Rozin, 2005;
Rozin et al., 2004; Spranca, 1992), is the neglect of an important potential
factor in naturalness preference: Many people may prefer natural options
because they think natural products are more environmentally friendly. Note
that environmental impact is an instrumental reason for naturalness prefer-
ence (Rozin, 2005; Rozin et al., 2004), as it provides actual benefits to the
individual indirectly through the environment. We addressed these limita-
tions in Study 3.
Study 3
Methods
Results
preference) and those who did not. Agreement to this statement was also
highly correlated with agreement to the suspicion on “unseen differences,”
r = .67, p < .001.
There were 135 participants who indicated sustained preference to natural
products even when material and functional differences do not exist and,
therefore, were asked about their reasons for such preference in the open-
ended question. Responses were coded by two independent coders using a
simple criterion: whether they mentioned each of the four attributes that were
explicitly excluded in the question as an advantage of natural products (taste,
safety, healthfulness, and environmental friendliness). This simple criterion
was adopted to avoid subjective interpretation of whether reasons listed in
the responses were instrumental or ideational reasons. Interrater reliability
was reasonably high, Cronbach’s a = .80, .83, .94, and .85, for the four
attributes, respectively. Disagreements between raters were solved through
discussion, and both raters agreed with the final coding. Among the 135
responses, 32 (or 23.7%) mentioned at least one of the four functional advan-
tages that were explicitly excluded in the question. The participants who
mentioned these instrumental advantages of natural options and those who
did not were comparable in all the other measures in the study, including
instrumental beliefs, agreement to “unseen differences,” and preference for
natural products. The remaining responses ranged from suspicion about
potential material and functional differences, reports for influence from
“culture” or the media (e.g., “It has just been advertised everywhere that
2872 LI AND CHAPMAN
natural is better, so I feel that this ongoing message in my head tells me that
natural is better even if everything about the two is the same”), to claims in
line with Rozin’s (2005; Rozin et al., 2004) definition of ideational reasons
(e.g., “Natural products are directly from the earth,” “natural things are
created by God for us,” “it just seems better”).
Discussion
The results from Study 3 support the idea that instrumental and idea-
tional reasons in naturalness preference are closely connected. The overall
favorable beliefs about natural products in terms of instrumental advantage
are consistent with instrumental reasons in naturalness preference. However,
these beliefs were also positively correlated with suspicion about unseen
material differences between natural and unnatural products even if there is
no detectable functional advantage, and with sustained preference to the
natural even if there is no detectable or undetectable functional advantage
(no instrumental superiority at all) in natural products. Thus, beliefs in
instrumental advantages of natural products could manifest in the preference
for natural products even under situations where such beliefs do not apply.
Responses to the open-ended question on reasons for naturalness prefer-
ence were consistent with our theorizing. About a quarter of the participants
who answered this question doubted that natural and unnatural products can
be identical, listing instrumental reasons that were explicitly excluded in the
question. There were also references to a “rule of thumb” cultured by media
messages or social norms. Some purely ideational reasons were also listed.
Despite potential post hoc rationalization involved in these free responses,
they seem to suggest a mix of instrumental and ideational reasons among the
participants who preferred natural for what might have been identified as
ideational reasons per se, as their preference persisted when instrumental
reasons were specified to be nonexistent.
Study 4
Method
The participants were willing to pay more for the organic coffee than for
an equivalent nonorganic coffee, with 82% of the participants indicating
prices higher than $10, and a median price at $13.50. A chi-square test
confirmed that more participants indicated a willingness-to-pay price above
$10, compared with prices equal or below $10, c2 (1, N = 266) = 108.65,
p < .001. Clearly, holding a preference for natural products has real-life
consequences in terms of cost. Perhaps a little worrisome is the fact that such
preferences could be sustained even without clear instrumental advantages of
natural products, as has been shown in the previous three studies. If “natural
is better” is a heuristic that people hold and apply to most situations where
they encounter a choice between natural and a comparable unnatural
product, it is very likely that people will spend more on products that do not
necessarily provide tangible benefits, just because they are “natural.”
General Discussion
In Studies 1 and 2, the majority of the participants indicated that they did
not believe the scenario statement, which cited scientific tests and authorities
as evidence to prove that the natural and unnatural alternatives were chemi-
cally identical and posed the same risk. The participants in Study 3 were
similarly skeptical. This sizable disbelief in the stated identity of alternatives
was largely ignored in previous research, which tacitly assumed full accep-
tance of whatever information was presented, possibly leading to an overly
strong conclusion that ideational, rather than instrumental, reasons play a
dominant role in naturalness preference. In contrast, we found a consistent
correlation between disbelief in the claimed identity of the two options and
preference for the natural, suggesting that at least part of what was originally
deemed as ideational-based naturalness preference is in fact due to suspected
differences between the options.
Study 3 demonstrated positive beliefs about instrumental advantages of
natural food products in general, as well as a correlation between such beliefs
and preference for the natural even when instrumental advantages of the
natural were specified to be nonexistent. When natural things are favored
without actual advantage, it would appear as if the preference is due to
ideational reasons. By demonstrating that such seemingly ideational prefer-
ence is stronger for those with stronger instrumental beliefs, we reveal the
close connection between such seemingly ideational preference and beliefs in
instrumental advantage of the natural.
Where do people obtain their favorable instrumental beliefs about natural
things? People may have acquired their beliefs about natural products from
multiple social sources, as indicated by the open-ended responses in Study 3,
such as media reports and advertisement, where messages in favor of natural
products are abundant. Or they can acquire such beliefs from personal expe-
rience or scientific evidence. For example, people could experience that
natural running (e.g., running across a field) is more enjoyable than artificial
running on a treadmill, and learn from scientific evidence that breast-feeding
is better than baby formula (Section on Breastfeeding, 2005). We propose
that the accumulated instrumental beliefs may lead to a “natural is better”
heuristic, which can be overextended to situations where a natural option
offers no more benefits than an unnatural one. Note that it does not mean
that people will always choose the natural option whenever it is available
(they will not when the cost of doing so is too great, e.g., most people will not
stay in the “natural” heat instead of an “unnaturally” air-conditioned room
in the summer). It just means that people will choose natural over unnatural
options when they are equivalent, or when there is no overwhelming evidence
to the superiority of the unnatural option.
An alternative explanation to the observed relationship between instru-
mental beliefs and naturalness preference is that the participants may have
WHY DO PEOPLE LIKE NATURAL? 2875
preferred natural options for inherently ideational reasons, and were simply
bolstering such preference by believing that natural products are instrumen-
tally superior. In the domain of vegetarianism, Rozin, Markwith, and Stoess
(1997) showed that compared with people who became vegetarians for health
reasons (instrumental), those who became vegetarians for moral reasons
(ideational) cited more nonmoral and nonhealth reasons for vegetarianism.
Although the causal direction from ideational to instrumental bases may
exist among some participants, we do not think it is the only story. We argue
that instrumental and ideational reasons are much more intermingled than
previously thought.
The heuristic that “natural is better” is especially influential in the natural
product market, because often when consumers choose between natural and
non-natural options—such as organic versus commercially farmed produce—
they cannot directly observe instrumental differences between the two, but
may have reasons to believe differences exist. And, even when two products are
“chemically identical” from the point of view of a chemist, they may be seen as
different because of their processing history. Indeed, research indicates that
consumers of organic food (a major category of natural products) tend to
think there are certain intrinsic (quality or safety) properties of organic
products (Vindigni, Janssen, & Jager, 2002), either observable or unobserv-
able. Giannakas (2002) noted that organic products are “credence goods”
(Darby & Karni, 1973; Nelson, 1970), meaning that producers know whether
the product is organic, but organic characteristics are not detectable by
consumers. Thus, believing that there are undetectable advantages of natural
products is not an unfamiliar aspect of consumer decision making. Judgments
made on the basis of believed but undetectable characteristics are consistent
with the psychological essentialism theory (Medin & Ortony, 1989), which
maintains that people hold a commonsense assumption that things have
essences that are responsible for their surface features, and make them what
they are. For example, a tadpole and a frog are perceived as the same animal
because there is an essence of the animal “frog” that maintains identity despite
the different appearances through the developmental stages. People are able to
grasp such essences without scientific understanding of genetics. It is possible
that what makes natural things natural in people’s minds is an “essence” that
distinguishes them from human-made things, even if they cannot point out the
exact nature of this essence. In other words, they may have an “essence
placeholder” (Medin & Ortony, 1989), for example, believing that there are
people who really know or could figure out what the essence is.
Our results indicate that instrumental beliefs play a critical role in the
naturalness preference. The question of whether this preference is rational,
however, needs discussion. If we adopt a consequentialist definition of ratio-
nality, a purely ideational naturalness preference is clearly irrational because
2876 LI AND CHAPMAN
References