Inviting Free-Riders or Appealing to Prosocial Behavior? Game

Transcription

Inviting Free-Riders or Appealing to Prosocial Behavior?
Game-Theoretical Reflections on Communicating Herd Immunity in Vaccine Advocacy
Cornelia Betsch, Robert Böhm, and Lars Korn
University of Erfurt
Running Head: COMMUNICATING HERD IMMUNITY IN VACCINE ADVOCACY
Health Psychology, in press
Author Note
Cornelia Betsch, Center for Empirical Research in Economics and Behavioral Sciences
(CEREB), University of Erfurt; Robert Böhm, Center for Empirical Research in Economics and
Behavioral Sciences (CEREB), University of Erfurt; Lars Korn, University of Erfurt.
CB and RB contributed equally to this paper. The authors are grateful to the master
students of the fall 2011 term, who helped set up the study and collect data. Tilmann Betsch,
Wasilios Hariskos, Fabian Kleine, Frank Renkewitz, Heather Fuchs, and the CEREB research
group made helpful comments on an earlier draft of the paper.
Correspondence concerning this article should be addressed to Cornelia Betsch,
University of Erfurt, Nordhäuser Str. 63, D-99089 Erfurt, Germany. E-Mail:
[email protected].
COMMUNICATING HERD IMMUNITY IN VACCINE ADVOCACY
2
Abstract
Vaccination yields a direct effect by reducing infection but also has an indirect effect,
herd immunity: If many individuals are vaccinated, the immune population will protect
unvaccinated individuals (social benefit). However, due to a vaccination’s individual costs and
risks, individual incentives to free-ride on others’ protection also increases with the number of
individuals who are already vaccinated (individual benefit).
Objective was to assess the consequences of communicating the social and/or individual
benefits of herd immunity on vaccination intentions. We assume that if social benefits are
salient, vaccination intentions increase (prosocial behavior), whereas salience of individual
benefits might decrease vaccination intentions (free-riding).
Methods: In an online-experiment (N = 342) the definition of herd immunity was
provided with one sentence summarizing the gist of the message, either making the individual or
social benefit salient or both. A control group received no information about herd immunity. As
a moderator we tested the costs of vaccination (effort in obtaining the vaccine). Dependent
measure was intention to vaccinate.
Results show that when a message emphasized individual benefit, vaccination intentions
decreased (free-riding). Communication of social benefit reduced free-riding and increased
vaccination intentions when costs to vaccinate were low.
Conclusions: Communicating the social benefit of vaccination may prevent free-riding
and should thus be explicitly communicated if individual decisions are meant to consider public
health benefits. Especially when vaccination is not the individually (but instead collectively)
optimal solution, vaccinations should be easily accessible in order to reach high coverage.
Keywords: public health, immunization, social dilemma, advocacy, communication strategies
3
Inviting Free-Riders or Appealing to Prosocial Behavior?
Game-Theoretical Reflections on Communicating Herd Immunity in Vaccine Advocacy
Vaccination is typically treated as an individual decision making task. In addition to
motivational factors such as adherence to social norms (Brown et al., 2011; Liao, Cowling, Lam
& Fielding, 2011; Sturm, Mays & Zimet, 2005; Ajzen & Fishbein, 1980), a vaccination’s
(perceived) individual costs and benefits are especially predicative of vaccination intention. This
is proposed by several theoretical models of preventive health behavior (for an overview see
Weinstein, 1993) and has been confirmed by empirical work (e.g. Brewer et al, 2007; Brewer &
Fazekas, 2007; Nguyen, Henningsen, Brehaut, Hoe & Wilson, 2011). Costs (barriers) of
vaccination can be monetary and non-monetary, such as the time needed to obtain a vaccination,
but also include the risks associated with vaccination such as the occurrence of vaccine-adverse
events. Benefits of vaccination vary according to the vaccine’s effectiveness as well as the
likelihood and severity of the disease the vaccine protects against.
Vaccination yields a direct effect by reducing infection. Moreover, vaccination against
contagious diseases has an additional indirect effect (Fine, Eames & Heyman, 2011): The
transmission of a disease is reduced with an increasing number of vaccinated individuals. An
indirect effect of vaccination can have two major implications: On the one hand, each
vaccination reduces the transmission of an infection in the population (Anderson & May, 1991),
which protects other susceptible individuals (for instance, those who are too young to vaccinate
or immunocompromised). With a critical vaccination level, herd immunity and disease
eradication can be reached (e.g. 95% vaccine coverage will allow for eradication of the measles
in Europe; Christie & Gay, 2011; Smith, 1970). Hence, vaccination yields a social benefit, as
vaccine coverage above the critical level is optimal from the collective perspective. On the other
4
hand, if everyone else is directly protected by vaccination, free-riders (or free-loaders; see Fine
et al., 2011) can benefit from the indirect effects of vaccination and henceforth avoid individual
costs of vaccination (e.g. money, time, vaccine-adverse events, inconvenience). The indirect
effect of vaccination, therefore, also yields an individual benefit. The presence and awareness of
both individual and social benefit from herd immunity result in a mixed-motive situation that
renders vaccination a strategic interaction (Schelling, 1960). As high vaccination uptake is of
major importance for society in order to reach public health goals, it is a fundamental question
how societies can increase vaccination uptake.
This contribution investigates how the communication of herd immunity may affect
vaccination uptake. More precisely, we examine how the awareness of a vaccination’s
individual benefit, social benefit, or both affects vaccination intention. According to the theory
of reasoned action and the theory of planned behavior, salient beliefs determine an individual’s
attitude towards a behavior and behavioral intentions (Ajzen & Fishbein, 1980). An individual’s
awareness of herd immunity may therefore either increase or decrease vaccination uptake,
depending on the salience of individual and social benefits that result from the indirect
protection of herd immunity.
Vaccination as a strategic interaction
The interactive structure of vaccination decisions has recently been discussed in the
literature (e.g. Bauch & Earn, 2004; Bhattacharyya & Bauch, 2010; Galvani, Reluga, &
Chapman, 2007; Manfredi et al., 2010). Based on this literature we devise a simple model of
vaccination as strategic interaction which is illustrated in Figure 1. The expected costs of a
disease (E[cD]) are typically treated as the product of the severity of the disease and the
probability to contract the disease. Similarly, the expected costs of a vaccination (E[cV]) are the
5
product of its costs (monetary and non-monetary) and their probability (e.g. the likelihood of
adverse advents): E[cD] = SEVD × PROBD and E[cV] = SEVV × PROBV (Weinstein, 1993). The
expected utility from vaccination (EUV) and non-vaccination (EU¬ V) results from the difference
between the expected costs of the disease and the expected costs of vaccination: EUV = E[cD] E[cV] and EU¬ V = E[cV] - E[cD]. Consequently, small differences in expected costs of the
disease and vaccination yield larger differences in the expected utility of vaccination and nonvaccination, since EUV - EU¬ V = 2(E[cD] - E[cV]). Furthermore, the probability of contracting a
contagious disease, and therefore also the expected costs of the disease, over-proportionally
decreases as a function of the number of vaccinated individuals, because the lifetime incidence
for unvaccinated individuals decreases (e.g., Fine et al., 2011). At the same time, the expected
costs of the vaccination remain constant, since, for instance, the probability and severity of sideeffects is not affected by the number of vaccinated individuals. Therefore, when the number of
vaccinated individuals (vaccine coverage) increases, the expected costs of the vaccination will at
some point exceed the expected costs of the disease (E[cD] - E[cV] < 0; cf. intersection in Figure
1; Chen, 1999). From an individual perspective, non-vaccination then becomes the best response
as EUV < EU¬ V. In contrast, it is collectively optimal to vaccinate until a vaccination level is
achieved that eradicates the disease, as the expected cumulative incidence is zero if coverage is
maintained above the critical vaccination level (Vc; Fine et al., 2011). Therefore, as soon as
E[cD] - E[cV] < 0 the vaccination decision contains a motivational conflict between the
individual and collective interest. If Vc is reached, non-vaccination is the best response from an
individual and collective perspective. Within these boundary conditions, vaccination constitutes
a N-person prisoner’s dilemma in which individuals may decide whether or not to contribute
(i.e. vaccinate) to a public good (i.e. herd immunity); see shaded area in Figure 1.
6
It is important to note that expected costs are subjective variables and can therefore
deviate from objective cost parameters. Today, wild forms of severe vaccine-preventable
diseases are rare and most individuals lack a concrete, vivid representation of the disease (= low
perceived costs of disease). At the same time, vaccination costs are more visible, vivid and
tangible due to high vaccination rates and immediate effort and inconvenience (= high costs of
vaccination; Chen, 1999). Many modern vaccination decisions (e.g. against polio or the
measles) can therefore be framed as a social dilemma.
As discussed, herd immunity can have two potential effects. On the one hand, if the
individual benefit of herd immunity is communicated, the individual’s selfish/egoistic
preferences might be activated (Dawes, 1980; Hardin, 1968) and accordingly affect behavior
(Ajzen & Fishbein, 1980). Therefore, the first hypothesis predicts:
H1. If the individual benefit of herd immunity is salient (but not the social benefit),
participants will show lower vaccination intentions compared to when the individual
benefit is not salient (free-riding hypothesis).
On the other hand, by making the social benefit of one’s own vaccination salient,
individuals’ positive, other-regarding preferences might be activated (for an overview see Fehr
& Schmidt, 2006). This leads to the second hypothesis:
H2. If the social benefit of herd immunity is salient (but not the individual benefit),
participants will show higher vaccination intentions compared to when the social benefit
is not salient (prosocial behavior hypothesis).
This study extends prior work (e.g., Hershey, Asch, Thumasathit, Meszaros & Waters,
1994; Shim, Chapman, Townsend, & Galvani, 2012) by orthogonally manipulating the salience
7
of individual and social benefits. We are, therefore, also able to explore the interaction between
the salience of individual and social benefits.
In Figure 1, if the expected costs of vaccination increase or decrease (grey dashed line
moves up or down, respectively), ceteris paribus, the intersection of E[cD] = E[cV] shifts further
to the left or right, respectively. Consequently, the difference in expected utility between
vaccination and non-vaccination on the right-hand side of this intersection increases or
decreases, respectively. This effect occurs due to the indirect effect of vaccination. If individuals
are unaware of this indirect effect (e.g. because herd immunity is not communicated), the
expected costs of the disease will not depend on the vaccination of others and will therefore be
constant. As, in this case, the solid grey line would be parallel to the dashed grey line, selfish or
other-regarding preferences should thus have no effect.
The (perceived) costs of vaccination should therefore interact with the effects of
communicated herd immunity: Free-riding entails the benefits of vaccination (due to others’
immunization) without carrying the costs. It follows that the incentives for free-riding are
especially high if costs to vaccinate are high. Likewise, individuals who vaccinate due to a
prosocial motivation (to protect the unimmune) take over costs for the society. Hence, if these
costs are high, prosocial behavior may be less likely. This leads to our third hypothesis, which
integrates the structurally equivalent sub-hypotheses of individual and social benefit salience:
H3. If the individual benefit of herd immunity is salient, participants are more inclined to
free-ride when the costs of vaccination are high than when the costs of vaccination are
low. Similarly, if the social benefit of herd immunity is salient, participants are more
inclined to prosocial behavior when the costs of vaccination are low than when the costs
of vaccination are high (vaccination costs interaction hypothesis).
8
Experiment
The hypotheses were tested in an online experiment that assessed vaccination intentions
regarding hypothetical diseases (for a similar methodological approach, see cf. Betsch &
Sachse, 2012; Betsch, Renkewitz, & Haase, in press; Betsch, Ulshöfer, Renkewitz, & Betsch,
2011; Vietri, Li, Galvani, & Chapman, 2012). The definition of herd immunity was provided
along with one sentence summarizing the gist of the message (Reyna, 2012), making salient
either the individual benefit, social benefit, or both. A control group received no information
about herd immunity. Additionally, we tested if the cost of getting vaccinated, operationalized
as the amount of effort required to obtain the vaccine, interacts with the communicated benefits.
Method
Participants and design. Participants were recruited via mailing lists and social network
websites (e.g. Facebook). As compensation, all participants took part in a raffle for one of five
gift certificates (25 Euro; ~ $ 31). N = 371 participants completed the questionnaire. 29
participants were excluded from the sample due to excessively long (> 30 min) or short (< 5
min) duration of participation, resulting in a mean duration of 12.76 minutes (SD = 5.14).
Hence, the final sample consisted of 342 participants, both students and non-students. Eightyeight percent of the sample had an Abitur (German University entrance diploma) or higher level
of education. The mean age of the sample was 30.34 years (SD = 12.5); 221 (64%) participants
were female.
The experiment used a 2 × 2 × 2 between-subject design with individual benefit of herd
immunity (communicated vs. not communicated), social benefit of herd immunity
(communicated vs. not communicated), and costs of vaccination (low vs. high) as factors. It was
9
realized with an online software program (EFS survey), which randomly allocated participants
to the eight conditions.
Herd immunity. In the control condition (individual and social benefits of herd
immunity were not communicated), participants received no information about herd immunity.
In the remaining three conditions, participants received the following definition of herd
immunity: “Herd immunity denotes the effect that occurs when acquired immunity against a
pathogen, generated through infection or vaccination, within a population (the “herd”) has
reached such a level that non-immune individuals in this population are also protected, because
the pathogen can no longer be transmitted”. Furthermore, depending on condition, one
additional sentence summarized the gist of the message, manipulating the salience of the
individual benefit, social benefit, or both. Individual benefit was highlighted by the sentence
“The more people are vaccinated in your environment, the more likely you are protected without
vaccination”. Social benefit was highlighted by the sentence “If you get vaccinated, then you
can protect others who are not vaccinated”.
Vaccination costs. Participants were either informed that they could get vaccinated
immediately (low cost) or that they would have to set up an appointment with the local hospital
and that this appointment would take almost three hours (high cost).
Measures.
Dependent measure. Vaccination intention was assessed (“If you had the opportunity to
vaccinate against [the illness] next week, what would you decide?”) on a 7-point Likert-type
scale ranging from 1 = definitely not vaccinate to 7 = definitely vaccinate.1
COMMUNICATING HERD IMMUNITY IN VACCINE ADVOCACY 10
Manipulation check. As a manipulation check, one additional item assessed the
perceived costs of the vaccination (“How do you estimate your personal vaccination costs?”; 1 =
very low to 7 = very high).
Procedure. After giving informed consent, participants were informed that all presented
information is fictitious. The questionnaire began with the measurement of demographic
characteristics. They were then asked to imagine a fictitious scenario: During a routine physical
examination, the general practitioner informs them about the severe infectious disease
Cornicoviszidosis. This recently discovered illness had been diagnosed increasingly often. The
participants received additional information about the origin of Cornicoviszidosis, the name of
the responsible virus (Cornicovi), the path of infection (smear infection), and the symptoms of
the disease (severe vomiting and diarrhea, severe dehydration, and high fever). Participants
received a data sheet for a fictional vaccine termed Macentat containing information about
vaccine-adverse events: hypersensitivity reaction of the skin with probability p = .1; headache, p
= .0001 to .001; vomiting, vertigo, and skin rash with a probability less than .0001. Information
about herd immunity was displayed afterwards. Before participants indicated their intention to
get vaccinated, they were informed about the costs of the vaccination. Finally, the manipulation
checks were recorded and participants fully debriefed.
Results
We present eta-squared as an effect size indicator along with all statistically significant
results. All non-significant comparisons have an F < 1 if not stated otherwise.
Manipulation check. As intended, the vaccination costs in the high cost condition were
perceived as significantly higher than in the low cost condition (Mhigh = 4.67, SD = 1.69; Mlow =
2.20, SD = 1.36; F(1, 340) = 222.06, p < .001, η2 = .40).
Vaccination intention. The mean values and standard deviations of vaccination intention
by experimental condition are displayed in Table 1. Inspection of the means in Table 1 suggests
that the control condition (without any herd immunity information) showed, on average, higher
vaccination intentions than when herd immunity was communicated in either manner. Indeed, a
post-hoc simple contrast analysis (control group vs. all other conditions) yields a significant
difference, F(3, 363) = 3.69, p = .012, η2 = .030. Still, what are the specific effects of
communicating individual vs. social benefits on vaccination intentions? H1 and H2 predict
lower vaccination intentions when the individual benefit is salient (free-riding hypothesis), and
higher vaccination intentions when the social benefit is salient (prosocial behavior hypothesis).
H3 suggests an interaction between benefit salience and costs.
To test the hypotheses, we conducted a 2 × 2 × 2 analysis of variance with vaccination
intention as the dependent variable. Results support the free-riding hypothesis (H1): when
individual benefit of herd immunity was communicated, vaccination intentions were
significantly lower than when it was not communicated, Mcomm = 3.72, SD = 1.84; M¬ comm =
4.18, SD = 1.77; F(1, 334) = 4.33, p = .038, η2 = .012. There was no difference in vaccination
intentions whether the social benefit was communicated or not, Mcomm = 4.01, SD = 1.86; M¬
comm
= 3.89, SD = 1.78, yielding no evidence in support of the prosocial behavior hypothesis
(H2). However, there was a significant interaction effect between the individual and social
benefit communication conditions, F(1, 334) = 3.90, p = .049, η2 = .011. As Figure 2A shows,
when social benefit was not communicated, vaccination intentions varied as a function of the
communicated individual benefit: the vaccination intention was significantly lower when the
individual benefit of vaccination was communicated than when it was not communicated, F(1,
177) = 9.38, Bonferroni-corrected p = .006, η2 = .051, indicating the tendency to free-ride on
others’ indirect protection. When the social benefit of herd immunity was communicated,
however, there was no such difference, regardless of whether the individual benefit was
additionally communicated. 2
In general, vaccination intentions were lower when vaccination costs were high than
when they were low, Mhigh = 3.83, SD = 1.87; Mlow = 4.10, SD = 1.79; F(1, 334) = 4.57, p =
.033, η2 = .013. As expected, vaccination costs interacted with the social benefit communication,
F(1, 334) = 7.43, p = .007, η2 = .021. As Figure 2B shows, when the social benefit of herd
immunity was communicated, vaccination intentions were higher when costs were low than
when costs were high, F(1, 163) = 11.07, Bonferroni-corrected p = .002, η2 = .064, indicating
conditional prosocial vaccination behavior. There was no such difference when social benefit
was not salient. However, vaccination intentions under individual benefit salience did not differ
between the low vs. high cost conditions. Taken together, results partially confirm the
vaccination costs interaction hypothesis (H3). The three-way interaction was not significant, F <
1.6, ns.
Discussion
The goal of this paper was to assess if and how the communication of herd immunity
may affect vaccination uptake. We devised a simple theoretical model of (non-)vaccination
utility as a function of the perceived costs of the disease and the vaccination contingent on the
number of vaccinated individuals in the population. Furthermore, some of the model’s
implications were tested in an experiment. The data shows that communicating the concept of
herd immunity can have two effects, depending on the gist of the message: First, when a
message emphasized the individual benefit of indirect protection through the “herd”,
individuals’ inclination to free-ride increased (H1). This was especially the case when the social
benefit of herd immunity was not communicated. Second, communicating the social benefit did
not result in a general increase in vaccination intentions (contradicting H2). However,
communicating the social benefit reduced free-riding and also had the potential to increase
vaccination intentions when the costs of vaccinating are perceived as low. Thus, depending on
which implication of herd immunity was made salient, vaccination intentions differed.
The lacking overall positive effect of communicated social benefit along with the
obtained free-riding effect pose the question whether the communication of herd immunity is
advisable at all. Strong emphasis on the social benefit, however, still seems advisable: even if it
might not have an overall positive effect, it might at least prevent free-riding.
This becomes even more clear when we consider that many cues in the decision structure
may invite free-riding: recent game-theoretic models of vaccination uptake have shown that the
level of vaccination may decrease dramatically and fall below the social optimum if the
expected costs of vaccination increase (e.g. due to a vaccine scare or anti-vaccination activism;
Bhattacharyya & Bauch, 2010) or if the costs of the disease decrease (Jansen et al., 2003). This
occurs mainly because of free-riding on the immunized herd (e.g., Bauch & Earn, 2004,
Galvani, Reluga, & Chapman, 2007; Manfredi et al., 2010). Furthermore, individuals are
sensitive to different levels of immunity in the population if these are varied in a withinsubjects’ setting (Vietri et al., 2012): the more people were vaccinated, the less likely
participants were to vaccinate themselves. Our results contribute to this literature by showing
that quite subtle communications are also sufficient to suggest a free-riding opportunity. Again,
this implies that strategies that prevent free-riding are needed. The current results suggest that
appealing to prosocial motives might be such a strategy to reduce free-riding tendencies (see
also Shim et al., 2012).
As previous research focused particularly on the negative effect of indirect protection
through herd immunity (free-riding; e.g., Bauch & Earn, 2004), more scientific attention should
be directed to its positive effect (i.e. prosocial behavior), in order to assess the boundary
conditions under which the communication of the social benefit of vaccination does increase
vaccine intentions (for instance, as in the present experiment, under low perceived costs of
vaccination; or when the risk for the self is low as shown by Vietri et al., 2012). This becomes
particularly important if the direct effect of vaccination is very small or even absent and the
indirect effect has important consequences for eradicating a disease (as in the most extreme case
of malaria control; Carter, Mendis, Miller, Molyneux, & Saul, 2000).
Limitations and Further Research
A number of possible limitations to the results should be noted. First, the hypothetical
scenario and the self-report data (intentions) might limit the external validity of the results.
Additionally, online-experiments may be subject to self-selection bias. Indeed, the participants
in our sample were typically well educated. However, including education in the analyses did
not affect the pattern of results. Moreover, the present study tests hypotheses derived from a
general game theoretical model. We do not assume different relations between utility functions
for individuals with different levels of education. We therefore conclude that external validity of
the results is given even though the sample might be skewed towards higher educated
participants.
Further, we neither manipulated nor measured the individual perception of vaccine
coverage and could therefore not test its impact on perceived costs of the disease. This should be
a next step in further research. For instance, an experimental public goods setting of vaccination
(see Chapman et al., 2012) may be a viable framework to test the dynamic relationship between
vaccination uptake and the number of others vaccinated. Nevertheless, the perceived risk of
both the vaccination and the disease were assessed in this study and can serve as proxies for
perceived costs (see footnote 1). As expected, the effects were stronger if E[cD] - E[cV] < 0 than
if E[cD] - E[cV] > 0 (see footnote 2). The presence of the effects in the total sample suggests that,
even if only a sub-sample of the population perceives higher vaccination than disease costs,
communicating the social benefit can have positive effects on vaccination intentions.
The present experiment was ambiguous regarding whether the other protected
individuals were not able to get vaccinated (because they are too young or
immunocompromised) or simply not willing to do so (free-riders) when the individual benefit of
herd immunity was communicated. A large amount of research in economics has shown that
people are willing to cooperate if others are also expected to do so (conditional cooperation;
e.g., Bolton & Ockenfels, 2000; Fehr Gächter, & Fischbacher, 2001; Fehr & Schmidt, 1999).
Moreover, in social psychology there is evidence that prosocial behavior is more likely when
uncontrollable factors created the situation of need (Weiner, 1980). Thus, if it is communicated
that others are explicitly not able to protect themselves, the communicated social benefit of herd
immunity should have a larger effect.
Similarly, the way in which costs were manipulated represents only one out of several
possibilities. As said before, costs accrue due to time, money, side effects, inconvenience, etc.
Future studies should use different approaches to manipulate costs. The perception or fear of
potential side effects (such as elicited in vaccine scares) decreases vaccination intentions and are
among the most prominent reasons against vaccination (Betsch, Renkewitz, Betsch, & Ulshöfer,
2010; Brown et al., 2010a, 2010b). Thus, it is possible that if costs are manipulated via potential
side effects of the vaccine, the obtained effects may be stronger.
Practical Implications
Overall, it seems advisable to stress vaccination’s social benefit in vaccine advocacy.
This is especially the case if the concept of herd immunity is communicated to the public, such
as during the process of eradicating diseases, e.g. the measles and rubella in Europe until 2015
(Christie & Gay, 2011). If the indirect effects of vaccination become obvious, free-riding might
increase, as vaccine coverage is usually already high (but not high enough). Therefore, stressing
the social benefit may help to reach critical vaccination levels in order to eradicate diseases.
Protection of others is especially important in contexts with highly vulnerable
individuals, such as immunocompromised patients in a hospital. For this reason, the WHO
recommends vaccination against influenza for health care personnel (HCP). Despite the
availability of an effective and well-tolerated vaccine, low seasonal and pandemic influenza
vaccine acceptance among HCP is a major problem detailed in many studies from all over of the
world (Salgado, Giannetta, Hayden & Farr, 2004; Talbot et al., 2010). The perception of
vaccination risks in addition to other expected vaccination costs are major reasons why HCP do
not get vaccinated against influenza (Betsch & Wicker, 2012; Wicker, Rabenau, Doerr &
Allwinn, 2009). One could speculate that HCP may generally be more prosocially oriented (e.g.,
Van Lange, 1999). Thus, building on the idea of tailoring health messages (Noar, Benac &
Harris, 2007), appeals to the social benefit of vaccination could be a viable strategy to increase
HCP’s vaccination rates.
Conclusions
We conclude that the social benefits of vaccination need to be explicitly communicated
if the individual decisions are meant to consider public health benefits. Even if it does not
generally raise vaccination intentions, it can prevent free-riding and has the potential to increase
vaccination intentions when the costs to vaccinate are low. A recent approach for vaccine
advocacy suggests that ‘Vaccination Adoption = Access + Acceptance’ (Thomson & Watson,
2012). Acceptance can be understood as E[cD] - E[cV] > 0 or as prosocially oriented behavior
under E[cD] - E[cV] < 0. Access can be understood as low costs to obtain the vaccine. The latter
also proved to be an important variable in our study. Especially when vaccination is not the
individually optimal solution and public health considerations suggest a collective benefit of
vaccination (e.g. cocooning newborns against pertussis, preventing school children from
spreading influenza to the elderly), access should be very easy in order to obtain high
vaccination coverage.
References
Ajzen, I., & Fishbein M. (1980). Understanding attitudes and predicting social behaviour.
Englewood Cliffs, NJ: Prentice-Hall.
Anderson, R. M., & May R. M (1985). Vaccination and herd immunity to infectious diseases.
Nature, 318, 323–329.
Bauch, C. T., & Earn, D. J. D. (2004). Vaccination and the theory of games. PNAS: Proceedings
of the National Academy of Sciences of the United States of America, 270, 1573-1578.
Betsch, C. & Sachse, K. (in press). Debunking vaccination myths – Strong risk negations can
increase perceived vaccination risks. Health Psychology.
Betsch, C., Renkewitz, F. & Haase, N. (in press). Effect of narrative reports about vaccine
adverse events and bias-awareness disclaimers on vaccine decisions: A simulation of an
online patient social network. Medical Decision Making.
Betsch, C., Renkewitz, F., Betsch, T., & Ulshöfer, C. (2010). The influence of vaccine-critical
internet pages on perception of vaccination risks. Journal of Health Psychology, 15, 446455.
Betsch, C., Ulshöfer, C., Renkewitz, F. & Betsch, T., & (2011). The influence of narrative vs.
statistic information on perceiving vaccination risks. Medical Decision Making, 31, 742753.
Bhattacharyya, S., & Bauch, C. T. (2010). A game dynamic model of delayer strategies in
vaccinating behaviour for pediatric infectious diseases. Journal of Theoretical Biology,
267, 276-282.
Bolton, G. E., & Ockenfels, A. (2000). ERC: A theory of equity, reciprocity, and competition.
American Economic Review, 90, 166-193.
Brewer, N. T., Chapman, G. B., Gibbons, F. X., Gerrard, M., McCaul, K. D., & Weinstein, N.
D. (2007). Meta-analysis of the relationship between risk perception and health behavior:
The example of vaccination. Health Psychology, 26, 136-145.
Brewer, N. T., & Fazekas K. I. (2007). Predictors of HPV vaccine acceptability: A theoryinformed, systematic review. Preventive Medicine, 45, 107-114.
Brown, K., Fraser, G., Ramsay, M., Shanley, R., Cowley, N., van Wijgerden, J., … Sevdalis, N.
(2011). Attitudinal and demographic predictors of measles-mumps-rubella vaccine
(MMR) uptake during the UK catch-up campaign 2008-09: Cross-sectional survey. PLoS
ONE, 6. doi:10.1371/journal.pone.0019381
Brown, K. F., Kroll, J. S., Hudson, M. J., Ramsay, M., Green, J., Long, S. J., ... Sevdalis, N.
(2010a). Factors underlying parental decisions about combination childhood vaccinations
including MMR: A systematic review. Vaccine, 28, 4235-4248.
Brown, K. F., Kroll, J. S., Hudson, M. J., Ramsay, M., Green, J., Vincent, C. A., … Sevdalis, N.
(2010b). Omission bias and vaccine rejection by parents of healthy children: Implications
for the influenza A/H1N1 vaccination programme. Vaccine, 28, 4181-4185.
Carter, R., Mendis, K. N., Miller, L. H., Molyneux, L., & Saul, A. (2000). Malaria transmissionblocking vaccines: How can their development be supported? Nature Medicine, 6, 241244.
Chapman, G.B., Li, M., Vietri, J., Ibuka, Y., Thomas, D., Yoon, H., & Galvani, A.P. (2012).
Using Game Theory to Examine Incentives in Influenza Vaccination Behavior.
Psychological Science, 23, 1008-1015.
Chen, R. T. (1999). Vaccine risks: Real, perceived and unknown. Vaccine, 17, 41-46.
Christie, A. S., & Gay, A. (2011). The measles initiative: Moving toward measles eradication.
The Journal of Infectious Diseases, 204, 14-17.
Dawes, R. M. (1980). Social dilemmas. Annual Review of Psychology, 31, 169-193.
Fehr, E., & Schmidt, K. M. (1999). A theory of fairness, competition, and cooperation.
Quarterly Journal of Economics, 114, 817-868.
Fehr, E., & Schmidt, K. M. (2006). The economics of fairness, reciprocity and altruism –
Experimental evidence and new theories. In S.-C. Kolm & J.M. Ythier (Eds.), Handbook
on the Economics of Giving, Reciprocity and Altruism (pp. 615-691). Amsterdam: NorthHolland Publishing.
Fischbacher, U., Gächter, S., & Fehr, E. (2001). Are people conditionally cooperative? Evidence
from a public goods experiments. Economic Letters, 71, 397-404.
Galvani, A. P., Reluga, T. C., & Chapman, G. B. (2007). Long-standing influenza vaccination
policy is in accord with individual self-interest but not with the utilitarian optimum.
PNAS: Proceedings of the National Academy of Sciences of the United States of America,
104, 5692-5697.
Hardin, G. (1968). The tragedy of the commons. Science, 162, 1243-1248.
Hershey, J.C., Asch, D.A., Thumasathit, T., Meszaros, J., Waters, V.V. (1994). The roles of
altruism, free riding, and bandwagoning in vaccination decisions. Organizational
Behavior and Human Decision Processes, 59(2), 177-187.
Jansen, V. A., Stollenwerk, N., Jensen, H. J., Ramsay, M. E., Edmunds, W. J., & Rhodes, C. J.
(2003). Measles out-breaks in a population with declining vaccine uptake. Science, 301,
804.
Liao, Q., Cowling, B. J., Lam, W. W. T., & Fielding, R.(2011). Factors affecting intention to
receive and self-reported receipt of 2009 pandemic (H1N1) vaccine in Hong Kong: A
longitudinal study. PLoS ONE, 6. doi:10.1371/journal.pone.0017713
Manfredi, P., Della Posta, P., d’Onofrio, A., Salinelli, E., Centrone, F., Meo, C., & Poletti, P.
(2010). Optimal vaccination choice, vaccination games, and rational exemption: An
appraisal. Vaccine, 28, 98-109.
Nguyen, T., Henningsen, K.H., Brehaut, J.C., Hoe, E., Wilson, K. (2011). Acceptance of a
pandemic influenza vaccine: a systematic review of surveys of the general public. Journal
of Infection and Drug Resistance, 4, 197-207.
Noar, S. M., Benac, C. N., & Harris, M. S. (2007). Does tailoring matter? Meta-analytic review
of tailored print health behavior change interventions. Psychological Bulletin, 133, 67393.
Reyna, V.F. (2012). Risk perception and communication in vaccination decisions: A fuzzy-trace
theory approach. Vaccine, 30, 3790-3797.
Salgado, C. D., Giannetta, E. T., Hayden, F. G., & Farr, B. M. (2004). Preventing nosocomial
influenza by improving the vaccine acceptance rate of clinicians. Infection Control and
Hospital Epidemiology, 25, 923-928.
Schelling, T. C. (1960). The Strategy of Conflict. Cambridge, MA: Harvard University Press.
Shim, E., Chapman, G.B., Townsend, J.P., Galvani, A.P. (2012). The influence of altruism on
influenza vaccination decisions. Journal of the Royal Society Interface, 9(74), 2234-43.
Smith, C. E. G. (1970) . Prospects of the control of disease. Proceedings of the Royal Society of
Medicine, 63,1181-1190.
Sturm, L. A., Mays, R. M., & Zimet, G. D. (2005). Parental beliefs and decision making about
child and adolescent immunization: From polio to sexually transmitted infections. Journal
of Developmental and Behavioral Pediatrics, 26, 441-452.
Talbot, T. R., Babcock, H., Caplan, A. L., Cotton, D., Maragakis, L. L., Poland G. A., …
Weber, D. J. (2010). Revised SHEA position paper: Influenza vaccination of healthcare
personnel. Infection Control and Hospital Epidemiology, 31, 987-995.
Thomson, A., & Watson, M. (2012). Listen, understand, engage. Science Translational
Medicine, 4.
Van Lange, P. A. M. (1999). The pursuit of joint outcomes and equality in outcomes: Integrative
model of social value orientation. Journal of Personality and Social Psychology, 77, 337349.
Vietri, J., Li, M., Galvani, A. P., & Chapman, G. B. (2012). Vaccinating to help ourselves and
others. Medical Decision Making, 32, 447-458.
Weiner, B. (1980). A cognitive (attribution)-emotion-action model of motivated behavior: An
analysis of judgments of help-giving. Journal of Personality and Social Psychology, 39,
186-200.
Wicker, S., Rabenau, H. F., Doerr, H. W., & Allwinn, R. (2009). Influenza vaccination
compliance among health care workers in a German university hospital. Infection, 37,
197-202.
Footnotes
1
Prior to intention we also assessed the perceived risk of the disease and the
vaccination (general risk [0-100], respectively). We did not expect any effects of
the manipulations on the assessed variables, as we did not provide any
information about vaccine coverage. We used the risk variables as proxies for the
expected costs of the disease and vaccination. The mean perceived risk of the
disease was Mdisease = 41.18 (SD = 23.92); the risk of the vaccination was
Mvaccination = 26.65 (SD = 22.14; t(341) = 7.89, p < .001). This mirrors the
experiment materials, as the disease was described as having severe symptoms,
while potential vaccination side-effects were rather moderate. As a consequence,
100 participants perceived a higher risk of vaccination than of the disease, while
the majority of 237 participants perceived higher disease than vaccination risks;
for 5 participants the costs of disease were equal to the costs of vaccination.
2
Strictly speaking, the pre-conditions for considering vaccination as a social
dilemma are only given for those participants who perceived the costs (risk) of
the vaccination to be higher than the costs (risk) of the disease (see Figure 1). In
support of this, two separate analyses showed that the effects were indeed
consistently stronger if E[cD] - E[cV] < 0 than if E[cD] - E[cV] > 0. The effect
sizes if E[cD] - E[cV] < 0 were η2 = .014 for individual benefit communication
(vs. η2 < .01 if E[cD] - E[cV] > 0) and η2 = .055 (vs. η2 = .001) for the interaction
between social and individual benefit communication. We conclude that this
demonstrates the validity of the proposed model.
Table 1. Means and standard deviations (in brackets) of the intention to vaccinate as a function
of communicated individual and social benefit as well as costs to obtain the vaccination.
individual benefit
social benefit
communicated
not communicated
high vaccination cost
communicated
3.44 (1.86)
3.62 (1.92)
not communicated
3.75 (1.91)
4.20 (1.63)
low vaccination cost
communicated
4.54 (1.50)
4.40 (1.92)
not communicated
3.30 (1.83)
4.42 (1.51)
Figure 1. Simplified general model of vaccination as a strategic interaction.
Note: n refers to the number of all other individuals in the population (N - 1) who decide to
vaccinate. The grey dashed line indicates the expected individual costs of the vaccination
(E[cV]) and the grey solid line indicates the expected individual costs of the disease (E[cD]). The
black dashed line indicates an individual’s expected utility if he/she decides to vaccinate (EUV),
whereas the black solid line indicates the individual’s expected utility if he/she decides not to
vaccinate (EU¬ V). Vc refers to the critical vaccination level (herd immunity threshold) that must
be achieved to eradicate the disease, which is simplified 1 - 1/R0, with R0 being the basic
reproduction number of the infectious disease (Fine et al., 2011). The shaded area indicates a
conflict between individual and collective interest, transforming the vaccination decision into a
N-person prisoner’s dilemma.
Figure 2. Intention to vaccinate as a function of communicated social benefit and communicated
individual benefit (A) or vaccination costs (B).

Inviting Free-Riders or Appealing to Prosocial Behavior? Game

Transcription

Similar documents