1 Introduction
Suppose you were the director of a hospital’s organ allocation team. Your job is to determine who among those patients awaiting organ transplants should be given priority for the organs that are currently available. How would you decide? One possible strategy might be to decide which candidates are “in the running.” That is, you would select (or “include”) the patients who you think should be seriously considered for an immediate organ transplant. The other strategy might be to decide which candidates are “out of the running.” In other words, you would eliminate (or “exclude”) those patients who you think should not be seriously considered for an immediate organ transplant.
Previous research suggests that if you use an inclusion rather than exclusion strategy you will end up with a smaller consideration set, i.e., more candidates will be screened out (Huber, Neale, & Northcraft, Reference Huber, Neale and Northcraft1987; Levin, Huneke, & Jasper, Reference Levin, Huneke and Jasper2000; Levin, Jasper, & Forbes, Reference Levin, Jasper and Forbes1998; Westenberg & Koele, Reference Westenberg, Koele, Borcherding, Larichev and Messick1990, Reference Westenberg and Koele1992; Yaniv & Schul, Reference Yaniv and Schul1997), and you will expend more effort (Levin et al., Reference Levin, Huneke and Jasper2000). Research also suggests that, under certain contexts, decision makers may have a natural preference for one strategy over the other (Heller, Levin, & Goransson, Reference Heller, Levin and Goransson2002; Levin, Prosansky, Heller, & Brunick, Reference Levin, Prosansky, Heller and Brunick2001). However, research has not fully explicated the conditions under which decision makers choose to include or exclude nor has it identified individual differences predictive of strategy selection.
1.1 Preference for inclusion/exclusion strategies
While the effects of inclusion/exclusion instructions on set size has been studied for roughly fifteen years, the preference that decision makers have for one strategy over the other is a relatively new phenomenon. This issue was raised initially by Ordóñez, Benson, Lehman, and Beach (Reference Ordóñez, Benson, Lehman and Beach1999). By comparing the number of options screened out under inclusion, exclusion, and control conditions, these researchers concluded that the normal screening process appeared to be to screen out bad options rather than to screen in good ones. Levin and his colleagues, however, questioned this conclusion. They asked subjects to select the strategy that seemed more “natural” to them in either a positive task of hiring someone or a negative task of firing someone. Results indicated that 81% of subjects in the hiring task chose inclusion as the manner in which they screened hypothetical job applicants, while 61% of subjects in the firing task chose an exclusion strategy (Levin, Prosansky, Heller, & Brunick, Reference Levin, Prosansky, Heller and Brunick2001). In a series of followup experiments, they were lead to conclude that strategy preference depends on the characteristics of the task. For example, Levin et al (Reference Levin, Prosansky, Heller and Brunick2001) revealed that for judgments involving either the “addition” or “deletion” of stocks, the majority of participants performing each task selected inclusion as the preferred strategy (70% and 74%, for addition and deletion of stocks, respectively). Inclusion was also the preferred strategy in a personal judgment task (e.g., Which of the following cities has influenced the U.S. economy and culture the most?) (Heller, Levin, & Goransson, Reference Heller, Levin and Goransson2002, Exp. 2), while exclusion was the strategy of choice for both verbal and quantitative decisions including a correct answer (e.g., Which of the following fractions has the greatest value?) (Heller, Levin, & Goransson, Reference Heller, Levin and Goransson2002, Exp. 1).
Levin and his colleagues have also attempted to identify individual differences predictive of strategy choice. Spurred by recent successes in using need-for-cognition to predict depth and breadth of information processing (Levin, Huneke, & Jasper, Reference Levin, Huneke and Jasper2000) and the Big-Five personality traits to predict differential reactions to positive and negative stimuli (Levin, Schreiber, Lauriola, & Gaeth, Reference Levin, Schreiber, Lauriola and Gaeth2002), Heller, Levin, and Goransson (Reference Heller, Levin and Goransson2002, Exp. 2) incorporated a variety of personality and cognitive style measures into their inclusion/exclusion strategy selection work. However, they found nothing. Individual differences in the Big-Five personality traits (neuroticism, extraversion, openness to experience, agreeableness, and conscientiousness), need-for-cognition, and faith in intuition, both separately and jointly in statistical analyses, were not predictive of the strategy selected.
One might use these results to argue that although there may be task or context differences in strategy selection, there appears to be no inherent individual differences. However, this line of research is still in its infancy, and obviously there are a number of individual difference factors yet to be studied. One such factor is political orientation. This individual difference variable has been all but neglected in decision making, but has important implications for social psychology and theories of distributive justice.
1.2 Political orientation and resource allocation
Allocating resources is a decision made everyday in the real world. Employers allocate jobs, wages, and rewards to employees; government policymakers allocate time, money, and personnel to various tax-funded social programs related to welfare, job training, education, and healthcare; and, as seen in our example, physicians and/or hospital administrators decide who gets an organ transplant and who does not via the use of a point system.
In psychology, one of the more intriguing questions is how does an ordinary person go about allocating resources? At the heart of many of the debates — both theoretical and empirical — is the widely assumed tradeoff between equality and efficiency. On the one hand, one wants to be fair, but on the other hand it’s important to get the most “bang for the buck.” Other factors that come into play include such things as the availability of the resource, the claimants perceived need, the attributions about the origin of social problems, and the perceived causes of one’s predicament (i.e., attributions of responsibility).
In the early 1990s Skitka and Tetlock (Reference Skitka and Tetlock1992, Reference Skitka and Tetlock1993) combined these elements together into a 4-stage contingency model of distributive justice in an effort to describe individual differences in resource allocation decisions. Specifically, they explored judgments of fairness in a variety of situations and consistently found that, when resources were abundant, i.e., there were sufficient resources available to help all whom required it, liberals were more likely than conservatives to help targets with internal and controllable causes for needing assistance, i.e., those who were personally responsible. More recent research from Skitka and colleagues has been aimed at explaining these ideological differences. Specifically, this work has tried to shed light on the cognitive strategies and motivational priorities of liberals and conservatives. Skitka (Reference Skitka1999), for example, has found that liberals and conservatives appear to see the world in relatively similar ways, and seem to be equally likely to make first pass personal attributions for the causes of others’ actions or problems. Liberals, like conservatives, for instance, perceive victims as responsible for their plight and report being angered about it. However, liberals and conservatives diverge in their attributions when these first pass judgments conflict with their ideological values or goals (Skitka, Mullen, Griffin, Hutchinson, & Chamberlin, Reference Skitka, Mullen, Griffin, Hutchinson and Chamberlin2002). Because of their egalitarian beliefs, liberals are motivated to correct the normal attribution-affect-action sequence and, in so doing, go out their way to try to generate sympathy for and find reasons to help the irresponsible.
Evidence to support this “motivated correction hypothesis” can be found in two studies. In the first, Skitka et al. (Reference Skitka, Mullen, Griffin, Hutchinson and Chamberlin2002, Exp. 5) showed that interfering with liberals’ ability to engage in systematic processing eliminated ideological differences in willingness to help the personally responsible. Specifically, Skitka et al. found that liberals were more willing than conservatives to help individuals who were personally responsible for their plight but only when they had the cognitive resources available to override the logical consequences of the attribution-affect-action sequence. When liberals were placed under conditions of high cognitive load, their ability to generate sympathy toward those personally responsible was impeded, and, therefore, the differences between liberals and conservatives disappeared.
In the second study, Skitka, Jasper, and Mullen (Reference Skitka, Jasper and Mullen2001) carefully examined the cognitive, decision-making processes that yield these differential patterns of helping behavior. Their thinking was as follows: if liberals were in fact correcting an initial personal attribution to make their reactions more consistent with personal standards, the process involved should be cognitively more effortful and demanding; in contrast, because attributional analyses and personal values lead to consistent conclusions, conservatives have no reason to correct the logical consequences of the attribution-affect-action sequence, and therefore, should engage in a less effortful process. Using a process tracing method developed by Jasper and Shapiro (Reference Jasper and Shapiro2002), Skitka et al. (Reference Skitka, Jasper and Mullen2001) confirmed this line of thinking. Liberals acquired more information about organ candidates (some of whom were personally responsible for organ damage), searched that information in greater depth and breadth, and took more time to make their decision than conservatives. This was true, however, only under conditions of scarcity, i.e., when there were more candidates than organs available. Under conditions of abundance, conservatives tended to exert more effort than liberals, although most of these differences were not significant.
1.3 The present study
So, what does this have to do with one’s preference for inclusion/exclusion strategies? First, allocating scarce resources seems to be a natural context in which one might choose an inclusion or exclusion strategy to narrow the original set of options. Second, it would seem that we have preliminary data indicating the choice of particular strategies. Based on our earlier review of the literature, we know that inclusion can be a more effortful strategy than exclusion (Levin, et al, Reference Levin, Huneke and Jasper2000). We also know from Skitka, et al (Reference Skitka, Mullen, Griffin, Hutchinson and Chamberlin2001) that liberals expend much more effort than conservatives under scarcity, while conservatives expend more effort than liberals under abundance. It appears then that the natural strategy (under abundance) for liberals is exclusion, while the opposite may be true for conservatives, i.e., they prefer inclusion. This would be consistent with Yaniv, Schul, Raphaelli-Hirch, and Maoz (Reference Yaniv, Schul, Raphaelli-Hirch and Maoz2002) who argued that individuals with an exclusion mindset have a more liberal (lax) decision criterion and Christiansen and Levine (Reference Christiansen and Lavine1997) who argued that liberals have a more liberal decision criterion in these situations than conservatives. According to Christiansen and Levine, this is because liberals are more concerned about misses, whereas conservatives are more concerned about false alarms. Interestingly enough, Christiansen and Levine also argue that liberals are more sensitive to the situational constraints and may be more likely than conservatives to change strategies. This then may mean that liberals are more likely to switch from exclusion to inclusion than conservatives are to switch from inclusion to exclusion when going from abundance to scarcity, respectively. The present study was designed to test these hypotheses. It was also designed, as a secondary purpose, to replicate four other findings, namely: 1) exclusion strategies lead to larger set sizes than inclusion, 2) abundant resource conditions result in larger set sizes than scarce resource conditions, 3) internal uncontrollable targets are chosen more often than internal controllable targets, and 4) liberals tend to choose more targets (particularly internal controllable targets) on average than conservatives.
2 Method
2.1 Design
The experiment used a 2 x 2 between-subjects design. The first variable was resource scarcity: scarce or abundant. The second variable was political orientation: liberal vs. conservative.
2.2 Participants
Seventy-six undergraduate students at the University of Toledo participated in the experiment. All students were enrolled in introductory psychology courses and earned extra credit for their participation. Thirty-eight subjects were randomly assigned to each scarcity condition.
2.3 Stimuli
Participants were presented with 16 organ candidates. Candidates were constructed to vary as a function of a 2 (Locus of control: Internal-controllable, internal-uncontrollable) X 2 (Contribution to society: High, low) X 2 (Need: High, low) X 2 (Efficiency: High, Low) factorial design.Footnote 1 Each of these manipulations is described in more detail below.
Locus of Control (labeled “Cause of Organ Damage”)Footnote 2
Internal Controllable: Despite a doctor’s repeated warnings about the damaging effects for this person’s health and the probability of severe organ damage, this person continued to eat high cholesterol foods, smoke, and not exercise. Consequently, this person now has severe organ failure and requires an organ transplant.
Internal Uncontrollable: This person has a genetically defective organ that cannot be corrected by any medical means to date. The organ has deteriorated to the state of severe organ failure and this person requires an organ transplant.
Contribution to Society
High: This person is actively involved in volunteer, civic, and charity activities.
Low: This person is not very actively involved in volunteer, civic, and charity activities.
Need (labeled “Urgency”)
High: This person has an 80% probability of dying before another organ is likely to become available.
Low: This person has a 50% probability of dying before another organ is likely to become available.
Efficiency (labeled “Life Expectancy w/ Transplant”)
High. This operation is expected to extend this patient’s life by 10 years.
Low. This operation is expected to extend this patient’s life by 2 years.
2.4 Procedure
Participants in each condition were given an experimental booklet which included a cover story describing that the study’s purpose was to examine the strategies that people use in narrowing down their options in medical situations. Specifically, they were told to “Imagine that you are the director of a hospital’s organ allocation team. It is your job to determine who among those patients awaiting organ transplants should be given priority for the organs that are currently available. There are 16 patients who require an organ transplant. However, you’re not absolutely sure how many organ are available.” Subjects in the scarce resource condition were told “You do know though that organs are currently very scarce. The latest estimates that you have indicate that somewhere between 3 and 5 organ are currently available; however, that number could be higher or it could be lower.” Subjects in the abundant resource condition were told “You do know though that organs are currently very abundant. The latest estimates that you have indicate that somewhere between 14 and 16 organs are currently available; however, that number could be higher or it could be lower.”
In addition to the experimental booklet, each participant was given an envelope containing 16 shuffled cards on which the organ candidates, identified by the numbers 1 through 16, appeared. Participants were given instructions to open their envelope and quickly look through all the cards and then pick one of two strategies (inclusion or exclusion) that they would use to complete the task. Inclusion was described as an available strategy in which “You could select patients who you think should be seriously considered for an immediate organ transplant.” Exclusion was described as an available strategy in which “You could eliminate (or rule out) the patients who you think should not be seriously considered for an immediate organ transplant.” Examples of both strategies were presented within the context of choosing from amongst a number of over-the-counter allergy medications.
After subjects had selected a strategy to use throughout the task, they were required to write down the word INCLUSION or EXCLUSION on the space within the following sentence: I would like to use the ___________ strategy. Participants were then asked to go back and look through the 16 cards carefully. Those who chose inclusion were instructed to circle the patients (listed by number in the booklet) they wanted to include; those who chose exclusion were instructed to cross out the patients they wanted to exclude.
Following the choice task, participants in each condition completed a questionnaire designed to measure their political orientation (or ideology). After completing all materials, participants were thanked and debriefed. Average time to complete the entire experiment was 20 minutes.
2.5 Political Orientation
Political orientation was measured using the short-form version of Kerlinger’s (Reference Kerlinger1984) Social Referent Scale. Fourteen items on the scale were liberal items (e.g., free abortion, socialized medicine, racial equality, freedom, social change) and the other 14 items were conservative items (e.g., free enterprise, law and order, faith in God, discipline, moral standards). Previous research in our lab has shown that responses to the conservative items are better predictors of differences in resource allocation decisions than responses to the liberal items. Therefore, only the 14 conservative items were scored. Participant responses to each item were made on a scale from +3 (agree very strongly) to −3 (disagree very strongly). Summing the responses to the 14 items defined our measure of political orientation; the range of possible scores was −42 to +42. Subjects scoring +23 or above scored high on the conservative subscale and were classified as “conservatives” (n=42), while subjects scoring 22 or below scored low on the conservative subscale and were classified as “liberals” (n=34). Political ideology is dichotomized for descriptive purposes. However, for all statistical analyses we treated political ideology as a continuous measure.
3 Results
Data will be reported and discussed separately for each of the following dependent measures: strategy choice, set size, and set content. The first measure tests the primary hypothesis that strategy choice varies as a function of scarcity condition and political ideology. The second and third measures assess the secondary hypotheses related to the number of candidates chosen under inclusion and exclusion and the composition of those candidate sets in terms of internal and external controllability (i.e., personal responsibility for damaging their own organs).
3.1 Choice of Strategy
Table 1 displays the number and proportion of participants who chose each strategy: inclusion or exclusion. These data were submitted to a 2 (condition) x 2 (ideology) x 2 (strategy) Chi-Square test of association. There was no clear preference, overall, for either strategy. Forty subjects (53%) chose to use an inclusion strategy while the other 36 (47%) chose to use exclusion. There was also no difference in preference between subjects receiving the scarce and abundant resource scenarios. Half the subjects (19 out of 38) chose inclusion to screen organ candidates under the condition of abundance; the other half chose exclusion. Under the condition of scarcity slightly more than half the subjects (21 out of 38) selected inclusion as the preferred strategy. Finally, there was no difference in strategy preference, overall, between liberals and conservatives. Roughly half of those designated as liberals (47%) chose inclusion; the other half (53%) chose exclusion. The same was true of conservatives, although there was a slightly larger preference for inclusion (57%) over exclusion (43%).
For the purposes of this paper, the most interesting finding was that liberals and conservatives differed in their preferences across scarcity conditions. Seventy-nine percent of liberals, but only 33% of conservatives opted to use exclusion under conditions of abundance, χ2 (1, n=38) = 7.24, p = .0071. In contrast, under scarcity the majority of liberals (65%) preferred an inclusion strategy, whereas slightly less than half (44%) of conservatives chose an inclusion strategy. The interaction was significant by a logistic regression predicting strategy choice from scarcity condition and political ideology (χ2−1=6.90, p=.0086).
3.2 Set Size
Table 2 presents the mean number of organ candidates chosen by liberals and conservatives under conditions of scarcity and abundance. In inclusion, number of candidates chosen (or set size) was defined as the number of options chosen. For exclusion, set size was defined as the number of options remaining or not crossed out.
As predicted, the mean set size under abundance (M = 10.35) was larger than the mean set size under scarcity (M = 6.49). This was seen statistically in a significant main effect of condition, F (1, 66) = 53.79, p < .0001, when the data were submitted to a three factor — condition x ideology x strategy — analysis of variance (ANOVA). As predicted, an exclusion strategy also led to larger set sizes (M = 8.40) than an inclusion strategy (M = 8.15), F (1,66) = 4.92, p = .0300. However, contrary to predictions, there were no differences in mean set size between conservatives (M = 8.11) and liberals (M = 8.44), and there was no interaction between condition and ideology. In the first case, although non-significant (F (1, 66) = .80, p = .3737), the effect of ideology was in the predicted direction; liberals tended to choose more organ candidates on average than did conservatives. In the second case, as predicted, we found that liberals tended to choose more candidates than conservatives under both scarcity conditions; the difference between ideologies, however, was not larger under abundance as seen in a non-significant scarcity condition X ideology interaction, F (1, 66) = .57, p = .4542.
A final effect worth noting was a significant interaction between ideology and strategy, F (1, 66) = 5.54, p = .0216. Liberals who chose to use exclusion had a larger mean set size (M = 9.17) than conservatives (M = 7.63) who chose exclusion. The opposite was true of those who chose inclusion. Conservatives (M = 8.59) had larger set sizes than liberals (M = 7.71). Put another way, an inclusion strategy resulted in larger set sizes for conservatives, while an exclusion strategy resulted in larger set sizes for liberals. Thus, our prediction of larger set sizes in exclusion rather than inclusion was true, but only for participants in our experiment who were designated as liberals. A follow-up analysis with only liberal subjects confirmed this finding statistically, t(31) = 3.05, p < .01.
3.3 Type of Candidates Chosen
Table 3 reveals the number of internal controllable (personally responsible) and internal uncontrollable (not personally responsible) candidates chosen by liberals and conservatives across conditions of abundance and scarcity. As predicted, subjects chose more internal uncontrollable (IU; M = 5.29) than internal controllable (IC; M = 3.15) candidates for immediate organ transplant. This was borne out in a significant main effect of candidate type, F (1,70) = 4.99, p = .0287, when the data were submitted to a three factor — condition x ideology x candidate type — ANOVA.
However, contrary to predictions, there was no difference between conservatives and liberals, and there was no interaction between condition and ideology. As discussed earlier, liberals tended to choose more organ candidates on average than did conservatives; nevertheless, the difference between ideologies was non-significant, F (1,70) = .42, p = .5181. In terms of type, liberals chose more internal uncontrollable (IU) candidates than did conservatives under both abundance (M = 6.00 and 5.91, respectively) and scarcity (M = 4.74 and 4.44, respectively). Liberals also chose more internal controllable (IC) candidates than did conservatives under abundance (M = 4.57 and 4.30, respectively), but not under scarcity (M = 1.84 and 1.94, respectively). However, there was no two-way, F (1,70) = .26, p = .6118, or three-way, F (1,70) = .12, p = .7325, interaction. Finally, although subjects chose more IU than IC candidates under both abundance (M = 5.96 and 4.44, respectively) and scarcity (M = 4.59 and 1.89, respectively), the difference between the two candidate types was much larger under scarcity; the interaction, however, between candidate type and scarcity condition was not significant, F (1,70) = 1.80, p = .1841.
Because some might argue (and rightly so) that the dependent variable is scaled differently (i.e., it potentially has different ranges) across the two scarcity conditions (which then makes comparisons between internal controllable and internal uncontrollable candidates tricky), we also ran two additional analyses. Specifically, we ran separate ideology x candidate type ANOVAs for each scarcity condition. The results of these 2-factor analyses were very similar to our 3-factor analysis. However, there are two effects worth noting. First, the separate ideology X candidate type ANOVAs for each scarcity condition indicated that the main effect of candidate type (seen earlier) was driven primarily by the scarce, F (1,35) = 7.60, p = .0092, rather than the abundant, F (1,35) = .35, p = .5589, condition. Second, in this area of research, one typically sees no difference in liberals’ willingness to help IC and IU claimants under abundance, but a reluctance to help IC relative to IU claimants on the part of conservatives. In other words, there is usually a two-way interaction between ideology and candidate type showing that the difference between the number of IC and IU claimants chosen is smaller for liberals than conservatives. In short, while the data descriptively support this effect under abundance, the interaction between ideology and candidate type was, nevertheless, non-significant, F (1,35) = 1.20, p = .2800.
4 Discussion
The primary goal of this study was to explore individual differences in strategy selection. Although previous research has found that the task or context of decision-making can affect one’s choice of inclusion/exclusion strategies, no one has ever revealed an individual difference factor predictive of strategy preference. We were encouraged in our pursuit by the knowledge that research in resource allocation and distributive justice (Skitka et al., Reference Skitka, Jasper and Mullen2001) has found that liberals tend to take more time and expend more effort in allocating resources to those who are irresponsible under conditions of scarcity. Conservatives, on the other hand, tend to take more time and expend more effort under conditions of abundance. Assuming that inclusion is a more effortful process than exclusion (Levin et al., Reference Levin, Huneke and Jasper2000), we surmised that the majority of liberals were probably “including” under scarcity and “excluding” under abundance, while the opposite was true of conservatives. A possible explanation for this was offered by Christiansen and Levine (Reference Christiansen and Lavine1997). They suggested that conservatives will normally use stringent inclusion rules for allocating aid. According to these researchers they do so to increase efficiency by minimizing waste; indeed, they’re worried about false alarms. In contrast, they suggest that liberals will set more liberal inclusion rules (arguably by using an exclusion strategy) to ensure that everyone who needs resources has access to them if at all possible. It was also suggested by Christiansen and Levine that conservatives probably allocate resources in a relatively similar manner regardless of the scarcity or abundance of the resource, whereas liberals will likely have to switch their strategy to compensate for their default position of helping as many as possible. In fact, this is what we have found.
Participants were asked to narrow down a list of 16 potential organ recipients after 1) reading information about each candidate’s medical urgency, life expectancy after transplant, contribution to society, and cause of organ damage, and 2) choosing a strategy of either inclusion or exclusion. The results confirmed our predictions in showing that liberals overwhelmingly chose to use exclusion under abundance and inclusion under scarcity. In contrast, conservatives tended, if anything, to prefer an inclusion strategy under abundance and exclusion under scarcity. Put another way, under abundance conservatives chose inclusion and liberals chose exclusion, and under scarcity conservatives slightly preferred exclusion, while liberals switched to inclusion.
The results also replicated previous research in showing that abundant resource conditions result in larger set sizes than scarce resource conditions, exclusion strategies lead to larger set sizes than inclusion strategies, and internal uncontrollable (non-personally responsible) targets are chosen more often than internal controllable (personally responsible) targets.
We also predicted that liberals would choose more targets (particularly internal controllable targets) on average than conservatives; although the findings were in the hypothesized direction, the results for these later predictions were not significant. We speculate that these non-significant effects had something to do with our design. Specifically, our manipulation of personal responsibility was not particularly strong. We suspect telling subjects that some targets led a non-healthy lifestyle (by eating high cholesterol foods and not exercising) is far less affectively charged than telling them that targets contracted AIDS from homosexual contact and now need immunosuppressant drugs, crashed their car because they were drunk and now need emergency medical assistance, or refused to look for work and now need help from the government. This then led to a less visceral, punitive reaction from conservatives. It could also be the case that our sample — college students — is much more egalitarian and much less judgmental in their views than other samples which include a wider range of ages, incomes, occupational histories, and life experiences.
This research is important for three reasons. First, it extends research on inclusion/exclusion strategy preference. Previous research has recently identified some task or context differences in strategy selection. But, as far as we know, this is the first time anyone has ever found an individual difference factor predictive of inclusion/exclusion strategy preference. Second, it adds to what we know about individual differences in decision making. We think that real progress in understanding and theorizing about decision behavior will come only when we use individual difference indices to account for significant variation in our data. The notion that political orientation might be a wide-ranging, stable predictor of other decision behavior is suspect. However, it does seem to predict, and thus, give us a better understanding of, effects in resource allocation. Finally, these results may help to explicate models of distributive justice. The finding that liberals choose an inclusion strategy under scarcity while conservatives choose the same strategy under abundance maps nicely onto what we know about their initial mindset and later cognitive processing, i.e., how they search through information and how much effort they expend in doing so. Nevertheless, these data may not support the motivated correction hypothesis, unless changing ones strategy (and by default) decision criterion is a better indicator of “correction” than information search processes. It could also be the case that conservatives and liberals are using inclusion and exclusion strategies in different ways or for different reasons. Only time and future research will tell.
Open access
