When it comes to the motivations behind redistribution, it turns out that fairness has little to do with it. Instead, researchers find1 that compassion, envy, and self-interest are the main drivers. From the abstract:
Why do people support economic redistribution? Hypotheses include inequity aversion, a moral sense that inequality is intrinsically unfair, and cultural explanations such as exposure to and assimilation of culturally transmitted ideologies. However, humans have been interacting with worse-off and better-off individuals over evolutionary time, and our motivational systems may have been naturally selected to navigate the opportunities and challenges posed by such recurrent interactions. We hypothesize that modern redistribution is perceived as an ancestral scene involving three notional players: the needy other, the better-off other, and the actor herself. We explore how three motivational systems—compassion, self-interest, and envy—guide responses to the needy other and the better-off other, and how they pattern responses to redistribution. Data from the United States, the United Kingdom, India, and Israel support this model. Endorsement of redistribution is independently predicted by dispositional compassion, dispositional envy, and the expectation of personal gain from redistribution. By contrast, a taste for fairness, in the sense of (i) universality in the application of laws and standards, or (ii) low variance in group-level payoffs, fails to predict attitudes about redistribution.
Let’s dive into the details:
We conducted 13 studies with 6,024 participants in four countries to test the hypothesis that compassion, envy, and self-interest jointly predict support for redistribution. Participants completed instruments measuring their (i) support for redistribution; (ii) dispositional compassion; (iii) dispositional envy; (iv) expected personal gain or loss from redistribution (our measure of self-interest); (v) political party identification; (vi) aid given personally to the poor; (vii) wealthy-harming preferences; (viii) endorsement of procedural fairness; (ix) endorsement of distributional fairness; (x) age; (xi) gender; and (xii) socioeconomic status (SES) (pg. 8422).
To test this prediction, we regressed participants’ support for redistribution simultaneously on their dispositional compassion, their dispositional envy, and their expected personal gain (or loss) from redistribution. As predicted, the three motives have positive, significant, and independent effects on support for redistribution. This is true in the four countries tested: the United States (US) (study 1a), India (IN) (study 1b), the United Kingdom (GB) (study 1c), and Israel (IL) (study 1d)—standardized regression coefficients (β values): compassion, 0.28–0.39; envy, 0.10–0.14; self-interest, 0.18–0.30. Jointly, these motives account for 13–28% of the variance in support for redistribution. Adding to the regression models age and gender, or age, gender, and S[ocio]E[conomic]S[tatus], does not appreciably alter the effect of the emotion/motivation triplet, or the total variance accounted for. We note that age did not have significant effects in any country. Gender had significant effects in the United States and the United Kingdom (females more opposed to redistribution), but not in India or Israel. SES had a significant (negative) effect in the United Kingdom, but not in the other countries (Ibid.).
Unsurprisingly, in the U.S. “self-described Democrats endorsed redistribution to a greater extent than Republicans and Libertarians did. Democrats also reported more compassion and more expected personal gain from redistribution than Republicans and Libertarians did; envy did not differ by party” (Ibid.). Interesting enough, “dispositional compassion was the only reliable predictor of giving aid to the poor” in all four countries. However, “support for government redistribution was not a unique predictor of personally aiding the poor in the regressions…Support for government redistribution is not aiding the needy writ large—in the United States, data from the General Social Survey indicate that support for redistribution is associated with lower charitable contributions to religious and nonreligious causes” (Ibid.).
Now consider these absurd answers from the survey respondents:
Participants in the United States, India, and the United Kingdom (studies 1a–c) were given two hypothetical scenarios and asked to indicate their preferred one. In one scenario, the wealthy pay an additional 10% in taxes, and the poor receive an additional sum of money. In the other scenario, the wealthy pay an additional 50% in taxes (i.e., a tax increment five times greater than in the first scenario), and the poor receive (only) one-half the additional amount that they receive in the first scenario. That is, higher taxes paid by the wealthy yielded relatively less money for the poor, and vice versa (63). To clarify the rationale for this trade-off, we told participants that the wealthy earned more when tax rates were low, thereby generating more tax revenue that could be used to help the poor. Fourteen percent to 18% of the American, Indian, and British participants indicated a preference for the scenario featuring a higher tax rate for the wealthy even though it produced less money to help the poor…We regressed this wealthy-harming preference simultaneously on support for redistribution, the emotion/motivation triplet, age, gender, and SES. Dispositional envy was the only reliable predictor (Ibid.).
In short, “Compassion and envy motivate the attainment of different ends. Compassion, but not envy, predicts personally helping the poor. Envy, but not compassion, predicts a desire to tax the wealthy even when that costs the poor” (Ibid.). The cries for fairness, though, have little to do with support for redistribution:
To sum up the set of fairness studies, in predicting support for redistribution, the effect of fairness as a group-wide concern is unreliable and of far smaller magnitude than the effect of the emotion/motivation triplet. This is true whether fairness is operationalized as uniformity in the application of laws and standards or as low (or null) variance in payoffs; whether distributional fairness is assayed between individuals (studies 2a, 2c, S1a, S2a, S2b) or between groups (“the rich,” “the poor”; studies 2b, S1b, S2c, S2d); and whether allocational decisions are hypothetical (studies 2a, 2b, S1a, S1b, S2c, S2d) or consequential (studies 2c, S2a, S2b) (pg. 8423).
So, why do people support redistribution?
- They care about the poor and want to help them (Compassion).
- They hate the rich (Envy).
- They expect to gain from redistribution (Greed).
Sounds about right.