Death and taxes may be inevitable, but it is only the latter which gather vocal apologists. Tax, these enthusiasts aver, is a good thing because it makes good things happen: hospitals, schools, roads, clean water and government inspectors maintaining standards.

Certainly, taxes may provide services at better standards and lower costs than could be obtained by individuals acting on their own. Rubbish collection is one example, military defence another. It is very probably true of public health, where small contributions can obtain considerable communal benefits. However, in the UK only a minor proportion of taxes are spent on such matters. Healthcare, education and defence account for 36% of spending. Another 36% is taken up by welfare and pensions, which in older times were considered a personal responsibility. The last third is general government activities. For example, “Protection” relates to police, fire services, courts and jails. “Interest” is the cost of government spending more than it raises in taxes. The “Other” category covers a very wide range of interventions, some 28 in total, several with welfare connections.

Proponents of higher taxes might be motivated purely by the overall advantage of government procurement of services as opposed to private provision. On the other hand, it might be due to the wish to obtain benefits. Could other motives explain support for higher taxation?

Chien-An Lin and Timothy C. Bates at the University of Edinburgh decided to find out.

Each is to count for one and none for more than one: Predictors of support for economic redistribution

https://psyarxiv.com/3jq4c/

There is a great deal in this paper, so I have had to summarize, and to concentrate on the main findings, particularly of the first study, and not the second replication.

They recruited a representative sample and then gave them questionnaires to complete. They set standards so that anyone who answered quickly and without thought was excluded (no one did that). They tested every questionnaire for consistency (Cronbach Alpha) and all of them are sufficiently consistent, so the answers are not distorted by a few freak questions. I have put in all the detail because these scales are not well known, so examples are helpful.

A total of 403 participants were recruited using Prolific Academic (268 females, mean age 37 years, SD = 12.19). We pre-registered a criterion that subjects who completed the questionnaire less than 20 seconds would be excluded. No subjects met this criterion. The racial mix of the sample was representative, with participants identifying as White (n = 366; 90.8%), Black (n = 14; 3.5%), Mixed (n = 14; 3.5%), Asian (n = 6; 1.5%) and other (n = 1; 0.2%), 2 participants (0.5%) chose not to answer.

Attitudes toward redistribution were measured with the 11-item support for economic redistribution scale Sznycer et al. (2017). An example reverse-scored item is “Wealthy people should not be taxed more heavily than others”. Each item used a Likert response scale from 1 (strongly disagree) to 5 (strongly agree). The Cronbach Alpha of economic redistribution in our sample was 0.90.

Communal fairness and instrumental harm were measured using the Oxford Utilitarianism Scale (Kahane et al., 2018). This 9-item instrument consists of two subscales: Impartial Beneficence, which we use to assess communal fairness; An example item is “It is just as wrong to fail to help someone as it is to actively harm them yourself”) and Instrumental Harm (example item: “It is morally right to harm an innocent person if harming them is a necessary means to helping several other innocent people”). Scores were on a Likert scale from 1 (strongly disagree) to 7 (strongly agree). In our sample, Cronbach Alphas were 0.63 and 0.69 for Communal fairness and Instrumental Harm.

Compassion, envy, and self-interest were measured as in Lin and Bates (2021).

The 10-item dispositional compassion scale Goldberg (1999); Sznycer et al. (2017) reliably (Cronbach Alpha = 0.80 in our sample) assesses compassion based on Likert responses from 1 (very inaccurate) to 5 (very accurate) to content such as “I suffer from others’ sorrows”.

Self-interest used a single item: “Imagine that a policy of higher taxes on the wealthy is implemented. What overall impact do you think the higher taxes on the wealthy would have on you?” with responses on a 1 to 5 scale: My own economic situation would 1: significantly worsen; slightly worsen; stay the same; slightly improve; 5 significantly improve.

The 5-item Malicious Envy Scale (Lange & Crusius, 2015) scores items from 1 (strongly disagree) to 6 (strongly agree) with example content including “If other people have something that I want for myself, I wish to take it away from them”. The Cronbach Alpha of Malicious Envy was 0.80 in our sample.

Wealthy-harming preference was measured using a scenario choice Sznycer et al. (2017). Scenario one (wealth harming) was “The top 1% wealthiest individuals pay an extra 50% of their income in additional taxes, and as a consequence of that the poor get an additional £100 million per year (the extra 50% in taxes paid in former fiscal years leaving the wealthiest with relatively less taxable income)”.
Scenario two (helping the poor) was “The top 1% wealthiest individuals pay an extra 10% of their income in additional taxes, and as a consequence of that the poor get an additional £200 million per year (the extra 10% in taxes paid in former fiscal years leaving the wealthiest with relatively more taxable income)”.

Finally, support for coercive redistribution was measured with a 19-item coercive redistribution scale generated for this study (see supplementary material detailing development of this scale and the refined, 5-item short version used in study 2). Example items include “People questioning redistribution of wealth should be punished” and “If the wealthy try to avoid tax, it would be permissible to use mild torture to reveal the money they are hiding from the poor”. Responses were on a Likert scale from 1 (strongly disagree) to 5 (strongly agree). Factor scores on the first component of a factor analysis of the 19-item coercive redistribution scale were used to score subjects.

So, sex and compassion do not have significant effect on whether respondents are willing to be coercive to achieve redistributive taxation. These “mild torturers” were motivated mostly by malicious envy, instrumental harm, self-interest and (least of all) communal fairness. The paper looks at the basis of “communal fairness” in more detail, and it has a sting in the tail, in that you have to be accepted into the commune before you can benefit from the proclaimed fairness. For example, communal fairness is a good explanation of honour killings: they are justified because they preserve the purity of the commune.

All these studies together account for over 40% of the variance in support for redistribution, more than achieved in any previous study.

In summary, not all requests for redistributive taxation arise from noble motives.

James Thompson