Rawls argues that parties in the original position would draw upon a “thin liberal theory of the good”, whereby the parties’ most compelling concern would be to secure the material and social conditions that enabled them to exercise their ‘two moral powers’. That is, parties behind the veil would seek to ensure their capacity to form and pursue a conception of the good, and their capacity for a sense of justice. Rawls argues that as a result, parties will evaluate potential conceptions of justice in terms of “basic” liberties – liberties that are essential to the development and the exercise of the ‘two moral powers’. Such liberties include liberty of conscience and freedom of thought, freedom of association, and the rights and liberties that determine the freedom and integrity of the person. Also included are equal political rights of participation and the rights and liberties that maintain the rule of law. Accordingly Rawls proposes his first principle of justice; the principle of equal basic liberties:
First principle of justice: “Each person is to have an equal right to the most extensive scheme of equal basic liberties compatible with a similar scheme of liberties for others”.
Following this Rawls develops the idea of equality of opportunity. Behind the veil of ignorance, Rawls believes, parties would seek to minimize the initial differential in advantages arbitrarily assigned to individuals in society. Furthermore, since Rawls believes that absolute equality of opportunity in this respect is not practically achievable, parties would seek to impose remedial steps whereby inequalities are minimized by benefiting the least advantaged members of society (known as the ‘difference principle’). Hence:
Second principle of justice: “Social and economic inequalities are to satisfy two conditions: first, they are to be attached to positions and offices open to all under conditions of fair equality of opportunity; and second, they are to be to the greatest benefit of the least advantaged members of society”.
Thus Rawls argues that parties will evaluate potential social contracts in terms of ‘primary goods’, those goods “which a rational man wants whatever else he wants”. The notion of primary goods encapsulates the notion of basic liberties given above, in addition to the powers and prerogatives of jobs, income, wealth, and the social bases of self-respect. Before proceeding, it is worth noting that Rawls considers the principles above to be lexically ordered. Basic liberties can only be limited in order to maintain other basic liberties. To quote Samuel Freeman: “Because of their role in defining the conception of moral persons that underlies Rawls’s view, justice as fairness assigns the basic liberties strict priority over other social goods”.
Many criticisms of Rawls’s ‘Theory of Justice’ have focused on the construction of the original position; it will elucidate the impending discussions to investigate this further. As an arena for deliberation and decision concerning different principles of justice, the original position can be understood as performing two distinct roles. The first of these is to act as a screening device for competing conceptions of justice, and the second is to rank conceptions of justice by their acceptability to the parties behind the veil. Since we are interested in whether Rawls is justified in his derivation of the ‘two principles of justice’, it is the screening role that concerns us here. Rawls believes that there are certain formal constraints that it is reasonable to impose on candidate conceptions of justice. The five formal constraints Rawls imposes are those of generality, universal applicability, publicity, that the conception of right must impose an ordering on conflicting claims (i.e. must be complete and transitive), and that of finality. Summarising these conditions, Rawls writes that they amount to: “a conception of right is a set of principles, general in form and universal in application, that is to be publicly recognized as a final court of appeal for ordering the conflicting claims of moral persons”. These constraints are fairly weak, (and some inherent in a contractarian standpoint), yet they serve to exclude certain forms of egoism whilst being satisfied by traditional forms of justice and Rawls’s own conception of justice. Rawls finds this screening component of the original position helpful in assessing different conceptions of justice because it prevents discriminatory principles from being candidates. For instance, racist or sexist principles would be vetoed by those parties in the original position who hypothetically assume the role of a disadvantaged agent. Due to the fact that behind the veil of ignorance any agent could find themselves in the disadvantaged position, discriminatory principles such as these would not be deliberated upon further.
To begin investigating various criticisms of Rawls’s methodology I shall consider the accusation that the veil of ignorance is “too thick”. Under this reading, the veil of ignorance rules out information that is morally relevant to the evaluation of competing conceptions of justice. Libertarians, for instance, would reject the idea of an ahistorical hypothetical agreement on principles of justice because knowledge of one’s property rights and bargaining position is relevant (provided one is entitled to them according to libertarian principles). Many utilitarians propose the imposition of a “thin” veil of ignorance whereby parties have full historical information of the society, yet are ignorant of their own position within it. In this way the problem of partiality is addressed, yet agents retain information concerning their desires and interests. The utilitarian argument is that this information must be relevant since ultimately, the purpose of justice is to promote existing human interests. The most compelling voicing of this position has been expressed by John Harsanyi, whose exploration of a contractarian approach has led very different conclusions to those of Rawls.
Harsanyi proposes an original position in which a thin veil of ignorance allows individuals to have complete knowledge of everyone’s preferences in society, yet supposes complete uncertainty regarding their own identity within it. In assessing how such an individual would maximise their own utility without knowledge of their preferences, Harsanyi argues that an individual should imagine that they have an equal likelihood of having the preferences of any member of society. Here Harsanyi appeals to a facet of Bayesian decision theory, namely the “principle of insufficient reason”. This principle states that, in order to maximise expected utility when there is no information available about the likelihood of one outcome over another, you should assign equal probability to each outcome. Harsanyi asserts that in order to maximise their own utility, agents should systematically identify themselves with the preferences of everyone in society, considering how much utility they might have in each of these positions. Thus Harsanyi argues the utility function of an individual can be expressed:
Urational agent = (1/n)Ui1 + (1/n)Ui2 + (1/n)Ui3 + …
where n is the number of different states (hence from equiprobability assumption 1/n is the probability of any state) and i is representative of different persons in society.
Harsanyi points out that the consequence of a rational agent seeking to maximise their utility according to the above function is the same as if they were applying the principle of average utility. This principle specifies that agents choose to maximize the total average utility across all members of society.
Harsanyi states that: “Rawls makes the technical mistake of basing his analysis on a highly irrational decision rule, the maximin principle, which has absurd practical implications”. Harsanyi demonstrates his point with the following payoff matrix. The agent deciding between states A and B has no knowledge of the likelihood of the outcomes, but is aware of the value of the payoffs.
Harsanyi contends that it would be irrational to make choice 1 when unaware of the likelihood of the outcomes. Indeed, this is the course of action recommended by equiprobability reasoning. Yet this is exactly what maximin reasoning compels us to do, since by selecting choice 1 the worst outcome is avoided – the 10 pence payoff.
The imposition of a “thin” veil of ignorance does not seem wholly permissible. This is because a “thick” veil of ignorance, by ruling out all historical knowledge, ensures that the favoured conception of justice is not tailored to any particular conception of the good. Intuitively, this seems like an attractive property of the original position – we would not want the favoured conception of justice to reflect the conditions of a particular social situation. Rawls’s “thick” veil of ignorance still ‘does work’ in the sense that it renders certain conceptions impermissible (recall the racist/sexist principles that failed the screening stage of evaluation). Rather than ensure fairness to competing conceptions of the good, Rawls seeks fairness to parties. Thus the original position promotes agents’ capacities to reflect upon and pursue their ends, rather than attribute importance to existing ends. All the same, Harsanyi’s criticism of the maximin principle seems to be problematic for Rawls derivation of his conception of justice.
In response to Harsanyi’s criticism, Rawls argues that it is not necessary that the maximin strategy always provides the rational course of action, but rather it is sufficient that maximin provides the rational course of action in some circumstances of uncertainty. Consider the following payoff matrix:
According to Harsanyi’s equiprobability reasoning, an agent faced with the above decision problem should make Choice 1, since the sum of expected utility is greatest in this instance. But suppose that the above decision problem represents the one made by parties behind the veil of ignorance. Rawls argues that rational agents would take account of the fact that the vast majority of individuals might end up with outcome A, since they are unaware of the respective probabilities. If this was the case, the sum of expected utility for Choice 1 could be far less than that of Choice 2. Behind the veil, the right to basic liberties, opportunities, income, wealth and the social bases of self-respect are all at stake. Forming one’s expectations on the basis of the principle of insufficient reason without any knowledge about the likelihood of outcomes does not seem entirely rational in this instance. Indeed, it may well be the case that the vast majority of agents end up with outcome A. In this case, the sum of expected utility would be far less if Choice 1 was selected than if Choice 2 was selected. Harsanyi asserts that the principle of insufficient reason is always the rational strategy to adopt in situations of complete uncertainty. Yet in the original position, when life itself is at stake, this assertion seems irrational. Whilst the maximin principle might not perfectly capture our intuitions concerning parties behind the veil, the failings of Harsanyi’s equiprobability principle are far more significant.
This paper has critically assessed Rawls’s derivation of his two principles of justice from the original position. I have explicated the structure of the original position and the veil of ignorance. With reference to the ‘two moral powers’ of parties behind the veil, I have demonstrated how Rawls arrives at his two principles of justice. I have considered the role of the original position as a screening device. Finally I have offered an in-depth discussion of Harsanyi’s critique of Rawls. I have argued against the imposition of a “thin” veil of ignorance and stated for what reasons I believe Harsanyi’s criticism of the maximin strategy to be inadequate.
I conclude that Rawls is successful in the derivation of the two principles of justice.
References
- Freeman, S., ‘Rawls’, (Routledge, 2007)
- Freeman, S., ‘The Cambridge Companion to Rawls’, (Cambridge University Press, 2003)
- Harsanyi, J., ‘Morality and the Theory of Rational Behaviour’ in Sen and Williams ‘Utilitarianism and Beyond’, (Cambridge University Press, 1982)
- Kukathas, C., Pettit, P., ‘Rawls’, (Polity Press, 1990)
- Martin, R., ‘Rawls’ (in ‘Political Thinkers’), (Oxford University Press, 2003)
- Rawls, J., ‘A Theory of Justice’ (Revised Edition), (Oxford University Press, 1999)