1. Introduction

This paper evaluates Martin Luther King Jr.’s (1992) second and third criteria for assessing the justice of laws. I argue that the second criterion is useful to a limited extent, but proves indeterminate for many laws. I also argue that the third criterion deems an undesirably large set of laws unjust. I begin by examining the second criterion. I argue that it must be understood substantively if it is to be applicable in a broad range of cases. Under a substantive reading, however, it follows that laws are rarely, if ever, universally binding. In these instances, a formulation of the preferences of groups not bound by a given law is required to determine if that law is just. I demonstrate that this process of formulating group preferences is inconclusive when questions of economic equity are concerned. Therefore the second criterion is typically inconclusive. Next, I consider the third criterion and argue that we should not place excessive emphasis on the appendage of universal suffrage, but understand the criterion as a requirement that all groups have real power in affecting decisions. I argue that this interpretation leads us to conclude that almost all laws are unjust. This conclusion renders the third criterion overly broad and therefore not useful. I conclude that the second and third criterion are limited in their usefulness for assessing the justice of a law.

King’s first criterion for the justice of a law is that it complies with the law of God (1992: 3). I set King’s first criterion aside for reasons of scope, in order to deal more thoroughly with the second and third criteria. The second criterion states that a law cannot be enacted by a majority on a minority, where that law binds the minority and not the majority and where the majority would not will that law on themselves (1992: 3). The problem with this criterion stems from how we conceive of a law as binding and how we conjecture what a majority would will. The third criterion states that a law cannot be inflicted upon a minority where that minority had no part in enacting or creating that law due to being denied the ability to vote (1992: 3). Essentially, this criterion requires that all groups have a role in the decision making process through their ability to vote. The difficulty with this criterion stems from how we conceive of a group partaking in the decision making process. I assume that a useful criterion is not one that deems all laws, or an extremely large subset of laws, unjust. I acknowledge that one could equally bite the bullet, so to speak, and accept the conclusion that unjust law-making is the norm almost everywhere, when that is the conclusion that follows from applying King’s criteria. For the purposes of this paper, I simply stipulate that such a conclusion renders the criterion overly broad and not useful.

2. Discussion

Starting with King’s second criterion, one first needs to consider who is bound by a given law. Furthermore, one needs to consider, in cases where some are bound by a law and others not, what those not bound would will if they were bound by that law. I refer to this as the role-reversal requirement. That is, in cases where a law binds one group and not another, there is no injustice provided the unbound group would will themselves to be bound by that law. Hence, were roles reversed, the unbound group would consent to that law. As for the first consideration, there are two ways to formulate a law’s scope. Either we interpret the law formally or substantively. I argue that the formal approach is too weak and does not preclude the possibility of targeted injustice. Laws can be formulated to be universally applicable but, in practice, target groups. Waldron (2016) points out that “even a norm like ‘A person who is of African American descent must sit in the back of any public bus they ride on’ applies, universalizably, to everyone. A formal requirement of generality does not guarantee justice”. The issue is that laws can have a conditional form where it will just so happen that only some groups will meet the antecedent condition, but the law is still universally applicable. The law “if you are African American, you must sit at the back of bus” applies to everyone, but in reality only African Americans must endure the law. Therefore, I discard this weak, formal reading of King’s second criterion and consider the substantive reading.

Under a substantive reading, we consider who the de facto target of a law is rather than simply de jure. That is, we go beyond what is formally stated in the law and consider how that law’s implementation will affect people. With a substantive reading, we avoid obvious injustices such as Waldron’s example. However, the issue is that the de facto target of a law is rarely the whole population. Almost every law will have some implicit conditionality to it that will lead to that law affecting groups unevenly. Inheritance tax affects disproportionately the wealthy whereas VAT regressively affects poorer people more. A tax hike on gasoline and diesel will disproportionately affect consumers of these fuels, which, as the Gilet Jaunes protests in France in 2018 and 2019 have exposed, is predominantly poorer, rural populations and long-distance truck drivers. The point is that differentials across people - be that in terms of income, geography, culture or some other factor - inevitably lead to laws affecting people differently. If it is the case that laws, by and large, de facto bind some and not others, then by implication of the second criterion, almost all laws are unjust. This then means we need to defer to the role-reversal requirement in each of these cases if we wish to avoid the conclusion that practically all laws are unjust. I argue, however, that this is not a successful system for determining justice, since we are rarely guaranteed a unique answer to the role-reversal requirement. Hence, in most cases the justice of a law is not discoverable by King’s second criterion.

To meet the role-reversal requirement, we need to formulate the hypothetical preferences of groups. There are several approaches to discerning what a group or person hypothetically desires, however I will consider just one approach - the Rawlsian (1971) approach. The Rawlsian approach places us behind a veil of ignorance and asks if we would will ourselves to be bound by that law without knowing if, on entering the world, we fall into the affected or the unaffected group. This approach delivers some desirable outcomes. We would not will that African Americans must sit at the back of a bus, since we would not know if when entering into the world we would be African American and so would not want to constrain our own rights. For cases such as this, the Rawlsian approach yields desirable answers to the role-reversal requirement. But in cases with distributive consequences, reaching a unique answer from this position may not be possible. Sen (2009: 57) argues that multiple, inconsistent answers may emerge. The outcome will necessarily be contingent on “utility, economic equity and distributional fairness, and the entitlement to the fruits of one’s unaided efforts”. Continuing the example of inheritance tax, arguably the best way to treat people impartially and unbiasedly would be to protect the property they accrue via their efforts and not tax wealth on the basis of that person dying1. Alternatively, behind a veil of ignorance we might argue that treating people impartially means trying to correct for arbitrary benefits accruing to individuals for which they cannot claim credit, such as limiting arbitrary economic advantages that exist by virtue of lineage. Both answers aim towards impartiality but start from different theories of economic equity. Since we are not guaranteed a unique answer when considering the role-reversal requirement, it then follows that the justice of a law is rarely discoverable. Under King’s second criterion, we constantly need to defer to the role-reversal requirement, since almost all laws de facto bind some and not others, and this system of formulating preferences of unaffected groups contains no guarantee of a conclusive result. Therefore, King’s second criterion for justice does not return an answer for an extremely broad class of laws.

The third criterion relates to minority participation in the law-making process. There is some ambiguity in how to understand the appendage of unhindered voting rights as a necessary or sufficient condition to satisfying this criterion. I consider three interpretations: firstly, I take universal suffrage as a necessary condition. Secondly, I take it as a sufficient condition. Thirdly, I de-emphasise this appendage and consider a more general reading of the third criterion in which it is read as a requirement that all groups hold real power to affect decisions. The immediate consequence of understanding universal suffrage as a necessary condition for just law making is that any non-democracy is ruled out by default. This means that every law formulated in Singapore, for example, is unjust. Although King’s first principle of justice – that a law agree with the law of God – is not the focus of this paper, I will use it here to show how this interpretation can generate a contradiction for King. Given King’s assertion that “one has the moral responsibility to disobey unjust laws” (1992: 3), one can conclude that we are morally obliged, for example, to refuse to testify in a Singaporean court to convict a murderer who we knew to be guilty. This contradicts King’s first principle, which would almost certainly require cooperation in the case of convicting murderers. Therefore, I argue we should discard the necessary interpretation due to such problems generated with regard to non-democracies.

Understanding universal suffrage as a sufficient condition is also problematic. Universal suffrage is ultimately a formal institution. It tells us nothing of the real power held by different groups in affecting or creating laws. Indigenous populations throughout the Americas can vote yet cannot be said to hold any significant, real influence in the law-making process. There also exist hegemonic democracies, Russia for example, where all people can vote but the governing party engages in voter suppression, strategic bribes, and other tactics to ensure they retain power. Injustices are prevalent against indigenous populations in functioning democracies and regular citizens in hegemonic democracies. Yet under a sufficient reading of the third criterion we need to defer to some other criterion to argue that these laws are unjust.

One may argue that perhaps we are improperly focussing on this universal suffrage appendage. We can redeem King’s criterion by noting that its motivation is inclusion of those affected by laws in the decision-making process and in the power to enact and change laws. I therefore employ a more substantive reading. There are some benefits to this interpretation. Firstly, we escape the conclusion that all laws enacted in non-democracies are unjust, as citizens in non-democracies may have alternate ways of affecting change in their governments. A vote is not the only way to influence change. Secondly, we avoid brushing over hegemonic democracies and marginalised, but technically enfranchised groups. Despite the advantages of this reading, we run into similar issues as we did with the second criterion. Just when, exactly, do all citizens have real, significant power to create and change laws? This can only really be said to occur under very direct forms of local democracy. Indeed, this is most likely true to King’s intention. Practically however, no country operates this way, and so we end up concluding that almost every law is unjust everywhere because it is almost impossible to have all citizens holding legitimate sway in the policy arena.

3. Conclusion

I have argued that King’s second and third criteria are limited in their usefulness for assessing the justice of a law. I have shown that King’s second criterion must be understood substantively, but that a substantive reading necessitates constant deferment to the role-reversal requirement. The role-reversal requirement produces no clear answer as to what groups would or would not will when economic equity is involved. It then follows that the second criterion is inconclusive for a broad set of laws. I acknowledge, however, that this argument examined only the Rawlsian method for discerning the preferences of unaffected groups. One could also observe actual behaviour in order to forecast preferences of unaffected groups or various other empirical approaches. This is a suggested avenue for further research. I have also argued that King’s third criterion is flawed, because it is overly demanding. A requirement that all individuals have real power to affect decision-making would require highly localised direct democracy, which exists virtually nowhere in the world.


1 – In the case of the UK, where the tax is paid by the deceased.


  1. King Jr, M. L. 1992. Letter from Birmingham Jail. San Francisco: Harper San Francisco.
  2. Rawls, J. 1971. A Theory of Justice. Cambridge, MA: Belknap Press of Harvard University Press.
  3. Sen, A. K. 2009. The Idea of Justice, 57. Cambridge, MA: Belknap Press of Harvard University Press.
  4. Waldron, J. 2016. The Rule of Law. The Stanford Encyclopaedia of Philosophy (Fall 2016 Edition), E. N. Zalta (ed.).