Proportional representation and party discipline


In recent years, much of the discussion of democratic reform in Canada has focused on two themes: proportional representation and party discipline. To some observers, this pairing seems paradoxical. After all, the movement for proportional representation stresses the importance of political parties in a representative democracy, while critics of excessive party discipline stress the ways that parties can undermine the representative responsibilities of individual legislators. Some critics even go so far as to argue that parties ought to be abolished altogether.

The appearance of a paradox, however, is an illusion. Although proportional representation cannot satisfy those who call for an end to party politics, there is a strong affinity between the case for PR and the concerns of the moderate critics of party discipline. This follows from the fact that although party cohesion tends to be higher under any form of PR than it is under FPTP, PR tends to produce a larger number of competitive parties.

The main criticism of excessive party discipline, I take it, is that it results in a legislature that is less reflective of the full range of public opinion on political matters. At its most extreme, this makes legislators mere placeholders for their party leaders. And because there are usually only two or three party leaders in the legislature, the legislature is completely dominated by just two or three perspectives. Although in practice a leader must pay attention to the views of their caucus in order to remain in power, members can only express themselves freely in caucus meetings that are closed to the public. And even then, the leader — who normally enjoys a separate mandate from the party membership and exercises unrestricted control over disciplinary measures up to and including the right to expel a sitting member from the party or block their nomination — has an enormous amount of leverage.

Proportional representation presents a partial solution to this problem. This works in two ways. First, PR almost always increases the number of parties in the legislature. Even assuming that the simplified picture presented above holds true at all times, this means that the legislature will still do a better job of reflecting the full range of public opinion. Second, PR reduces the leader’s leverage by lowering barriers to entry for new parties. The threat of expulsion or a vetoed nomination has significantly less force if it is relatively easy for a group of renegades to set up a new party with a reasonable prospect of being returned to the legislature in the next election. This is likely to lead to a freer exchange of ideas within closed caucus meetings as well as in public proceedings of the legislature and its committees.

Finally, it should be noted that enacting PR is entirely consistent with other measures that have been proposed to restore balance to party discipline, such as legislating a procedure that caucus members can use to remove party leaders, transferring sole control over nominations to riding associations or regional bodies, and subjecting the most severe sanctions for members to the collective control of the party caucus rather than the leader.

These measures may sound familiar; all of them were included in the early drafts of Michael Chong’s Reform Act. By the time the Reform Act passed, unfortunately, every one of its provisions had been watered down to the point of meaninglessness. While the movement for proportional representation in Canada remains strong and continues to bear fruit, the other major theme in the conversation about democratic reform seems to have petered out for now. Given that these themes are in fact closely related, however, citizens who are still concerned about excessive party discipline in BC and the country at large have good reason to get behind proportional representation.


Meta-ethics and progressive politics, part 6: Equality, contractualism and capabilities

1296-1Reasonable agreement between equals with distinct properly valued ends

In my previous post in this series
, I argued that all persons possess an equal intrinsic non-monetary value. While this conclusion rules out the kind of argument against extreme economic inequality propounded by Danny Dorling, it provides the basis for a distinct, fundamentally egalitarian critique of particular economic, social and political inequalities.

Before proceeding, here’s a recap of the story so far. Reasons are real; there are considerations that count in favour of being, doing or believing some things rather than other things. Our beliefs about what reasons we have can be justified in the pursuit of reflective equilibrium between general principles and our considered judgments about particular cases. In this pursuit we frequently rely on social processes — including political struggle and society-wide “experiments in living” — to counteract biases that distort our thinking, to gather new information, and to test the reasonableness of our judgments.

Last time, I left open the precise nature of our equal intrinsic value. However, this summary hints at an historically popular answer: the intrinsic value of a person is closely related to the capacity for autonomy, or self-governance on the basis of reasons. For some philosophers, such as Kant, this is an identity relation — the value of a person is their capacity for autonomy. But many people, especially today, find this hard to swallow because it implies that young children and persons with severe cognitive disabilities have no value and are owed nothing as a matter of justice. The intuitive connection between our ability to recognize and act on the basis of reasons and our distinctive value can be vindicated by loosening the relation so that the equal intrinsic value of a person consists in their being the kind of creature that has the capacity for autonomy. Very young children and persons with severe cognitive disabilities meet this criterion. Their lack of autonomy is merely a contingent fact about what they happen to be like; it does not change what kind of being they are.

If this is the basis for the intrinsic value of persons, what is the proper response to that value? What does it mean to treat a person according to their worth? Monetary values are easy to deal with — if something is valued at $4.99, the proper response to that value (if you’re going to respond to that value at all) is just to throw down $4.99 (plus applicable taxes). But in a way, the principle is the same. Just as you honour monetary values with money, you honour the capacity to recognize and act on the basis of reasons with reasons — by treating others in ways that can be justified to them, insofar as they are reasonable. This basic idea can be fleshed out in a variety of ways and has taken a number of well known forms including the Golden Rule, the Confucian concept of shu, Kant’s categorical imperative, and — particularly relevant to the present discussion — the contractualist ideal of reasonable agreement on principles of conduct or institutional design.

This clarifies why, if the equal value of persons has nothing to do with money, vast economic inequalities would still be objectionable. The distribution of income and wealth in society is not a purely natural artifact; it is the outcome of a massive system of social cooperation that includes a wide range of institutions across both the public and private sectors. Property law, tax policy, industrial relations frameworks, public services, the legally approved governance structures of private firms, and the essential features of national constitutions — all of these interrelated elements collectively comprise what Rawls calls the basic structure. The basic structure has a pervasive effect on our life prospects; this gives each of us a compelling interest in the design of that structure, and generates the requirement that the principles underlying the structure’s design be justifiable to each person. The degree of economic inequality Dorling observes in his society fails to meet this standard; such vast inequalities would be permitted by principles that could reasonably be rejected or would be forbidden by principles that would be accepted by all. This explains at least one reason why egalitarians should object to extreme economic inequality.

But let’s not stop there. Although it seems plausible that principles permitting extreme inequality could be rejected and principles forbidding extreme inequality would be accepted, a good theory of justice should spell out the bases of rejection and acceptance. In doing so, we gain a better idea of what the guiding principles of a just society’s basic structure would actually require, a more vivid picture of what that society would look like, and more practical guidance concerning the kinds of policies that might get us there.

Given the realist stance towards normative reasons that I have adopted in this series so far, a natural starting point is to ask what we have reason to want out of life. One influential approach to this subject follows the work of Martha Nussbaum, who has proposed the following list of capabilities that are central to social justice, where capabilities are defined as effective freedoms or real opportunities to achieve functionings (“beings and doings”) that one has reason to value:

From “Capabilities as Fundamental Entitlements: Sen and Social Justice” by Martha Nussbaum

Unlike the other theorists discussed in this series so far, Nussbaum is not a contractualist. But she has acknowledged that there are strong affinities between contractualism and her own approach; where the two do not actually converge with each other, they are at least compatible with each other.

Nussbaum’s capabilities approach is not a full theory of justice. One reason for this is that it does not specify rules to govern the distribution of capabilities (or the goods that are necessary to secure the central capabilities). But it does impose a constraint that any distributive rule must be able to meet. The capabilities spelled out in Nussbaum’s list are of fundamental importance for all of us. In contractualist terms, this means each of us can reasonably reject principles for the design of the basic structure that would deprive us of any of these capabilities. It also gives each of us reason to accept principles that would secure the central capabilities, even if this means giving up chances for the unlimited accumulation of riches. In order for a society to be considered minimally just, then, the central capabilities must be guaranteed to all, at least up to some threshold of sufficiency. It is impossible to determine a priori exactly where this threshold is located. We can only grope towards it through a combination of social experimentation and the ongoing quest for reflective equilibrium. In practice, this kind of uncertainty does not pose much of a practical problem, because it is safe to say that we are far short of reaching this threshold wherever it is, and there is no shortage of excellent ideas for how to get there.

The egalitarian approach to theorizing about justice that I’ve been discussing here flows smoothly from the realist view of normative reasons with which this series began. Our capacity to act on the basis of reasons generates the requirement that the principles underlying major social institutions must be justifiable to all. It also helps establish a universal, objective standard by which the justifiability of these principles can be assessed: the ability of a basic structure designed in conformity with those principles to secure the central capabilities. And unlike Dorling’s meritocratic principle, this standard explains why both monetary and non-monetary inequalities — in access to health care, personal safety, recreational opportunities, and so forth — are objectionable. Given its explanatory success and its strong meta-ethical foundation, the egalitarian approach is clearly worth pursuing further.

All posts in this series

1. Meta-ethics and progressive politics
2. Meta-ethics and progressive politics, part 2: Partners in crime
3. Meta-ethics and progressive politics, part 3: Moral knowledge and reflective equilibrium
4. Meta-ethics and progressive politics, part 4: Social struggle and moral knowledge
5. Meta-ethics and progressive politics, part 5: Meritocracy versus equality

Why the minimum wage is like free trade

Much like the debate on the minimum wage, the Darvaza gas crater fire has been raging for decades and shows no sign of abatement

Over the last week or so, Ontario’s scheduled minimum wage hike has sparked a new round of debate on the wisdom of this policy, and in a column at the National Post, Andrew Coyne has reiterated his preference for a guaranteed minimum income rather than a minimum wage. The column is worth reading, but his conclusion should be rejected.

Coyne’s reasoning is that raising the minimum wage can end up hurting some of the people it is meant to help. If it becomes more expensive to employ people, then some people who would otherwise have low-paying employment will end up having no employment at all. Raising the minimum wage may well increase overall labour income, he admits, but gains for those who are lucky enough to keep their jobs do not justify the burdens imposed on those who lose out. If minimum wage hikes were the only means available to increase the incomes of poorly paid workers, then we might face a real dilemma. Fortunately, he points out, this is not the only means available. Supplementing workers’ low pay with a basic income scheme would achieve the same goal without unfairly burdening the worst off. Given our collective refusal to pursue this alternative, minimum wage supporters’ indignant rhetoric about greedy employers seems hypocritical.

I agree with some of the points Coyne raises, especially where he invokes sound Rawlsian principles holding that the well-being of the worst off should be our paramount concern, the aggregate benefits of a policy do not cancel out the burdens imposed on actual individuals, and the task of ensuring a decent standard of living for all is a job for the basic structure of society as a whole rather than employers in particular. However, Coyne’s conclusion should be resisted, because these principles are not incompatible with the minimum wage as such; they can easily be reconciled by implementing a minimum income scheme alongside a minimum wage.

Coyne’s reluctance to address this third option is puzzling, especially because the kind of dilemma he points out is extremely common. He even points out a comparable situation himself: free trade. Increased competition in open markets promotes innovation and leads to lower prices for consumers; while the overall gains can be substantial, however, so too may be the costs to the worst off. Even if we assign absolute priority to the interests of the worst off, however, we are not obligated to erect trade barriers to shield them from these costs. Instead, we can redistribute some of the gains from free trade in the form of cash transfers and active labour market policies, thereby maintaining or improving the situation of the worst off without having to forgo the benefits of a more open domestic market.

The analogy between free trade and the minimum wage is appropriate because Coyne grants the validity of a Bank of Canada finding that minimum wage hikes will increase labour income overall. Coyne does not contest that this is a legitimate policy goal; his objection is just that this goal must not be pursued at the expense of the worst off. But he does not consistently apply the principles his argument invokes.

A Rawlsian focus on the basic structure works both ways; the obligation to ensure a decent standard of living for all doesn’t fall on employers or the government in particular, and specific responsibilities arising from this more general obligation may be assigned to either of them depending on what division of responsibilities would result in the best overall basic structure consistent with the requirements of justice. A basic structure that secures a decent standard of living for all and increases overall labour income by assigning employers the responsibility to pay a minimum wage and assigning government the responsibility to provide a basic income is preferable to one that achieves only one of these objectives. Coyne — and the rest of us — ought to support both of these policies, not try to pick one or the other.

Meta-ethics and progressive politics, part 5: Meritocracy versus equality

primeThe ultimate basis of Optimus Prime’s authority is egalitarian rather than meritocratic, and if it’s good enough for giant alien robots it’s good enough for me

In a post at the LSE blog (adapted from one of his books), Danny Dorling criticizes gaping inequality as a threat to the moral sentiments (feelings of sympathy and kindness towards one’s fellows, for example, which reflect the obligations we owe to them) and the welfare state. Although Dorling makes several sharp observations about some of the negative consequences of extreme inequality, his complaint that extreme inequality fails to respect the value of persons misses the mark:

The 1 per cent, by definition, will always be those taking the largest slice, but not always such a great fat slice, leaving slithers [sic] for the rest. Question those who say that it can only be this way. Try to question them kindly rather than with incredulity. A society based on merit would be remarkably equitable compared with what we face today. No one is worth 3,000 times another person. The three-thousand fold inequalities within the 1 per cent are just as indefensible as those between them and the other 99 per cent.

As this passage makes clear, the basis of Dorling’s complaint is meritocratic, not egalitarian. On this view, it seems, each person has a cash value that varies according to individual merit, and people ought to be paid according to their cash value. Some people are more meritorious than others, and so should receive higher pay; a person who is twice as valuable as another should have their salary doubled, for example. But in our society, the richest people can earn thousands of times more money than the poorest. This cannot be justified by Dorling’s meritocratic criterion, because (plausibly) no one is worth thousands of times more than anyone else.

This is not a very satisfying line of reasoning, for three reasons. First, the inequalities that Dorling is prepared to accept seem likely to have the same deleterious effects on the moral sentiments and the welfare state as the inequalities he objects to. If I believe I am, at the fundamental level, worth twice as much as another individual, I am unlikely to regard them as sympathetically as someone just as valuable as or more valuable than me. The fact that they are fundamentally worth less than others means that they are less deserving of my sympathy than others. Likewise, if I believe that the economic system has been arranged on a meritocratic basis so that each person gets at least roughly what they are worth (again, in a fundamental sense), it is difficult to see why I should support the welfare state; it would seem as though each person already receives what they deserve, so social spending would involve taking from the deserving and giving to the undeserving. If concerns about the moral sentiments and the welfare state are part of the reason why Dorling objects to extreme inequality, then, they are also reasons to object to the milder meritocratic inequalities Dorling’s view supports.

The second problem with the meritocratic view is that it requires the state to make intrusive, degrading and presumptuous assessments of each individual’s worth. There is even more at stake here than the ability to sympathize with one’s fellows. Democratic, liberal states are defined by a commitment to essentially egalitarian norms within the political sphere, including the equal weighting of votes and a system of equal basic rights and liberties. But to maintain meritocratic inequalities in the economic sphere, the state would also have to commit to a set of essentially inegalitarian norms operating alongside the egalitarian norms of liberal democracy. This arrangement seems unstable to me; I think it is extremely unlikely that the state can sustain a full commitment to both of these fundamentally incompatible sets of basic norms at the same time. After all, if some people are viewed as being worth less than others in a fundamental sense, it is not clear why there should be a commitment to egalitarian norms in any sphere. I also find it hard to believe that state’s determination of each person’s fundamental value could be effectively insulated from political influence. Even if it were possible for the state to accurately determine each person’s merit, there would be a powerful incentive for political factions to skew these determinations in favour of their own interests.

Finally, it is implausible that the fundamental value of a person is a cash value. Money itself is only instrumentally valuable. For something to be instrumentally valuable, it must be useful for some intrinsically or non-instrumentally valuable purpose. So what is the intrinsically valuable purpose for which money is useful? It is natural to think the answer is that having money is necessary for us to thrive. But if the fundamental value of ourselves is itself a monetary value, then the value of persons cannot provide the necessary grounding for the instrumental value of money. In the absence of any other plausible candidate for the source of money’s instrumental value, we must conclude that persons possess intrinsic value from which the instrumental value of money can be derived.

If the value of persons is intrinsic, does it follow that it must also be equal? Perhaps not. But what could account for inequality in the intrinsic value of persons? Instrumental considerations (such as a person’s usefulness for some purpose) and judgments of character are extrinsic factors — by definition, they cannot enhance or diminish one’s intrinsic value. Because we have already ruled out the dependence of morality on God’s commands, divine decrees cannot explain inequality in intrinsic value either. This leaves the possibility that variation in the value of persons could be a brute fact — a fact with no further basis, reason or explanation.

Allowing that there could be brute moral facts, however, would be inconsistent with much of the account of morality and moral knowledge provided in this series. Moral facts cannot be directly observed, and brute facts cannot be directly inferred. Brute moral facts, then, would be inaccessible to both reason and observation. Without confidence in any of our moral beliefs, we would have no grounds to believe in morality at all, and if there are brute moral facts, we could not have confidence in any of our moral beliefs. And if we have no grounds to believe in morality at all, then we certainly have no reason to believe that there are brute moral facts. It follows that belief in brute moral facts cannot be justified, and therefore inequality in the intrinsic value of persons cannot be a brute fact.

Having established the intrinsic value of persons and ruled out all possible bases of inequality in this value, we must conclude that all persons possess equal intrinsic value. So what does this imply about economic inequality, the issue that kicked off this discussion? If income and wealth have no bearing on the value of a person, does that mean extreme inequality of income and wealth is consistent with each person possessing equal value? Not at all. In fact, compared to meritocracy, this egalitarian perspective provides a much stronger, more rationally coherent foundation for generalized sentiments of sympathy and kindness and their institutionalization in the welfare state. It also gives us reason to close the social distance created by material inequalities, and to ensure that those inequalities are restricted to levels that benefit everyone. Overall, egalitarianism gives a much more satisfying explanation of why extreme economic inequality is objectionable — it brings us closer to reflective equilibrium than the meritocratic view. The next part of this series will explore the egalitarian perspective in more detail, clarifying its connections with both the meta-ethical foundations established so far and the realm of progressive social and economic policy.

All posts in this series

1. Meta-ethics and progressive politics
2. Meta-ethics and progressive politics, part 2: Partners in crime
3. Meta-ethics and progressive politics, part 3: Moral knowledge and reflective equilibrium
4. Meta-ethics and progressive politics, part 4: Social struggle and moral knowledge
5. Meta-ethics and progressive politics, part 5: Meritocracy versus equality

A brief response to Daniel Zamora on basic income

In Jacobin, Daniel Zamora has an essay criticizing basic income, a topic I’ve written about on this blog a number of times. I found Zamora’s argument unsatisfying, for reasons I outline briefly here:

1. Zamora points out that Guy Standing’s small UBI costs 6.5% of GDP without achieving a large reduction in the poverty rate, whereas the cost of eradicating poverty only amounts to around 1% of GDP. This is very odd, because Zamora is using the same trick as the UBI advocates who like to pretend the program would be cheap. A program that only brings people up to the poverty line is functionally equivalent to a UBI with 100% clawback at the poverty line. It’s not a realistic option, so it’s not a reasonable comparison.

2. Zamora complains that a UBI sufficient for a decent standard of living — the only kind worth considering, in his view — would be too expensive; in France, he figures this would cost about 35% of GDP. But this assumes that the full amount would be paid out to everyone (including children) and none of it would be recovered through a clawback rate or taxes. In other words, he only considers the least efficient implementation you could possibly come up with. This weakens the cost-based argument tremendously. A UBI with more realistic parameters would still be very expensive, but not that expensive; as I’ve discussed here before, it could be fully funded in Canada by raising the GST to roughly the same level as Denmark, Finland, Sweden and Norway.

3. Zamora complains that proposals for an UBI assume that the unemployed don’t want to work or that cash can compensate for unemployment. This is false. While some advocates of a UBI surely believe one or both of these claims, there is no necessary connection between them. A UBI solves the problem of being cash-poor (not having enough money) without solving the problem of being work-poor (not having valued employment), but so what? There’s no a priori reason why the same program has to address every dimension of poverty. Perhaps it’ll turn out that the best solution to being cash-poor will also solve the problem of being work-poor, but we can’t just assume that this is the case.

4. Early on, Zamora complains that a small UBI will not be sufficient to increase workers’ bargaining power, so the promise of transforming the labour market will remain unfulfilled. But then later on, he seems to forget that he thought a large UBI could increase workers’ bargaining power. If the division of labour in our society sucks and people don’t like it, a large UBI gives workers real leverage to change it. So the idea that a UBI necessarily props up an unacceptable division of labour doesn’t hold water.

5. Finally, maybe I’m just not in the target audience. Close to the very end, Zamora asks:

Isn’t the best way to fight against capitalism to limit the sphere in which it operates? Establishing a base income, by contrast, merely allows everyone to participate in the market.

As a social democrat, I don’t have a problem with capitalism in the minimal sense, i.e. an economy where there is some role for private ownership and market exchange. I just want it to be restricted to certain spheres (e.g. it’s good for candy bars and clothing, but not so good for health care and a just distribution of income and wealth) and everyone has opportunities for participation on fair terms. In other words, I don’t care if the best way to fight against capitalism as such is to indiscriminately limit the sphere in which it operates, because I don’t think capitalism as such is bad. If I did, I might be more sympathetic to Zamora’s argument.

Site C and sunk costs

In recent days, a number of commentators and news outlets have discussed the sunk cost fallacy in connection with the government’s decision to proceed with construction of the Site C hydroelectric project. Without exception, I’ve found the handling of this issue completely unsatisfactory and highly misleading with respect to the meaning of the sunk cost fallacy and/or the reasoning cited in support of the government’s decision. Before getting to the problem, let’s get clear about what this fallacy involves.

Suppose I buy a movie ticket, but I find the film boring, amateurish, aesthetically repellent and morally repugnant. I could walk out and find something more worthwhile to do with my evening, but I stay put. After all, I’ve already paid for my ticket. I can’t bring myself to waste it, so I suffer through the rest of the film. I have just committed the sunk cost fallacy, allowing the unrecoverable costs I incurred in the past to take precedence over expected returns in the future. Assuming my goal is to have a fun evening, those costs are irrelevant; the only thing that’s relevant is whether I will have a fun evening if I stay in my seat.

Some critics of the decision to proceed with Site C claim that this is exactly what’s wrong with the government’s decision. Ministers have pointed out that if the project is abandoned, the government will have spent $4 billion with nothing to show for it; superficially, that does look like the fallacious reasoning in the movie theatre example. But it is not fallacious to consider sunk costs to the extent that these costs actually have some bearing on the future.

The total bill for Site C is expected to be around $11 billion. About $2 billion has been spent. Completing the project should cost about $9 billion. At this point, cancelling the project would cost an additional $2 billion. Therefore, cancelling Site C can only be justified on cost-benefit grounds if there is an alternative power project with equal value to the province but a total bill of only $7 billion or less. In the government’s assessment, there is no such alternative; thus, cost-benefit calculation favours completing the project. So, costs that have already been incurred in the course of the Site C project are cited in support of the decision only because they relate to the amount of money the province will have to spend in the future. That is why the government has not committed the sunk cost fallacy.

This is not to say that the government is not open to criticism on other grounds. Not committing basic errors of reasoning is a low bar to clear, after all. The government’s critics may wish to argue that the cost of completing the project has been underestimated, or that the costs of alternatives have been overestimated. Or they may argue that cost-benefit considerations should not have been allowed to override property rights, treaty rights, or Indigenous rights. For those who wish to make these arguments, credibility is a precious resource. It would be a mistake to squander this resource on misrepresentations.

Full disclosure: I’m a member of the NDP and I opposed continuing the Site C project.

Meta-ethics and progressive politics, part 4: Social struggle and moral knowledge

Screen Shot 2017-12-04 at 8.17.40 PM
This is what moral philosophy looks like

The moral philosophers who have put in an appearance in this series so far — Elizabeth Anderson, Stephen Darwall, John Rawls, and T. M. Scanlon — are all contractualists. According to contractualism, morality — or at least some particular moral domain (e.g. social justice in Rawls’s case, or interpersonal morality in Scanlon’s) — is based on the idea of an agreement between equals; our obligations derive from principles that no one could reasonably reject, or which all would agree to. Note that this is only an idea of an agreement; contractualists do not suppose that there has ever been (or could be) a real contract specifying the basic principles of morality. Nevertheless, the fact that we would agree to or could not reject these principles should be sufficient for us to recognize them as binding.

Even though the agreement envisioned by contractualists is strictly hypothetical, contractualists tend to agree that we have good reason to pay attention to real agreements and real disagreements in the social world. Neither moral philosophy nor everyday moral reasoning can be a strictly a priori enterprise. This distinguishes morality from other normative domains like logic and mathematics. As Elizabeth Anderson argues:

With respect to mathematics, it is plausible to suppose that the social identities of inquirers are irrelevant to how we think about the subject matter. This idea is harder to credit with respect to moral inquiry. Moral reasoning is supposed to help diverse people live together, come to terms with their differences, and promote peaceful cooperation on fair terms by supplying mutually acceptable principles for adjudicating the conflicting claims they make on each other, and for coordinating our moral sentiments to fit the demands of living together. We should expect that people’s social positions affect the claims they regard as intuitively legitimate, as well as their moral sentiments.

Because of this, Anderson advocates for a more empirical approach to moral philosophizing that goes beyond the armchair, the seminar room, and even the laboratory settings of “experimental philosophy”; moral philosophy must take seriously the real claims that people make on each other in the course of social life. This is especially important in the case of political struggles, where sharply conflicting and often mutually inaccessible moral intuitions are brought to light.

To this end, Anderson has devoted many years to the study of historical egalitarian social movements, especially the abolitionist movement in the United States and Britain. In practice, Anderson finds, strictly a priori methods of moral theorizing were not sufficient to defeat pro-slavery arguments; for every bedrock moral principle the abolitionists invoked to criticize slavery, supporters of slavery seemed to have one of their own at the ready to defend it. Even on the abolitionist side, whites tended to balk at the implications of anti-slavery arguments that presupposed the equality of persons, as it was apparent that a robust principle of equality would call into question hierarchies of class and sex that were widely approved. Today, most people would be inclined to endorse such a principle of equality precisely because it rules out domination based on sex and class as well as race. But at the time, whites were conditioned to regard this as a highly counterintuitive implication of a general egalitarian principle. The reasons that would eventually come to be recognized as the proper grounds for opposing slavery were not articulated until black abolitionists arrived on the scene and produced their own critiques of slavery, rooted in their own experiences.

As Anderson’s study of the abolitionist movement makes clear, moral bias and the limits of our individual experience can lead us into moral error. While cultivating a “philosophical” detachment from moral issues can help us avoid rash judgments, it can also deprive us of indispensable sources of insight. Political struggles in which victimized individuals raise their own voices against oppression can break down the walls that deprive us of the insights necessary for moral progress. If moral philosophy is, in part, concerned with moral truth, and moral progress involves advancing moral knowledge, then the advancement of moral philosophy must depend in part on political struggle.

Note: Anderson presents her view in part as a critique of the method of reflective equilibrium, the subject of the previous post in this series. I do not think this critique is successful (one reason being that it seems to be based on a caricature of the method), but nor does her account of moral progress seem to depend on it. In my view, the two approaches are complementary rather than opposed; I hope to return to this subject in the future.

All posts in this series

1. Meta-ethics and progressive politics
2. Meta-ethics and progressive politics, part 2: Partners in crime
3. Meta-ethics and progressive politics, part 3: Moral knowledge and reflective equilibrium
4. Meta-ethics and progressive politics, part 4: Social struggle and moral knowledge
5. Meta-ethics and progressive politics, part 5: Meritocracy versus equality

Anderson and Rawls on the limits of reflective equilibrium

At least there is likely to be a reflective equilibrium for these possible worlds

In her 2015 Presidential Address, “Moral Bias and Corrective Practices: A Pragmatist Perspective”, Elizabeth Anderson criticizes two dominant approaches to moral and political philosophizing: “the ascent to the a priori” (i.e. searching for fundamental, self-evident truths from which the rest of morality can be derived) and the method of reflective equilibrium. In the latter method, she says,

we move between intuitively appealing general moral principles and intuitions about particular cases. We use each to modify the others until we arrive at a set of principles that accounts for our moral judgments of all particular cases. Carried to its logical conclusion, this method can also lead to moral principles for all possible worlds, as long as we entertain thought experiments about sufficiently bizarre cases to elicit intuitions against which to modify our general principles. (p. 22)

Part of the problem, she argues, is that we have no grounds for confidence in our intuitions about truly bizarre cases. Our moral intuitions about this world are unreliable enough, but at least they can be tested and corrected by experience, whereas intuitions about people seeds, teletransporters, and runaway trolleys don’t even have that going for them.

But here’s Rawls in A Theory of Justice:

Some philosophers have thought that ethical first principles should be independent of all contingent assumptions, that they should take for granted no truths except those of logic and others that follow from these by an analysis of concepts. Moral conceptions should hold for all possible worlds. Now this view makes moral philosophy the study of the ethics of creation: an examination of the reflections an omnipotent deity might entertain in determining which is the best of all possible worlds. Even the general facts of nature are to be chosen. Certainly we have a natural religious interest in the ethics of creation. But it would appear to outrun human comprehension. From the point of view of contract theory it amounts to supposing that the persons in the original position know nothing at all about themselves or their world. How, then, can they possibly make a decision? A problem of choice is well defined only if the alternatives are suitably restricted by natural laws and other constraints, and those deciding already have certain inclinations to choose among them. (p. 159 in the original edition)

This seems like an explicit rejection of the excessive ambition and insensitivity to empirical considerations that Anderson objects to. Those vices are not essential to the method of reflective equilibrium after all; from very beginning, Rawls denied that this method — or indeed any method — could be used to find principles for all possible worlds.

Why electoral reform does not require a supermajority

BC Parliament Buildings Victoria BC

At the Province, Mike Smyth complains that it is unfair for the Greens and NDP to set the threshold for victory in the upcoming electoral reform referendum at 50 percent plus one. The problem, he argues, is that this makes it far too easy to bring about a fundamental change in a democratic mechanism that needs broad support in order to function properly. Supposing that half of all eligible voters participate in the referendum, and the reform side barely reaches the threshold, a new system would be adopted with the support of just a quarter of all voters. In Smyth’s view, this would clearly not meet the standard of “broad support”.

I think this argument is misguided. The expectation of low turnout actually works against Smyth; supporters of the status quo can effectively raise the threshold for victory simply by mobilizing more non-voters to vote No. Requiring supermajorities makes much more sense when turnout is expected to be very high — in a legislature, for example, or a referendum on independence.

Furthermore, the argument that a supermajority ought to be required for electoral reform because it affects such a fundamental part of the democratic system seems to beg the question in favour of the status quo. A 60/40 supermajority requirement allows the status quo to win with the support of only 40 percent plus one. This makes sense only if there is a shared presumption in favour of the status quo. According to many supporters of electoral reform, the current system is democratically defective. Thus, there is no shared presumption in favour of the status quo, and no justification for setting the threshold for approval of the status quo as low as 40 percent plus one.

In fact, one of Smyth’s own arguments seems to imply that there ought not to be any presumption in favour of first past the post. One of his concerns is that the Greens and NDP stand to benefit if proportional representation is adopted, which gives them an interest in slanting the process towards the Yes side. But a corollary of this — which he freely admits — is that the Liberals stand to benefit from retaining the current electoral system. If we should avoid adopting thresholds that increase the likelihood of victory for one side or another, and we should not adopt a 50 percent plus one threshold because it increases the likelihood of a favourable outcome for the Greens and NDP, we cannot then adopt a 60 percent plus one threshold that increases the likelihood of a favourable outcome for the Liberals.

As you may have noticed, we’ve come to a dead end. Every possible threshold favours one side or another compared to some other threshold. Every electoral system favours one party over another compared to some other electoral system. A principle that forbids favouring this or that faction (in the relevant sense) provides no practical guidance for the design of referendums or electoral systems. Therefore a principle for ensuring the legitimacy of the referendum result will have to be found elsewhere.

Meta-ethics and progressive politics, part 3: Moral knowledge and reflective equilibrium

Try to achieve reflective equilibrium in advance next time, Doctor

Last time, we saw that moral facts are normative — they tell us what to do — and primitive — they are not reducible to some further, non-moral facts. Moral facts share these characteristics with facts about logic and mathematics. It is commonly held to be unproblematic that we know certain facts about logic and mathematics; thus, it is not clear why facts about morality should raise any special problems. However, while it may be unproblematic to say that we know normative facts, it is more difficult to say how we know normative facts, given that normative facts cannot be inferred from non-normative facts alone.

Knowledge of normative facts depends on knowledge of other normative facts. According to a standard definition, knowledge is justified, true belief. Normative facts are true in virtue of other normative facts. But how can we get from truth to knowledge? How, in other words, do we justify normative beliefs?

Nelson Goodman proposed that we can (and routinely do) justify certain normative beliefs — those concerned with principles of logical inference — by attempting to derive general principles from our judgments about the strength or validity of particular inferences. We then apply these principles to new cases to see if they generate conclusions that accord with our judgments about those cases. When a conflict arises, it may seem best to revise our judgments in light of these general principles, or it may seem best to revise the general principles and test them against another set of new cases. This process continues until our judgments about particular inferences and our general principles are fully aligned and mutually supporting. This endpoint is only hypothetical — there are always new cases to consider. But we can make progress towards it, and our beliefs about what counts as a good inference and why are justified in proportion to the progress we have achieved.

Goodman’s method for justifying beliefs about the principles of logic was taken up by John Rawls, who thought that it could also be applied to beliefs about morality, or at least a particular domain of morality — the domain of social justice. He called this the method of reflective equilibrium. We start with moral judgments in which we have a high degree of confidence under good conditions for making judgments of that kind (e.g. when we are refreshed and cool-headed rather than tired and upset), and a set of supporting principles that explain those judgments. We then test those principles against different cases, checking whether they yield judgments that we are, on reflection, prepared to endorse, and then revising principles or judgments as appropriate. Wash, rinse, repeat. Rawls says that when we apply this method, we are describing our sense of justice. But this does not mean that our conception of what justice requires remains fixed. Describing our sense of justice is an exploratory and deliberative process through which we discover principles that we can justify as constraints on our behaviour. At any point in this process, any judgment or principle is open to revision; we may end up with a fully developed conception of justice that would seem very surprising from where we began.

Recall that we are using the definition of knowledge as justified true belief. If we are justified in believing that the principles and judgments we hold in reflective equilibrium are true, and those principles are true, then the method of reflective equilibrium is a method for acquiring moral knowledge.

Religious belief plays no role in this method, which makes it particularly appealing to atheists. It should also be appealing to many progressives who do hold religious beliefs. After all, many religious ethicists would agree with St. Paul that morality was “written in our hearts” by a supreme being, and with Aquinas that it can be apprehended by reason without the aid of holy books. Reflective equilibrium can support these views by providing a fuller account of how morality was written in our hearts (i.e. by creating a universe from which rational creatures can emerge and flourish), and how reason can be used to read that which was written.

Others, however, may find this account troubling because it threatens to make the moral content of sacred texts superfluous. But I think this threat is illusory. A religious text may still provide moral guidance by suggesting new judgments about particular cases (e.g. “God was right to kill every firstborn child in Egypt to secure the Hebrews’ release from slavery”) and candidate principles (“Love your neighbour as yourself”). Some of these judgments and principles should be accepted, and others should be rejected. The method of reflective equilibrium provides a non-arbitrary standard for deciding which parts of a text to reject or reinterpret, while preserving the overall text as a potential source of moral insight. Far from making religious texts superfluous, the method actually gives even non-believers good reason to pay attention to the conclusions of religious ethicists. Elizabeth Anderson puts the point nicely:

That God as people imagine it does not literally exist no more undercuts the value of their moral reasoning than Mackie’s critique of metaphysical moral realism undercuts critical moral reasoning today. (And even Mackie, after trashing moral realism, finds himself unable to avoid moralizing himself.) The critical moral point of view is essential and productive even if it is projected onto a fictional entity. What matters is the alternative perspective it affords, not its ontological status.

Perhaps this all makes moral knowledge sound too easy to come by. If the method of reflective equilibrium does imply that moral knowledge is very easy to come by, that counts against the method. To vindicate the method, we need to show that it is consistent with the slow rate and inconsistent direction of moral progress throughout history, and the phenomenon of apparently intractable moral disagreement both within and between cultures today. The answers to these puzzles will take us from the very abstract level at which reflective equilibrium has been described towards the concrete circumstances in which we pursue it. This will be the subject of the next post in this series.

Note: In addition to the blog by Elizabeth Anderson cited above, this post draws on discussion of reflective equilibrium by John Rawls in A Theory of Justice, T. M. Scanlon in Being Realistic About Reasons, and Norman Daniels in the Stanford Encyclopedia of Philosophy article “Reflective Equilibrium”.

All posts in this series

1. Meta-ethics and progressive politics
2. Meta-ethics and progressive politics, part 2: Partners in crime
3. Meta-ethics and progressive politics, part 3: Moral knowledge and reflective equilibrium
4. Meta-ethics and progressive politics, part 4: Social struggle and moral knowledge
5. Meta-ethics and progressive politics, part 5: Meritocracy versus equality