Wednesday, August 24, 2022

The Nature of Moral Progress: Definitions, Types and Measures

Moral progress is something to be celebrated. But what is it, exactly? In answer to that question, many people point to paradigmatic cases of moral progress: the abolition of slavery, the extension of legal rights to women and racial minorities, the decriminalisation of homosexuality, and so on. But what is it that unites these cases? What makes them all instances of moral progress? Can we identify progress as it happens or does it only become obvious in retrospect?

These are important questions. They are important from a social perspective since past episodes of moral progress have improved the state of the world for many people. We might like to accelerate such progress in the future. They are also important from an individual perspective since we want to be on the right side of history. We don’t want to be reactionary, conservative, relics of the past. At least, most of us don’t.

But it is not always easy to say what moral progress is or to understand how it comes about. Philosophers and social scientists have been studying this topic for some time and there is considerable disagreement about what it is and whether it exists. Indeed, as some academic commentators have noted “for much of the 20th century, it was taken as a sign of moral progress that we had stopped believing in it” (Sauer et al 2021).

Still, we can say some things about the nature of moral progress. In particular, following a recent review by Hanno Sauer, Charlie Blunden, Cecilie Eriksen and Paul Rehren, we can say something about: (i) the definition of moral progress; (ii) the different forms of moral progress; and (iii) the epistemic challenge of identifying episodes of moral progress. In what follows, I will consider each of these in more detail. In doing so, I am inspired, but not constrained, by what Sauer and his colleagues have to say. Much of what I write will summarise their insights; but some of what I write will expand upon or criticise what they have to say. It should be obvious when the latter is happening.

1. The Definition of Moral Progress

Let’s start with the definition of moral progress. As Sauer et al note, there are broad and narrow conceptions of moral progress. They characterise this as a distinction between any kind of morally desirable change (broad) versus specific kinds of morally desirable change (narrow). I think the distinction is useful but prefer to distinguish between broad and narrow forms of progress on the basis that they involve different mechanisms of change. As follows

Broad Moral Progress: Any change in the world that results in a morally improved outcome/state of affairs, irrespective of how this change was brought about. For example, a decline in the murder rate over time would count as moral progress under this broad conception, even if the decline was caused by changes in economic well-being and not moral perception and reasoning.
Narrow Moral Progress: Changes to how people perceive and reason about moral situations/actions that results in a morally improved outcome/state of affairs. For example, a decline in the murder rate over time that was brought about by more people perceiving certain kinds of killing to be impermissible.


The focus on mechanisms is, in my view, important. To be sure, it can be hard to draw the line between the moral and the non-moral, but most of us tend to think that moral behaviour and thought has a particular content and purpose. We think and act morally when we attend to what is good/bad and right/wrong. We develop, morally, when we change how we think about what is good/bad and right/wrong.

It is possible to achieve morally desirable outcomes without changing people’s morally thought and action. For instance, you could reduce road deaths from speeding by either (a) encouraging people to acknowledge that speed poses risks to others (and themselves) and to drive more slowly because they wish to protect them or (b) by installing in each and every car a speed limiting device that stops people from driving too fast. Both methods could achieve the same outcome and both would, in my view, be morally desirable. But it’s hard to say that the latter would count as moral progress. It works, to the extent that it works, largely because it bypasses moral reasoning.

That said, the picture is more complicated than this example suggests. Sometimes, morally desirable changes that initially result from an amoral mechanism could, in turn, influence our moral beliefs and practices such that the desirable changes come to depend, in the long term, on a change in moral reasoning. Continuing with the speeding example, suppose we install the speed blockers in every car. Initially, this might be resisted — perhaps because it undermines freedom and autonomy — but, over time, if it does result in a significant drop in road fatalities, people might change their tune. They might come to believe that forgoing one’s freedom and autonomy, at least in this instance, is morally preferable and, indeed, that someone who drove a car without a speed blocking device was doing something immoral.

In other words, amoral progress could generate moral progress or, at least, a new understanding of what counts as moral progress. It’s possible that the rise of market thinking and the acceptance of market norms is attributable to this phenomenon. Economic growth is, to a considerable extent, dependent on people accepting transactional market norms. Some people think that these norms are amoral or, in some instances, immoral. There has been a lot of opposition to them over the years for this very reason. Market norms, we are told, promote selfishness, interpersonal coldness, distance and, ultimately, the triumph of efficiency over equality. But, at the same time, since economic growth often delivers real benefits for society (increased wealth, opportunity, innovation and so on) people have also started to moralise market norms and suggest that to follow them is to act morally.

There is a lot more to be said about this but the recursive relationship between amoral progress and moral progress is, I think, something worth exploring.

2. The Types of Moral Progress

So much for the definition of moral progress. What about its forms? In principle, any moral change can, under the right circumstances, count as moral progress. Morality consists in beliefs about what is good/bad (axiology) and right/wrong (deontology). Consequently, any change in what we think is good/bad or right/wrong could count as a form of moral progress. This means that the forms of moral progress are as diverse and bountiful as the forms of moral change.

In their paper, Sauer et al try to be a bit more precise and identify six major forms of moral progress. These six forms track with what has been argued to constitute moral progress. They are:

1. Gains in welfare: Improvements in well-being, such as gains in life expectancy, and reductions in poverty and childhood mortality, make us better off and hence, under a broad conception, count as a form of moral progress. Documenting and debating these gains in welfare has become something of a fad in recent years. Famous proponents of the view that there have been significant welfare gains over the past 250 years, or so, include Steven Pinker, Hans Rosling, Angus Deaton, among many others (see my paper on Techno-optimism for more).
2. Expansions of the moral circle: The extension of moral concern to people or things that were once excluded is often considered to be the quintessential form of moral progress. Many of the paradigmatic cases of progress, mentioned in the introduction to this article, involve expansions of the moral circle, e.g. the abolition of slavery and the vindication of the rights of women.
3. Proper demoralisation: Removing moral censure or condemnation from certain activities, e.g. gay sex and gay marriage, is often taken to be a form of moral progress. Usually, this is because doing so is taken to support or promote some important value, e.g. sexual freedom, individual autonomy, social respect/dignity.
4. Proper moralisation: Attaching moral censure or condemnation to certain activities, e.g. sexual harassment in the workplace and the casual sexual assault and exploitation of women, is often taken to be a form of moral progress. This is the inverse of the previous form of progress and, similarly, is thought to promote important values, e.g. equality, respect/dignity.
5. Improvements of moral concepts: Episodes of moral disruption or social unrest often lead to changes in our moral concepts, either by expanding the scope of existing concepts (freedom, equality) to cover new phenomena or by developing wholly new concepts to address a new or underappreciated moral issue. Sauer et al give the example of the concept of sexual harassment. This was, historically, something that was not recognised and taken seriously. By naming the problem and giving people the means to articulate their concerns about it, we facilitated a form of moral progress.
6. Improvements of moral motivation: Increasing people’s desire to follow moral norms and to act morally can count as a form of moral progress. In a sense this is necessary for all forms of progress, at least under a narrow conception of moral progress. It is only when people recognise that something is good/bad or right/wrong, and are willing to act appropriately as a result of this recognition, that they can be handmaidens to moral progress. But, as Sauer et al point out, there is another dimension to this. The more people rely on their own moral reasoning and moral perceptions, the less pressure is put on other social or institutional mechanisms for enforcing desirable change. To go back to the speeding example, it would be less costly to get everyone to recognise that speeding is a problem and to act accordingly, than it would be to install a speed limiting device in every car.


Identifying and labelling these six types of moral progress is interesting but I am not sure that is particularly informative. It’s odd, in particular, that both moralisation and demoralisation can count as forms of progress.* In essence, it boils down to what I said at the start of this section: any change in how we think about morality can, under the right circumstances, count as a form of progress. The question is ‘what are the right circumstances?’. This brings me to the last issue I want to address in this article.


3. Measuring Moral Progress

Suppose that moral change has taken place. What makes this change progressive? The answer to this leads to something we can call the measurement problem in the study of moral progress. To classify any particular instance of change as progressive you have to have a moral measuring stick against which to evaluate it. This doesn’t have to be a precise, quantitative measuring stick, but it has to allow for some ordinal rankings of different states of affairs. It needs to be able to tell you, at a minimum, that the new situation is better than the old one.

There are, of course, plenty of moral measuring sticks lying around. Most normative theories generate their own measuring sticks. Sauer et al suggest that, in the available literature, you can see people appealing to utilitarian measuring sticks — the new state of affairs makes people better off — and deontological measuring sticks — the new state of affairs respects people as morally autonomous individuals. They also suggest that there can be pluralistic measuring sticks that rely on multiple different values to assess instances of moral change.

And therein lies the rub. The measurement problem arises from the fact that there may be too many measuring sticks and they might not all reach the same verdict about a particular instance of moral change. What’s more, these measuring sticks might be contested, with some groups preferring one over another. The demoralisation of homosexuality might be progressive when measured against the values of autonomy and individual well-being but, according to conservative critics, would be regressive (or transgressive) when measured against the values of purity, naturalness, and social cohesion.

And the problem may go even deeper than this. If moral measuring sticks are themselves subject to progressive moral change, then it might be even more difficult to classify instances of change as progressive. You have to have some fixed set of values against which to measure change as progressive. If nothing is fixed, then all progress seems illusory (or at least highly contingent and relativistic).

These are not new problems. They have been part and parcel of moral philosophy for a long time, but they do affect the study of moral progress. I tend to think there is no entirely satisfactory resolution to them. The best we can do is to be clear about the measuring sticks we are using.

* I should also add that I don’t like the terms ‘moralisation’ and ‘demoralisation’. What’s really happening in cases of moralisation or demoralisation is that certain acts are undergoing a deontic reclassification: what was once permissible, becomes impermissible; what was once impermissible, becomes permissible. It’s not that the acts fall outside or inside the scope of moral reasoning.


  1. Interesting and well rounded discussion. I appreciated definitions piece and differentiations made for clarity: axiology/deontology. On another related front, a blog posted something I found puzzling asserting class is not an identity. That piece did not view class as it exists in some cultures, instead defining it in more political than cultural parameters. So, now do these sorts of distinctions square with moral progress, if indeed they square at all? I wanted to ask a question of that blog, but it rarely is accessible to questions.

  2. Agreed, this is a very interesting discussion -- and, on a self-interested note, thanks for engaging with my and my colleagues work!

    I think your point about the relationship between broad and narrow moral progress (or amoral and moral progress, as you put it) is well taken. For one thing, as you point out, when conditions of social life improve (say because there are fewer car accidents, or less physical violence, within a society) then its plausible and morally welcome that people might become more morally sensitive to car accidents or physical violence. For another thing, we also have some evidence from empirical sources like the World Values Survey that certain kinds of broad moral progress (increasing levels of wealth, increasing levels of education, etc.) correlate quite strongly with changes in people's moral values (such as them becoming more tolerant of homosexuality and divorce) that many philosophers who work on moral progress take to be narrow moral progress. On caveat is that, while the correlation is clear, it's not 100% clear what the mechanism is. Christian Welzel, one of the academics who carries out the WVS, has a theory about why certain changes in value go along with changes in background conditions like wealth and education (he calls it the "utility ladder of freedoms" theory), but this is the least evidenced part of his book on the subject, in my opinion.

    The 'measuring moral progress' problem is a biggy. One of the things that I find unsatisfying about the current debate is the way that philosophers present examples of moral progress that are straightforwardly parochial. Philosophers typically present canonical examples of moral progress and say that these examples are obvious and non-controversial, and specifically that the examples are less controversial than any of the competing ethical theories (e.g., utilitarianism, contractarianism, care ethics...) that might allow us to justify why any particular example is moral progress. I don't think that this is a bad strategy per se, but then when you look at these obvious non-controversial examples they typically include the expansion of rights for women and the expansion of rights for gay and lesbian people. I find these very plausible as examples of moral progress, but you don't have to look very hard at all to find evidence that support for these developments varies wildly across the world (for example, this survey on the topic of acceptance of gay people: So it seems that several key examples are parochial, and obviously so. More than that, they seem to be the kind of examples that appeal primarily to WEIRD people.

    Maybe the correct position is to just admit this, and say that we're using WEIRD measuring sticks. Another option is to look for less controversial instances of moral progress, to preserve a universalist flavour for these claims. I think that many broad kinds moral progress belong in this box: greater levels of health, wealth, sustenance, etc. in the past 200 or so years. One has to reckon with how unevenly these developments are spread around the world, but I think this is possible: moral progress needn't mean that things are now how we'd like them to be.

    Anyway, this comment is already way overlong! Thanks again, it was great to read your post, and I hope we can chat about this topic some more sometime.