Can you measure the good you’ll do

The idea that it’s impossible to measure which career lets you make the most difference is silly.

If it were true, then packing meat for a living would be, for all we know, as good for the world as running Oxfam or being a great President.

Why, then, do we so often meet the idea that ‘you can’t measure the good done by a career’? – an idea that quashes debate about what’s best to do, and thus leads millions of ambitious young people to do less for the world than they could.

Here’s the mistakes I think are being made.

1. Thinking measurement needs to be perfectly precise

Thinking back to school science classes, it’s easy to think ‘measuring’ means ‘precisely quantifying.’ A classic example is using a pair of scales to weigh something, and having a ‘precise’ number as the result. In choosing between careers, we don’t get precise numbers, so the good you do is not measurable.

But this isn’t a good way to think about measurement. No measurement is perfectly precise. Weighing with scales actually only gives you a range of probable values for the true weight, because the scales are not 100% accurate. In reality, there’s always some uncertainty. Our choice is only ever how much uncertainty is OK given our purposes.

If there’s always some uncertainty left, then we can see that in practice what it is to measure some quantity is to reduce uncertainty about its magnitude. That’s all that’s actually possible. But now that we’ve given up the unrealistic definition of measurement, we can see that lots of things can be measured – all we’re saying is that there is something we can do to reduce our uncertainty about it.1

Turning to the question of which careers have the most potential for impact, there are all kinds of things we could find out that would be relevant to the question: how many people you affect, how good those people say these effects are, how impactful experts say your work is, and so on. These are the ‘measurements’ we could make.

When we measure the good done by a career, there’s much more uncertainty than when we normally measure weight, but it’s just a difference of degree. It doesn’t mean one is measurable and the other is not. In both cases we make observations to reduce uncertainty, and that’s all that we ever mean by measurement.

2. Not defining what they mean by ‘making a difference.’

Lots of abstract terms seem immeasurable when we leave them only vaguely defined. ‘Making a difference’ is no different. Left unexpanded, it seems impossible to say which career makes more difference than another. But once you start to really define what you mean by ‘making a difference’, it becomes clearer that the concept can be measured.

For instance, we might think that making a difference has something to do with improving people’s welfare. So, we can immediately see that number of people affected is going to be relevant to how much difference you make. And that’s clearly observable.

Of course, you can debate this definition, but then the problem is no longer with measurement, but with what definition is good enough.

3. Not knowing what progress has already been made

People who claim ‘it’s impossible to measure the difference made by different careers’ are often not aware of the progress that has already been made on other relevant questions. Often, there are already answers proposed to relevant questions, and sometimes entire research movements dedicated to answering them.

For instance, the question of which charitable donation makes the most impact has been taken on by Givewell and Giving What We Can. On a larger scale, the Copenhagen Consensus has brought together some of the world’s top economists to assess which government policies could have the most positive impact on development.

Psychologists and economists have worked extensively on the question of what leads people to have happy and satisfying lives. This has involved creating ‘measures’ of life satisfaction, and then working out what kinds of life changes have the biggest effect on satisfaction. This is a big part in estimating which interventions lead to the most people with the most satisfied lives, a big part of estimating which careers are best.

Knowing about these efforts makes the idea of comparing careers in terms of impact immediately more tractable.

4. Thinking measurement is possible in principle, but too costly to be worth it in practice

More precisely – the value of information from measuring the impact of different careers isn’t high enough to justify the costs. This is a sensible thought, that might turn out to be well founded. But so far we haven’t seen claims of this sort justified with real back of the envelope value of information calculations.

On the contrary, since careers seem to differ a huge amount in impact, there’s a lot of value making just very slightly better choices about them. Moreover, since no-one has vigorously taken on the project of comparing careers in terms of impact before, there doesn’t seem to be any reason to think there aren’t useful discoveries that can be made with relatively little effort. So, it seems like we should expect high value of information for little cost.

5. Thinking that what counts as making a difference in a career completely depends on the individual’s values

We attempt to provide general statements about the potential value of different careers. But, so the objection goes, what’s valuable depends on the individual.

Well, firstly, that’s a controversial thesis in ethics. Some philosophers hold that value is an objective thing to be discovered. Others hold that it’s purely subjective. Raising this objection means taking sides in a controversial debate.

Secondly, even if it does depend on the individual, peoples’ values tend to overlap quite heavily in reality. We can all agree that, all else equal, saving lives is good and causing suffering is bad. This means that, although differences in value exist, there’s quite a bit we can usefully say about the value of different careers in general. And when it really does come down to a value judgement over which people reasonably disagree, we can just flag that decision point, rather than give up on the whole project.

6. It’s impossible to compare the value of different options

The thought is that different careers result in different kinds of outcomes, and that it’s impossible to weigh these options against each other. For instance, one career saves lives while another contributes to artistic culture. How could these possibly be compared?

This idea that some values are ultimately incomparable (or incommensurable in philosopher language) is again a controversial thesis in ethics with many arguments on either side. So saying that some values are incomparable is not an easy end to the debate.

More to the point, in reality we have to make trade-offs between different options. Your selection of one career over another is implicitly a judgement that one is better than the other. To do otherwise is just to choose randomly or thoughtlessly.

7. It’s measurable in principle, but there are more important things to focus on right now

This is strikes me as one of the most sensible objections. The idea is that although 80,000 Hours could focus on assessing the impact of different careers, we could get more value from investigating something else. In a recent interview, Holden suggested that we focus more on evaluating how to get a good person-job fit rather than on immediate social impact. This is a tough question, depending on judgements about the relative ease of investigating the two things, and how much value you get out of the corresponding reductions in uncertainty. Our current best guess is that it’s worth us doing some work on comparing the social impact of different careers, but we’ll constantly assess whether our efforts might be better spent elsewhere.

So those were the seven objections which I think most commonly lie behind our project to assess the potential for impact of different careers, and why I think they’re mistaken.2 I hope our most convincing rebuttal will be demonstrating that the project is possible by
actually doing it! Stay tuned for more on that over the next couple of months.

You might also be interested in:

Notes and References

  1. This view of measurement is eloquently put forward by Douglas Hubbard in How to Measure Anything 
  2. Also note that even if the good you might doesn’t have a scale, so in a sense isn’t ‘measurable’, it might still be possible to make a (partial) (ordinal) ranking of careers in order of goodness.