Usually, we think of charity as something nice to do, but no one would say you were doing anything wrong by spending all your money on yourself. It's your money, after all.

There's a different way of seeing things. Here's an argument that it's wrong to not give to charity:

### Argument #1:

1. Many people suffer terrible diseases which are cheap to cure.

2. If we can prevent something bad from happening, without giving up anything important, we should.

3. I can prevent people dying by giving more to effective organisations.

4. By giving more to effective charities than I do, I would not be sacrificing anything important.

5. Therefore, I should give more to effective charities.

Quite a few groups act out some version of this, in the form of 'tithing': you can see it in Mormonism, Islam and Judaism. One group, the strong consequentialists, go further though:

### Argument #2

1. Many people suffer terrible diseases which are cheap to cure, and this is really bad.

2. If we can prevent bad things without giving up anything equally important, we ought to.

3. It is in our power to cure instances of disease.
• Most rich-world people have enough disposable income for many suffering people.
• There are organisations who deliver healthcare amazingly efficiently.

4. My having luxuries is not as important as the fatal preventable disease of others.

5. Therefore, give away whatever you earn, after you've bought necessities for yourself.

### What's the difference?

The change is in line (2): it points out that, even if it's important to me to visit my friend who lives across the globe in Japan, it's not as important as someone else's life or death. And that this difference should affect my actions.

This line of reasoning leads quickly to something much rarer: the idea of a maximum wage. Call the amount of money a strong consequentialist is allowed the maximum consequentialist income (MCI).

Three ideas for how to calculate the MCI 100:

### A: Give yourself the world average.

First let's try from the top down. I'm no less worthy than anyone else, so maybe I can just give myself the global average income.

• What's the average world income?: Something like £6,486 ($74tn / 7.13bn / £). But using the mean is ridiculous, because it puts too much weight on the unbelievably rich. If we take the median instead we get the abysmal £1,772 at PPP. 99 • As you've probably noticed, this method has gone off the rails already: for a start, the most saintly (or monstrous) utilitarian must accept not the median but the marginal income ($730 a year). But much more because no-one can begin to live independently on £2000 a year, in Britain, except via homelessness or backwoods survivalism, neither of which conduce to many effective altruist jobs. (So the Singer principle is breached, and we try something else.)

Of course this impossibility (even when you adjust for purchasing power) is the point about poverty, or 'average world income' as we're calling it.
• If that were not bad enough, there are also hard limits to the aggregate amount we should give, and these fall well before the marginal utility limit (at which we’re no better off than those we’re helping). There are least three large rule-utilitarian problems with a whole society becoming Singerian:
1. Economic sustainability for donors. We're imagining an effective 75% remittance rate: the economic destructiveness of this cannot be overstated. A massive increase in giving is also a decrease by orders of magnitude in consumer spending, and so a proportionately tinier economy (millions of lost jobs, deflation, brain drain). There's also the foregone replacement of depleted capital stock, estimated at "11%" a year for Britain, whatever that means (chart 28 here) 200bn? I concede this point (as does Singer by the way), though the more idealistic of us can argue there’d be non-negligible positive effects from decreasing consumption, if the change was gradual enough. (Think about the environment alone.) When conceded, this leads to the oldest sticking point in the distributive book: the smaller slice of the bigger pie. We might give more overall with 15% of GNP than 40%, since total GNP would be that much bigger (and the country would still be there the following year). Where the line should be drawn is legendarily hard to say.

2. Perverse effects for recipients. Resource flows just one percent as large as the donation from hypothetical hyper-altruist Britain have in the past had 'hundreds' of severe perverse economic effects, from inflation to corruption to civil war. We sidestep this somewhat by emphasising that our work would be evidence-based (the 'effective' bit in effective altruism), but that much money is very unpredictable and genuinely hazardous.

3. Cultural decline. Don't underestimate this; what becomes of culture, if no one pays for gigs or exhibits or such? (Even DIY nonprofit things rely on somebody willing to make a loss.) The consequentialist should be careful not to tear up all other kinds of value in pursuing economic justice, and become a mere 'machine for the redistribution of wealth'.

But these are irrelevant for our decision, since we can safely assume that we're acting 'at the margin', where there will be no mass movement toward radical giving.

One issue which does need solving now is the problem of composition. Note first that we know that the NGO Schistosomiasis Control Initiative can deworm someone for $1.28 including admin - a permanent substantial boost to one person's health, nutrition, and mental performance. Nothing I can get for myself can match this. (Nothing I could spend$1,000 on could.) So, naively it seems that any figure I set for the MCI could be eroded further, a dollar at a time, by applying the marginal utility rule - "Is this pie worth more to me than a deworming to someone in need? No." (ad infinitum).

The answer is to reject the use of marginal utility on its own; you need to derive a minimum baseline first. On to which:

### B: Give yourself the cost of necessities.

Let's try from the bottom-up instead, adding basic annual costs, again for a childless Briton in OK health living in the average town:

• Rent: £2700. 101
• Food: £1800 102
• Toiletries: £100? 103
• Clothes: £100. 107.
• Bills: £1100 104
• Transport: £1300. 105
• 'Participation' (going out, gifts, misc): £500. 109

• MCI = £7400 + tax 106

Surprisingly neat. Of course most British people would consider this anti-poverty measure to be poverty. But that's an abuse of the word. This does propose you live a student life forever, but that's the price of virtue by these lights.

### C: Give yourself a 'living wage'

We could accept the estimates of rich-world activists for a rich-world 'living wage'. Lovely people like the Joseph Rowntree Foundation and MIS calculate a decent domestic minumum of about £16,000 (at 2013 prices).

In pushing in the direction of the much lower MCI, we might seem to oppose the push for better wages for working-class British people (who are nonetheless, and at worst, still in the global top 20%).

But thinking that the money should come from the lower end of the income spectrum is the same terrible mistake that people who rant about benefit scrounging instead of tax avoidance make: there's a vast amount of room for far less harmful redistribution at the top end.

The calculator here gives the UK living wage as £16,852 for a singleton, at minimum (though it doesn't handle the flat-sharing setup I used in [B]). This is the amount polled British people think one can really live on - "participating fully in the economic, social, and political life of the society in which they live".

Who am I to say this is 'too much'?

Well it's not too much. In fact it's not enough, given most people's life goals. It's just that the Schistosomiasis Control Initiative's $1 offer remains, and really should force you to think about how much you actually need another holiday. Using method B's baseline instead of the above living wage is equivalent to giving 4,800 more children, every year. Well, what was the point of that? Does anyone actually live like this? A few do, yes. But I always get annoyed by people who dismiss consequentialism for being too demanding: why should it be easy to take the best course of action? Even if it was impossible for humans to live like this (and it is not), it would be extremely valuable information, to know how far from the best course I am trailing. 1. World median household per capita divided by$:£ rate.
2. For a single, childless, working-age person with no disabilities.
3. (Excluding London.)

Average non-London rent: £684 a month; 684 x 12 = £8208 a year. This is for "homes", so it covers massive 5 bedroom places as well as bedsits though. The mode is 3 bedrooms (p.26 here) so let's run with that.

Assuming you can find people to share with, we'll divide this by (say) three tenants = £2736 each. The median will be lower, but I don't want to just guess.
4. i.e. £5 a day. This is with many pulses and power smoothies.
5. Yeah I dunno.
6. Heating:
Electricity:
Water: £
Internet: £120
Council tax:

UK average is £1,440, which seems high. (Mine was £940 last year, split between three people.)
7. £26/week according to the JRF.
8. Tax depends entirely. It is £6500 if you're on the national median wage: on a £25k income, £5k income tax, £1.5k council tax.

Also without juggling tax brackets with deductibles.
9. Thrift store threads plus underwear and socks.
10. JRF says more like £2,400 for 'participation', but they don't have bookish consequentialists in mind.
Because giving money is regarded as an act of [unnecessary] charity, it is not thought that there is anything wrong with not giving. People do not feel in any way ashamed or guilty about spending money on new clothes or a new car instead of giving it to famine relief. (Indeed, the alternative does not occur to them.) This way of looking at the matter cannot be justified... To give money away is not charitable, or generous.
– Peter Singer

That's why a bear can rest at ease with just the bare necessities of life...
– Terry Gilkyson

What amount can you allocate to yourself, if you accept maximizing consequentialism? Call this amount the maximum consequentialist income (MCI), all earnings above which an ideal utilitarian gives to the most effective causes.

We all know the minimum wage, but the idea of a maximum wage, like this MCI, is not common. The Soviet nomenklatura had one at first, though it was easily subverted and later abandoned completely. The Venezuelan civil service supposedly has one at the moment, though I imagine the same forces are acting on it as we speak. Those forces include our assumption of a right to unlimited accumulation, the Lake Wobegon effect, and magical thinking about the role of our effort in determining our earnings. Yet at least one group of people voluntarily impose a maximum wage on themselves to help others. The classic argument for this, in case you're wondering what I'm on about:

### Singer's argument

1. Many people suffer from lack of food, shelter or medical care.

2. Suffering lack of food, shelter or medical care is bad.5

3. If it is in our power to prevent something bad from happening, without thereby sacrificing anything of comparable moral importance, we ought to do it. 7

4. It is in our power to alleviate people's lack of food, shelter or medical care.
• Since most British people have disposable income, broadly defined.
• Since there are anti-poverty organisations who verifiably do alleviate poverty.4

5. Aside from what I spend on necessities, my consumption is not of comparable moral importance.

6. Therefore, give away what you earn above the cost of necessities (the MCI). 2

### Argument from the nature of world wages

A further argument for the MCI, from the nature of world wages:

1. Accidents of birth place some of us in positions of economic advantage, and most others in more or less terrible poverty. (Global moral luck)
2. Your income has much more to do with the average productivity of your country than your own productivity. 6
3. We can only have earned things that we have some control over. 1
4. Therefore, you didn't strictly earn the global surplus you receive.

### Argument for moderate consequentialism

A less demanding and thus less unpopular form of the argument, which doesn't entail the MCI, runs as follows:

1. Suffering and death from lack of food, shelter, and medical care are bad.
2. "If it is in our power to prevent something bad from happening, without thereby sacrificing anything morally significant, we ought to do it.
3. I can prevent people dying by giving more money to effective anti-poverty causes than I do.
4. By giving more money to anti-poverty causes than I do, I would not be sacrificing anything morally significant.
5. Therefore, I should give more to anti-poverty causes.

If these arguments are right, it falls to the consequentialist to give away what they have not really earned, and what they cannot use better than the world's extremely poor could. The effective altruism movement tends to focus on percentages of income given or pledged - and this is only prudent, since it prevents the relatively low-income among us being turned off by stories of billionaires giving millions to thousands.

(Incidentally, this calculator shows just how rich you are, even on UK minimum wage, as I was when I first wrote this. 3)

Three ideas for calculating the MCI 100:

### A: Give yourself the world average.

First let's try to calculate the MCI from the top-down. Since I'm no less worthy of consideration than anyone else, perhaps it's as simple as allocating oneself the global average income. For one childless person in OK health, then:

• What's the average world income?: Something like £6,486 ($74tn / 7.13bn / £). But the mean is ridiculous, taking into account as it does the aneurysmally rich as if normal; let's take the median instead: the abysmal £1,772 at PPP. 99. • Given the unemployed (6-9%), retirees (14%), children (26%) and those otherwise unable to work, only about two-fifths of the world is employed. So what's the average world salary? This is a hard question. Maybe £11,290. (Anyway it's not relevant for the fully strict utilitarian, see below.) • As you've probably noticed, this method has gone off the rails already: for a start, the most saintly (or monstrous) utilitarian must accept not the median but the marginal income ($730 a year). But much more because no-one can begin to live independently on £2000 a year, in Britain, except via homelessness or backwoods survivalism, neither of which conduce to many effective altruist jobs. (So the Singer principle is breached, and we try something else.)

Of course this impossibility (even when you adjust for purchasing power) is the point about poverty, or 'average world income' as we're calling it.

### Perverse dynamic effects

There are hard limits to the aggregate amount a society should give, and these fall well before the marginal utility limit (at which we’re no better off than those we’re helping). There are least three large rule-utilitarian problems with a whole society becoming Singerian:
1. Economic sustainability of donors.
We're imagining an effective 75% remittance rate: the economic destructiveness of this cannot be overstated. A massive increase in giving is also a decrease by orders of magnitude in consumer spending, and so a proportionately tinier economy (millions of lost jobs, deflation, brain drain). There's also the foregone replacement of depleted capital stock, estimated at "11%" a year for Britain, whatever that means (chart 28 here) 200bn? I concede this point (as does Singer by the way), though the more idealistic of us can argue there’d be non-negligible positive effects from decreasing consumption, if the change was gradual enough. (Think about the environment alone.) When conceded, this leads to the oldest sticking point in the distributive book: the smaller slice of the bigger pie. We might give more overall with 15% of GNP than 40%, since total GNP would be that much bigger (and the country would still be there the following year). Where the line should be drawn is legendarily hard to say.

2. Perverse effects for recipients.
Resource flows just one percent as large as the donation from hypothetical hyper-altruist Britain have in the past had 'hundreds' of severe perverse economic effects, from inflation to corruption to civil war. We sidestep this somewhat by emphasising that our work would be evidence-based (the 'effective' bit in effective altruism), but that much money is very unpredictable and genuinely hazardous.

3. Cultural decline.
Don't underestimate this; what becomes of culture, if no one pays for gigs or exhibits or such? (Even DIY nonprofit things rely on somebody willing to make a loss.) The consequentialist should be careful not to tear up all other kinds of value in pursuing economic justice, and become a mere 'machine for the redistribution of wealth'.

But actually the above warning are irrelevant for our personal decision, since we can safely assume that we're acting 'at the margin', where there will be no mass movement toward radical giving.

One issue which does need solving now is the problem of composition. Note first that we know that the NGO Schistosomiasis Control Initiative can deworm someone for $1.28 including admin - a permanent substantial boost to one person's health, nutrition, and mental performance. Nothing I can get for myself can match this. (Nothing I could spend$1,000 on could.) So, naively it seems that any figure I set for the MCI could be eroded further, a dollar at a time, by applying the marginal utility rule - "Is this pie worth more to me than a deworming to someone in need? No." (ad infinitum).

The answer is to reject the use of marginal utility on its own; you need to derive a minimum baseline first. On which note:

### B: Give yourself the cost of necessities.

Let's try from the bottom-up instead, adding basic annual costs, again for a childless Briton in OK health living in the average town:

• Rent: £2700. 101
• Food: £1800 102
• Toiletries: £100? 103
• Clothes: £100. 107.
• Bills: £1100 104
• Transport: £1300. 105
• 'Participation' (going out, gifts, misc): £500. 109

• MCI = £7400 + tax 106

Surprisingly neat. Of course most British people would consider this anti-poverty measure to be poverty. But that's an abuse of the word. This does propose you live a student life forever, but that's the price of virtue by these lights.

### C: Give yourself a 'living wage'

We could accept the estimates of rich-world activists for a rich-world 'living wage'. Lovely people like the Joseph Rowntree Foundation and MIS calculate a decent domestic minumum of about £16,000 (at 2013 prices).

In pushing in the direction of the much lower MCI, we might seem to oppose the push for better wages for working-class British people (who are nonetheless, and at worst, still in the global top 20%).

But thinking that the money should come from the lower end of the income spectrum is the same terrible mistake that people who rant about benefit scrounging instead of tax avoidance make: there's a vast amount of room for far less harmful redistribution at the top end.

The calculator here gives the UK living wage as £16,852 for a singleton, at minimum (though it doesn't handle the flat-sharing setup I used in [B]). This is the amount polled British people think one can really live on - "participating fully in the economic, social, and political life of the society in which they live".

Who am I to say this is 'too much'?

Well it's not too much. In fact it's not enough, given most people's life goals. It's just that the Schistosomiasis Control Initiative's $1 offer remains, and really should force you to think about how much you actually need another holiday. Using method B's baseline instead of the above living wage is equivalent to giving 4,800 more children, every year. ### Bibliography It was never contended by a sound utilitarian that the lover should kiss his mistress with an eye to the common weal – John Austin The above took it for granted that strong individualist economic consequentialism is the right way to think about the right thing to do. I don't necessarily, though I respect it deeply. The real problem of utilitarianism (and other very demanding ethical schemes) is not that they require too much from us, but that they could easily be destructive of other (lesser) components of goodness when taken beyond the margin. In short, a nasty consequence of maximizing consequentialism is that it invites us to disparage perfectly good activities (e.g. chess, poetry and anthropology) in the name of "People are dying!". The world according to Singer presents us with the choice of being irritatingly single-minded, or immoral. My shonky solution is that we only have to maximise the extremely publicly-consequential components of our lives - roughly, career, consumption pattern, and ideology. Only those are worth the time it takes to do the research and run the optimisation properly, anyway. Maximise the big, satisfice the small. In this gap I plan to live. 1. This is the 'Control principle': as applied here it states 'if an accident of birth put me on track for higher wages, then I haven't earned the surplus wage I receive.') 2. This form of the argument mentions poverty relief for reasons of availability: these things don't work well without an example, and extreme poverty relief is an uncontroversially good thing. But the case for animal rights work or existential risk mitigation is often very strong indeed and I intend no prejudice. 3. I wrote the first draft of this in 2011, but this claim was a bit underhand even then: I had an economics degree, a stock of human capital which (usually) disqualifies people from being systemically poor. 4. Given reasonable interpersonal values, these organisations can use a unit of income thousands of times better than we can use it for ourselves, a margin far beyond econometric error or admin deadweight.) 5. If you skip to the "Simple" version of this post, you'll see that there is no premise for "this is bad". It takes a certain level of sophistication before you need telling. 6. This follows from the Baumol effect plus closed borders. 7. This is known in the literature as the Strong Singer Principle, and this is Peter Singer's phrasing. 8. World median household per capita divided by$:£ rate.
9. For a single, childless, working-age person with no disabilities.
10. (Excluding London.)

Average non-London rent: £684 a month; 684 x 12 = £8208 a year. This is for "homes", so it covers massive 5 bedroom places as well as bedsits though. The mode is 3 bedrooms (p.26 here) so let's run with that.

Assuming you can find people to share with, we'll divide this by (say) three tenants = £2736 each. The median will be lower, but I don't want to just guess.
11. i.e. £5 a day. This is with many pulses and power smoothies.
12. Yeah I dunno.
13. Heating:
Electricity:
Water: £
Internet: £120
Council tax:

UK average is £1,440, which seems high. (Mine was £940 last year, split between three people.)
14. £26/week according to the JRF.
15. Tax depends entirely. It is £6500 if you're on the national median wage: on a £25k income, £5k income tax, £1.5k council tax.

Also without juggling tax brackets with deductibles.
16. Thrift store threads plus underwear and socks.
17. JRF says more like £2,400 for 'participation', but they don't have bookish consequentialists in mind.