Working to Give vs. Working to Improve

The de facto wisdom within the Effective Altruist community seems to be that the good that most people pursuing Earning to Give comes primarily by their donations, rather than the work that produced those donations.

For instance, take a finance professional earning and donated 30% of his/her income. This may be approximately $30,000 per year, and it’s very easy to understand how this money improves the world if spent wisely. At a lower bound (for those who believe in even more important causes), this money would directly fund malaria nets via the AMF, saving approximately 15 lives and preventing more sicknesses. It’s much more difficult to estimate the impact of the more direct labor that the professional provides via their work in finance. However, given the magnitude of the benefit from the donation, it generally been been assumed that the direct work is significantly less useful to society.

Still, I hear some claims that “Working to Give” in conventional fields like finance or software development (as opposed to unconventional fields like meta-research) may be as impactful or more-so than “Working to Give.” The primary example that comes to mind is Holden, Exective Director of Givewell, who has stated,

Claim #1: Some careers provide many (10-10,000x) times the work benefit as others. Claim #2: If some “decent” careers are approximately as useful as donations to the AMF, then “amazing” careers must be substantially better still. Claim #3: If “Working to Give” were more impactful than “Earning to Give”, then donations to work education may be substantially more important than ones to existing Givewell approved charities. Claim #4: For most software engineers, “Earning to Give” is much more useful than “Working to Give”.

Claim 1

Consider the following careers. How much value or utility to you think each one of them produces to the world through their work?

  • A high-frequency trader
  • A pornography website host / web engineer
  • A Farmville advertising engineer
  • A LinkedIn email marketing automater.
  • A software engineer at Google
  • A software engineer at (their ‘do-good’ division)
  • The Wikipedia software team

I’m going to guess that almost everyone would agree that at least one or two of these provide negligible or negative world value. There seem to be zero-sum markets where benefits to one firm directly result in less profit to another, like high-frequency trading. In other markets, web companies definitely do things that aren’t in the benefit of most people. For example, I bet that most spam emails are a net negative for society, and yet web companies have entire departments set around spreading these.

On the other hand, there have been some software projects that definitely have been quite useful. Wikipedia comes across as one that started out with a tiny team but produced what seems to be lots of value. So did Wikileaks. Peter Singer recently mentioned that he considers that work for internet services in developing countries could provide huge returns. So there have been software engineers / hackers who have provided lots of utility (though the amount is unclear), and others in different fields who have likely produced very little.

Perhaps more obvious is the fact that most other fields seem to have power laws when describing the amount of world benefit provided. We know that in the charity sector some charities are at least 10-1000x as efficient at others at converting money into utility. With individual researchers this seems to be the case, with entire research fields it seems to be the case. The 80/20 principal appears incredibly common, and it would be very surprising to me if it didn’t exist in terms of the benefits in different jobs. Therefore it seems very likely that different careers would provide expected output utility of different orders of magnitude.

Claim 2

If it’s true that average employees or software engineers produce more value in their work than in their donations(if spent optimally), it seems like careers that produce results 10-100x as high would make the AMF donations inconsequential in comparison. This would appear to just be common sense if one assumes a Power Law.

Claim 3

If it’s true that for people who choose “efficient careers”, they will produce 10-100x as much value from their work than their donations, then what’s by far the most important is that they maximize their work rather than that they maximize their donations. If this is true, we should be doing much more of the following.

1: Personal Education & Training to Insane Levels

Personal coaches and tons of investment in self promotion would be in order, to very high amounts.

$80,000 per year, $20,000 donated per year.

20x as efficient as donation, so that’s effectively $400,000 of world productive value per year.

A 5% improvement in productivity would be worth %20,000. My guess is that $20,000 of money spent on personal efficiency could go much further than increase productivity by 5%. $20,000 per year would be about $1.7k per month. This could buy you several monthly coaching sessions, food and all home issues dealt with by fancy personal assistants, home cleaning etc.

Much of this would be to work as much as possible to produce as much output as possible. Weekends and long nights as much as one doesn’t burn out or could sustain.

2: Very high investment in one’s happiness

as sa fas sf asa as aasd

2: Investment in education of others.

Claim 4 So my guess is that anything matters. It also includes that nothing is really what it would seem otherwi