The Leaky Bucket Experiment

I’ve recently finished reading the 1975 book Equality and Efficiency: The Big Tradeoff by Arthur Okun. Although it’s a little older and some of the subject matter is no longer super relevant for today, it’s nonetheless a very interesting book for anyone who likes thinking about political philosophy and justice. The title accurately represents the subject matter, Okun is making the case that equal rights (for anything from free speech to access to firefighters) can never have economically efficient outcomes. On the other hand, economic efficiency will never lead to equality (unless by pure chance). In a perfectly free and well functioning market we should expect no deadweight loss, but we should not expect, for instance, everyone involved to have equal incomes. Or to put it another way, if you distribute equal quantities of something to all citizens, be it rights or wealth or anything, and you do not allow people to sell/trade these things, then you cannot expect the distribution pattern to be Pareto optimal – we always face this tradeoff.

Okun is clearly a very well read and thoughtful guy, and although I found myself disagreeing with him through much of the text, I had to stop and give pause to his clear and intriguing logic and thought experiments. One such example of this is what he called “the leaky bucket experiment”.  

At the time Okun was writing, he estimated that taking less than 10% of the income from the top 5% of income earners in America, could give the bottom 20% of earners 25% more income. This program of income distribution has an unsolved problem however, that “the money must be carried from the rich to the poor in a leaky bucket”. Okun continues: “some of it will simply disappear in transit, so the poor will not receive all the money that is taken from the rich”.

The leak in Okun’s metaphor is meant to represent all inefficiency. Therefore, it doesn’t merely represent administrative costs, but also the leak represents the negative economic effects such as incentive distortions on things like work effort and savings/investment. In short, a leak in the bucket is anything that shrinks the overall economic pie size (or at least slows its growth more than otherwise).

So the question is how much leak do you think society ought to allow? In looking at this answer, I think it’s most interesting to look at the two extremes – people who wouldn’t tolerate the distribution even if there was no leak at all, and people who would tolerate the distribution even if it all leaked out.

If you’re someone in the first camp who thinks redistribution is wrong even if there is no leak whatsoever, remember, this commits you to the position that taking from the rich and giving to the poor is bad even if (by design of the thought experiment) there are no negative economic consequences. You think taking from someone is inherently wrong.

Likewise, if you say the distribution is justified even if it all leaks out (the equivalent of just burning or destroying 10% of the wealth of the richest members of society) than you are concluding that this is good even though there are (again by design of the thought experiment) only negative economic consequences. You think that inequality is inherently wrong.

Of course people’s preference could fall anywhere between these two extremes, but what I find so interesting about this thought experiment is it exposes a misconception that I think people often face – that people often underestimate the knowledge of their opposition. For example, I often hear arguments against wage equality going something like this:

“People who favour equality see the economy as a static pie that doesn’t grow and they want to divide it up as “fair”, but what they don’t understand is that the market process allows for the pie to grow, and interference with this can make everyone worse off. If only they understood that market outcomes lead to bigger total pies, faster pie growth and less absolute poverty, then they would get why equality of income shouldn’t be advocated for.”

But then here is a guy like Okun, who clearly understands economics more than most (not too many economists get laws named after them) and understands the fixed pie fallacy quite well, and yet he says we should go for this transfer, and he’d tolerate leakage. We cannot say he comes to his conclusion from lack of understanding or knowledge, he just inherently values equality in and of itself. You may disagree with him, but he’s not necessarily wrong or mistaken.

In fact, what I think we sometimes forget is that this is ultimately a question about what you subjectively value, as Okun says, “your answer cannot be right or wrong – any more than your favorite flavor of ice cream is right or wrong”.

So what amount of leak would you tolerate?


p.s. If you’re curious, Okun says he would probably tolerate such a wealth redistribution with up to a 60% leakage.


3 thoughts on “The Leaky Bucket Experiment

  1. Hi Casey,

    That’s a good question but to answer it… you have to assume that you have the possibility to measure these leaks. Or that’s precisely the problem because we can’t erase and do a different political decision and measure the difference. You can measure the cost of bureaucracy but not the disruption of the market process.

    We can explain and show all leaks by using reason or empirical evidences but we can’t measure it.
    For this reason, we are kept on normative statements and arbitrary decisions…

    You know that I’m in the first extremist group 😀 but I’m open to discuss.

Leave a Reply