What's the goal? What's the problem?
A good system doesn't provide equal outcomes. An education system should provide better outcomes to those who work hard. An education system which provides equal outcomes to both the hardworking and the lazy is not fair. But an education system which provides more benefit to a talented, hardworking Korean than it provides to a talented, hardworking Mexican is also not fair.
Whether we're talking about systems of education, business or government. The goal is a fair and just system. A system that benefits people based on what they *do* and not based on who they *are*. A system where the outcomes are not determined by a person's ethnicity, gender, religion, etc.
What's fair?
It is often super easy to see when a person has been treated unfairly. We are all pretty good at recognizing injustice when we look at the experience of an individual. But at a larger scale, it can be really hard to say just how just or unjust a system is. No system is ever perfectly just, every system is going to eventually treat some people unfairly. But how bad is it? Do we need to throw out the whole thing and start over, or can we just tweak some incentives here and there and get a better result.
So how do we measure the relative justice of a system? Its not perfect, but the best way I know of to try to measure fairness is to look at the long term distribution of outcomes across ethnicities, genders, religions etc. If the outcomes are statistically similar across these groups, it is likely the system is relatively fair.
But here's the thing, statistically similar outcomes across people groups does not mean a system is fair. Its a useful measurement, but not a guarantee. A system which randomly assigns outcomes will result in statistically similar outcomes, but that is not a fair system. If I've been accused of a crime, I want a judge or jury to decide, not a random number generator. Statistically similar outcomes is not the goal, fairness is the goal. Justice is the goal.
If we have an unfair system, replacing it with randomly generated outcomes does not solve the problem. But for those who have lost sight of the real goal and become fixated on statistically similar outcomes, random results perfectly hides the problem. This is why I don't like quotas. If we use quotas to guarantee equally distributed outcomes we haven't made the system fair. Quotas do not magically make unfair systems into fair systems. Quotas just skew the results in a way that makes it more difficult to see the unfairness of the system. They undermine the one useful metric we had for measuring the fairness of a system.
Quotas don’t make systems fair they just make it more difficult to see the injustice of the system.
Ashley
A good system doesn't provide equal outcomes. An education system should provide better outcomes to those who work hard. An education system which provides equal outcomes to both the hardworking and the lazy is not fair. But an education system which provides more benefit to a talented, hardworking Korean than it provides to a talented, hardworking Mexican is also not fair.
Whether we're talking about systems of education, business or government. The goal is a fair and just system. A system that benefits people based on what they *do* and not based on who they *are*. A system where the outcomes are not determined by a person's ethnicity, gender, religion, etc.
What's fair?
It is often super easy to see when a person has been treated unfairly. We are all pretty good at recognizing injustice when we look at the experience of an individual. But at a larger scale, it can be really hard to say just how just or unjust a system is. No system is ever perfectly just, every system is going to eventually treat some people unfairly. But how bad is it? Do we need to throw out the whole thing and start over, or can we just tweak some incentives here and there and get a better result.
So how do we measure the relative justice of a system? Its not perfect, but the best way I know of to try to measure fairness is to look at the long term distribution of outcomes across ethnicities, genders, religions etc. If the outcomes are statistically similar across these groups, it is likely the system is relatively fair.
But here's the thing, statistically similar outcomes across people groups does not mean a system is fair. Its a useful measurement, but not a guarantee. A system which randomly assigns outcomes will result in statistically similar outcomes, but that is not a fair system. If I've been accused of a crime, I want a judge or jury to decide, not a random number generator. Statistically similar outcomes is not the goal, fairness is the goal. Justice is the goal.
If we have an unfair system, replacing it with randomly generated outcomes does not solve the problem. But for those who have lost sight of the real goal and become fixated on statistically similar outcomes, random results perfectly hides the problem. This is why I don't like quotas. If we use quotas to guarantee equally distributed outcomes we haven't made the system fair. Quotas do not magically make unfair systems into fair systems. Quotas just skew the results in a way that makes it more difficult to see the unfairness of the system. They undermine the one useful metric we had for measuring the fairness of a system.
Quotas don’t make systems fair they just make it more difficult to see the injustice of the system.
Ashley