Hacker News Reader

Comment by jandrewrogers | Hacker News Reader

Social mobility measures based on income (PPP) percentiles are misleading because they don't account for the hugely different income variances between countries. In countries with low income variances like Denmark, a small change in your absolute income will be deemed "socially mobile" whereas a much larger increase absolute income in countries with high income variances like the US will be deemed as not "socially mobile" even when it is vastly better for the average individual in absolute income terms.

Individuals don't care about their ability to change their percentile, they care about their ability to change their total income. I think most individuals would prefer a "non-mobile" $60k increase in income versus a "mobile" $40k increase in income. You can't spend percentiles.

View on HN

>"Individuals don't care about their ability to change their percentile, they care about their ability to change their total income"

This is completely wrong. People care intensely about their pecking order (looks, money, accolades, etc). Beyond this, there are purely functional reasons to care about your relative income. Limited resources lead to competition.

Nobody cares if their GPA is higher than their parent's was. They care if it's higher than their peers who are trying to get into the same universities. Ditto for housing prices. What matters isn't the dollar value of the property you can afford, only whether or not you can afford to be in a top school district. Politicians don't win by being able to spend more on campaigns than politicians of the previous generation either. They care if they can spend enough to win against their current opponents.

In human society, the relative comparison dwarfs the absolute.

But these percentiles are a zero sum game (the distribution of percentiles is always uniform 0-100).

Which makes me wonder why one would laud upward mobility in percentiles, when any such upward mobility must be balanced out by equal and opposite downward mobility.

Because the lack of downward mobility suggests the same problem as the upward mobility: your success in life is based on your parents income not your own merits.

If you want a meritocracy you have to accept that untalented/unmotivated kids of rich parents may become relatively poor.

Well if you really want to optimize for social churn, couldn't you just randomize all the newborn babies at hospitals or something?

This has nothing to do with genetics.

>Individuals don't care about their ability to change their percentile, they care about their ability to change their total income. I think most individuals would prefer a "non-mobile" $60k increase in income versus a "mobile" $40k increase in income. You can't spend percentiles.

I highly dispute this. Take it as a completely different world view but I think if people were given a choice:

1 >a nominal income of $100k for you and a nominal income of $200k for all your neighbours

2 >a nominal income of $75k for you and a nominal income of $60k for all your neighbours

A lot of people would opt for choice 2. I myself am not sure whether I'd go for 1 or 2.

Depends. Can I opt for the $100k income and then move to the $60k neighborhood?

Yes, that is choice 2

Yes, the Economist has an article on this relativity every two years or so :-)

There are two reasons for why people invariably choose 2>

- people are competitive and want to feel superior to others (or at least above average)

- the price of everything adjusts to median income, so if you live in a $60K state/country with a $75K income, everything is more affordable for you.

This is a good point, but I think individuals care a lot more about their percentile than you would expect - their benchmark for whether their income is 'a lot' or 'a little' is how much they see the people around them making.

Of course, you can't have very much social mobility in high income variance countries anyway if you want to preserve those precious income inequalities.

Reply on news.ycombinator.com