The poverty rate is an awful metric. It simply counts the number of people who earn below a certain income. So people dipping $1 below the line are counted the same as those dipping $10,000 below the line. So, poverty-rate alleviation measures can be implemented simply by focusing on the top end of those under the limit - of course, this doesn't help much. What we should really worry about is the poverty gap - the amount of income people need to get people above the poverty line.
> What we should really worry about is the poverty gap - the amount of income people need to get people above the poverty line.
Is that not showing an increase right now too? I mean, I don't disagree that serious analysis needs more info than a single number. But... what's the policy distinction here that you're trying to argue?
There's no mystery to the current situation: people are out of work due to the pandemic and making a lot less money, and that's bad.
Edit: two replies have responded by repeating the theoretical distinction and proposing scenarios where the two metrics would diverge. I understand that part. I'm asking if there is any evidence RIGHT NOW that this is the wrong metric to look at. Because naively, you'd really expect that people losing a lot of jobs to covid would simultaneously increase the population of people below the poverty line AND the average/median incomes of those people already below it. And if that's true, and it seems to be, why are we arguing about minutiae about metrics? The cause, and the solution, is the same in either case.
"Reduce the number of people in poverty by 50%" is a different goal than "Reduce the net poverty gap by 50%". A real headline can be that fewer people are in poverty than last year but the dollar sum of poverty is higher.
Effective policy considers the true goal - people are living better - and takes into account Goodhart's law as possible. I would say OP is arguing that the latter goal metric is better for humans than the former, and I agree.
If there are 5mm people below the poverty line and they all make $1 under the limit then you can close the gap, and reduce the poverty rate to 0 by giving each person a dollar.
If they are all making $10 below the limit then you can evenly distribute the money and reduce the gap by 10% while keeping the poverty rate the same. Or you can give all of the money to only 10% of the population and reduce the poverty rate by 10% while the gap remains the same.
I don’t know how this is typically measured (or measured in this specific case because I can’t find the actual paper), but do economists generally apply the same dollar threshold to everyone across the country, or do they account for regional cost of living?
Apparently the standard poverty threshold does not vary by region. There is a measure called the "Supplemental Poverty Measure" which does take into account certain cost-of-living adjustments, amongst other things. More info here: https://www.census.gov/content/dam/Census/library/publicatio...
It does not look like this research used that measure. Given the income disparity between urban and rural areas, I'm worried that this could paint a misleading picture about who exactly is suffering here.
(I'm definitely no expert here, just a lay person trying to interpret this.)
I don't think the arbitrary nature is as big of a deal as the first user suggests. If policy makers were truly optimizing for poverty rate reduction of that single metric than Goodhart's law would come into play - but they don't really do that. Other people use other thresholds and I've never seen any evidence that people just below the poverty line getting help is responsible for the shift.
If in the worst case more people are unemployed b/c the minimum wage increase, the poverty rate would increase. But the converse wont be true - even if the economy is booming and unemployment is zero, these folks are still going to be poor
