I like looking at the extremes. If it were actually true that capitalism causes selfishness, then what are the odds that the world’s most free market country would also be the world’s most egalitarian country? But it is.
If it were actually true that fertility rates are falling because it’s too costly to raise kids these days, then what are the odds that the biggest fall in fertility in the entire world would have occurred in the country with the fastest economic growth over the past 60 years? But it did.
Today, I found another example:
As homelessness skyrockets across the country, Mississippi’s numbers are trending in the right direction.
According to USAFacts, 3.3 out of 10,000 people experienced sheltered homelessness in Mississippi last year. With 982 people experiencing homelessness in Mississippi in 2023, the state has the lowest rate of homelessness nationwide. This accounts for about 0.0015% of America’s homeless population. This is over five times lower than Mississippi’s share of the country’s overall population.
This link provides data for all fifty states:
Unfortunately, the color coding here is total homeless population, which is why Wyoming appears lower than Mississippi. If you go to the link and click on each state, you’ll find the per capita figures, which are more meaningful. New York and Hawaii have the highest rates, although elsewhere I read that California has the most homeless people actually living on the streets. Many homeless New Yorkers are in shelters. Mississippi is lowest in per capita terms.
So here’s my question: If poverty actually did cause homelessness, then what are the odds that the lowest rate of homelessness would occur in America’s poorest state? Even if there were no correlation, the odds would be only 1 in 50. If there actually was a positive correlation between poverty rates in a state and homelessness, then it would be even more surprising to find Mississippi having the lowest homeless rate.
In my view, homelessness has multiple causes. For homeless people in shelters, restrictions on building are a major cause. For those on the street, drugs and mental illness also play a big role. Poverty is far down the list.
READER COMMENTS
Devin
Jun 3 2025 at 10:19pm
Supplemental poverty, which accounts for housing costs, may be a better measure.
California leads the nation in that measure. I’d speculate supplemental poverty in urban areas specifically might be the measure with the greatest correlation.
steve
Jun 4 2025 at 9:32am
That measure or some other measure that accounts for relative poverty would seem to be more accurate. If you have little money and you live in a relatively poor area that wouldn’t seem to be as bad as living in a relatively wealthy area. In the poor area of coal country where I worked for a long time there was no homelessness even though there were a lot of poor. However, the area has tons of deserted homes. On top of that, having worked with homeless at times, there isn’t much infrastructure to hide out in like there would be in a large city and the weather is awful much of the year.
There are lots of poor people and relatively few homeless. However, of those that are homeless the huge majority are poor. Those that actually have means and are homeless largely have mental illness problems.
Steve
Craig
Jun 3 2025 at 10:24pm
“If poverty actually did cause homelessness, then what are the odds that the lowest rate of homelessness would occur in America’s poorest state?”
‘Red dirt rich'{?} and from what I have seen in rural areas of TN is that poor people aren’t so much homeless but rather they live in straight up dilapidated/poor quality housing. Very common the old single wide, when that gets too old many will buy an RV and plop that on the property and just live in that. Beyond that they do the tiny home/shed.
“sheltered homelessness”
I am also curious about this term because this sounds like something where homeless individuals go and take shelter in some kind of homeless shelter? Google maps showing a complete dearth of homeless shelters, just generally, in the Southeast. Is it possible there are homeless individuals and they just aren’t going to shelters or being counted? The most common form of homeless I see in TN and FL are people living in cars. In NYC I’d see individuals living rough out on the street.
john hare
Jun 4 2025 at 4:31am
I have been living in old campers for the last 15 or so years. Does that make me technically homeless. Or just someone choosing to invest elsewhere?
Jon Murphy
Jun 4 2025 at 7:04am
This has been my (albeit limited) experience as well. When I lived in New Hampshire, I volunteered at a homeless shelter during the winter. Most of the individuals we had were not poor. They had families or other significant resources they could draw on. Just, for whatever reason, they chose not to draw on those resources.
Garrett
Jun 4 2025 at 8:23am
The first “but it is” link is broken for me
Scott Sumner
Jun 4 2025 at 3:01pm
Sorry, I fixed it.
Jose Pablo
Jun 4 2025 at 10:23am
For homeless people in shelters, restrictions on building are a major cause.
It would be interesting to fact-check that claim. What would be a good indicator, or at least a useful proxy, for “restrictions on building”?
Warren Platts
Jun 4 2025 at 1:36pm
Median housing costs?
robc
Jun 4 2025 at 1:39pm
https://Constructioncoverage.com/wp-content/uploads/2015/04/Chart2_Housing-Growth-by-Step.png
robc
Jun 4 2025 at 1:44pm
This linked failed if any mod wants to delete.
robc
Jun 4 2025 at 1:42pm
https://constructioncoverage.com/wp-content/uploads/2025/04/Chart2_Housing-Growth-by-State.png
Scott Sumner
Jun 4 2025 at 3:04pm
Housing prices are certainly one indicator. In addition, California has a number of regulatory restrictions that you don’t see in other states, including CEQA.
Warren Platts
Jun 4 2025 at 4:33pm
I think the problem is a triple-whammy of poverty, high cost of living (especially housing), and lack of social cohesion. For example, Hawaii has the highest median housing of around $800,000. It also has a pretty good per capita income of over $61,000. But when adjusted for cost of living, Hawaii is about as poor as Mississippi. The story is the same in California. Yes, California is the biggest, richest state with a great per capita income, but the poverty rate is 20% — the highest in the nation. Meantime, the minimum wage in Hawaii is twice that of Mississippi. Also, when most people contemplate a move to another state, states like Mississippi or West Virginia are not usually at the top of their list. That entails that just about everyone that’s living in Mississippi or West Virginia has been living there for generations and so have a dense family and friend network that a person could rely on when times get tough.
Craig
Jun 5 2025 at 2:38pm
“That entails that just about everyone that’s living in Mississippi or West Virginia has been living there for generations”
Indeed, I’ve also seen extended adolescence become more intergenerational homesteading. The child et ux will get the starter single wide (that sells new for 40-50k) and they’ll stick it on their family property which in the South often does include family cemetaries.
Grant Gould
Jun 4 2025 at 4:53pm
In the absence of high vacancy rates, it’s hard to see how anything except supply could cause homelessness: If there are 10% more households than houses, approximately the worst-off 10% of households will be homeless, regardless of poverty levels.
Poverty is going to inescapably correlate with homelessness; mental illness, drug use, and so forth likewise; but the underlying hydraulics, not the poverty level, is doing all the work. In a world where nobody was poor, mentally ill, or drug addicted, but the housing supply was 10% short, you’d have 10% homeless.
The argument to the contrary has to be that if people weren’t poor the elasticity of supply would be higher? I’m pretty sure that’s the argument that Mississippi works against, though.
Thomas L Hutcheson
Jun 5 2025 at 7:56am
What exactly is the question?
In place X, which policy would reduce homelessness the most?
If reduction in homelessness is valued at $Z per person year, how does that affect the cost benefit analysis of Policy Y?
“Poverty” is not a policy variable so asking its effects on anything is imprecise.