State Migration Increasingly Driven by Taxes

[dead0man]

but it's from economists from CATO being reported on by Reason, so I'm sure it's all lies

You must be logged in to read this quote.

[Torie]

Well now that I have a Red Avatar, it might be interesting to discuss the policy issue raised thereby and how to address it. Cato-Reason of course is implying that the high tax states should have taxes equal to the low tax states, and life will all be better all across the Fruited Plain.

The Red Avatar might suggest that the real problem is the race to the bottom problem. States that have a suck social safety net, and treat workers like serfs (a little extreme hyperbole to perk me up this morning, and I can cuz I am a Mod and thus special), of course get employers to race down there, because it is a win-win. They can reduce their own taxes, while increasing their profits, thanks to the abundance of serfs. And that can help pay for a good private school education for their kids, so they don't need to worry about suck schools that warehouse the issue of the serfs.

The Red Avatar might suggest that the policy fix is to have the low tax states raise their taxes closer to the high tax states. That way, the race to the bottom game will be slowed down, leaving only Switzerland as a place to flee to. But American employers suck at languages, and English is not a recognized language in Switzerland, so that is one little inconvenience for them anyway. And Switzerland while low tax, doesn't have too many available serfs to boot. And they can't go to Mexico, because they will be kidnapped and held for ransom (and then the damn language thing again).

Thanks for putting this up Dead0.  As always, you are my hero on this Forum.

[Brittain33]

"That new study on Tax Reform and Interstate Migration is from Chris Edwards, a tax expert at the Cato Institute. Using 2016 data from the Internal Revenue Service, he finds that 578,269 people moved, on net, from the 25 highest-tax states to the 25 lowest-tax states."

If California and New York are in the top group, and Texas and Florida are in the bottom group, what % of 578,269 do they comprise? Is it potentially even greater than 100%?

How do they factor out the cost of housing as a factor?

[Torie]

That is a very insightful and pertinent question. Maybe it is more driven by snob zoning (a great evil stalking parts of the fruited plain).

I think the three of us can all agree to launch a jihad against snob zoning, no?

[DINGO Joe]

Well, do they separate retirees from the general population or does this include everyone?  It's a basic fact that the boomers are hitting 65 and exiting the workforce.  It's a fact that boomers are a much larger group than the previous generation.  I've heard people leave NY to go retire in Florida (or maybe elsewhere). 

Also, wrap your mind around this fact, according to the US Census since 2010, the population has increased 17 million.  Of that, the population 65 and over has increased by 10.6 million while the population under 65 has increased by 6.4 million.

[Torie]

Well, do they separate retirees from the general population or does this include everyone?  It's a basic fact that the boomers are hitting 65 and exiting the workforce.  It's a fact that boomers are a much larger group than the previous generation.  I've heard people leave NY to go retire in Florida (or maybe elsewhere).  

Also, wrap your mind around this fact, according to the US Census since 2010, the population has increased 17 million.  Of that, the population 65 and over has increased by 10.6 million while the population under 65 has increased by 6.4 million.

Not that there is much there yet, but here is a thread on that topic. I came up with the one article by googling, and the list of states purportedly getting retirees on a net basis was in part surprising.

[Rules for me, but not for thee]

A factor for sure, but it's the amount of money people make and driving up the cost of living in such places.  People making $100,000+ are ok in Cali, NYC, Chicago, etc.  Low and middle-income families are the ones moving to where the cost of living is much lower (and usually, so are the taxes).

For California, the graphs show this quite well: https://lao.ca.gov/LAOEconTax/Article/Detail/265

Incoming from NY/IL in the high income range, exodus to neighboring states and Texas for lower incomes.

[DINGO Joe]

Well, do they separate retirees from the general population or does this include everyone?  It's a basic fact that the boomers are hitting 65 and exiting the workforce.  It's a fact that boomers are a much larger group than the previous generation.  I've heard people leave NY to go retire in Florida (or maybe elsewhere).  

Also, wrap your mind around this fact, according to the US Census since 2010, the population has increased 17 million.  Of that, the population 65 and over has increased by 10.6 million while the population under 65 has increased by 6.4 million.

Not that there is much there yet, but here is a thread on that topic. I came up with the one article by googling, and the list of states purportedly getting retirees on a net basis was in part surprising.

Yeah, I'd love to get better info since the boomers are such a big group, but I had about as much success googling as you did.  Boomer births crossed 4 million in 1954 and stayed above that level for a decade so the biggest wave is yet to come

https://www.infoplease.com/us/births/live-births-and-birth-rates-year

[Florida Man for Crime]

There doesn't seem to be much actual substance here, and this certainly does not qualify by any reasonable definition as a "study."

This apparently is the full "study":

https://www.cato.org/publications/tax-budget-bulletin/tax-reform-interstate-migration

But as far as I can see all it does is draw a 1 variable scatterplot and run a simple linear regression on it, and then say that 2 variables are correlated.

You must be logged in to read this quote.

And that's all there is, unless I am missing something. Anyone who has ever run any sort of regression higher than the introductory college statistics level can tell you just how little that actually means.

The Reason magazine article claims:

You must be logged in to read this quote.

But that is patently untrue and a gross misrepresentation. Edwards only views and dissects the data in one single, highly simplistic way.

There are many other ways you could view the data and many potential obvious problems with the way Edwards is looking at the data. For example, he uses the ratio of in-migration to out-migration as his variable to supposedly capture how much migration there is. Implicitly, this gives all states equal weight regardless of the actual size of the in-migration/out-migration, and it also introduces a lot of noise into the analysis, especially from small states that have relatively small amounts of in-migration/out-migration, because an increase in 100 people in-migration or an increase in 100 people out-migrating could substantially change the ratio (whereas in a large state with higher absolute numbers, a similarly sized change would have much less impact on the ratio). But one could alternatively use the absolute amount of in-migration/out-migration as your variable. Why not do that? Perhaps it gives a different result.

But we wouldn't know that, because he never does that (more likely, however, he made many simple correlation graphs and picked out the one he liked the most to highlight in this "study"). There are many other such ways to re-organize and re-analyze any sort of data like this, especially when one is dealing with broad, large scale sociological/demographic macro-variables. Typically (really, essentially always), the results you get are extremely sensitive to slight changes in how you do these things. But the "study" says nothing about any sort of checks on the robustness of the analysis.

There is also such a thing as "effect size" - even if the object of your study is found to be "statistically significant," that does not mean that it is actually important in actually explaining a phenomena - statistical significance is an entirely different concept from the size of the impact. A fly could land on my shoulder and that might be statistically significant to a very scientific study of my day. But nonetheless, it doesn't have a significant impact on my day.



The other part of the article that dead0man cuts out to emphasize is a different "report":

You must be logged in to read this quote.

Possibly that might say something more substantive, since it claims to control for "for climate, the cost of living, and other factors." But judging by the quality of the first "study," it doesn't inspire much confidence.

I might look at that as well, but at some point there get to be too many "studies" to look at them all. When an organization presents pseudoscience in their first "study," then given that one has limited time to read all the studies/reports put out by every possible organization, I am not sure if it is worthwhile to spend time seriously looking at their second report to see whether it is actually a report, or just another "report" similar to their "studies."

Still, if dead0man wants to go into that other report and pick out one particular claim and point to one particular argument/analysis that is supposed to support it, then perhaps we could take a look at that.



None of this says anything one way or the other about the underlying issue - just that anyone who takes this as a serious piece of academic research that establishes anything whatsoever is pretty gullible.