Hacker News new | comments | show | ask | jobs | submit login
Where star scientists choose to locate: the impact of US state taxes (microeconomicinsights.org)
96 points by mcenedella on Nov 12, 2017 | hide | past | web | favorite | 71 comments



It is interesting that this is a summary of a published journal article and yet suffers from a rather glaring flaw.

The paper looks at corporate and income tax but completely ignores sales tax, so it treats Oregon as being the highest tax state (with a 9.1% income tax rate, 0 sales tax) while a state like Texas is treated as a 0% income tax state when it has an 8.25% sales tax rate for major urban areas.

Looking at their scatter plots the correlations they present look weak (the summary doesn't provide the numbers) and there is no credible theory for causation.

"Star Scientists" (defined as 99th percentile income earners, 300K+) don't make location decisions based on taxes, they make decisions based on who has the infrastructure to support their work and money to pay them.

Corporations might make decisions based on tax policy and they might have to take employee income taxes into account when creating a competitive compensation package for a high value employee, but one or two percentage points on the salary of a few high end employees is not going to move the needle on the overall corporate budget.

Obviously taxes are part of the big picture for corporations deciding where to locate but they are probably not a big part. A state advertising low taxes may be signalling that they are "business friendly" but they are also signalling that they have an overall poor business environment and are desperate to get people in the door.

Market theory suggests that taxes will adjust to an efficient level which is the maximum that individuals and corporations are willing to pay to locate somewhere. Consequently, you can judge that high tax states like California offer the best environment for business because the market is willing to pay their high taxes in order to locate there, while a hypothetical no-tax locale would be the worst environment for business because even a small tax would cause residents to leave.


> "Star Scientists" (defined as 99th percentile income earners, 300K+)

No! The definition they use is "95th percentile of patents awarded", then they assume that these people are probably in the 99th income percentile. But they present no data to validate this assumption (at least in this summary of the journal article).

Personally, I'm not convinced that the assumption holds.

Edit: Also, the figure of 300K+ doesn't appear in the post, and a very quick poke around Wikipedia didn't confirm that number for the 99th income percentile. Do you have a source?


Finding an objective definition for "star scientists" isn't easy, but the one they've chosen does have the deficiency of being more likely to identify patent trolls than most Nobel Prize winners. Suspect the former group's choice of domicile is much more closely correlated with tax advantage than the latter's.


Yeah, the methodology is weird.

I got the income info from here: https://seekingalpha.com/article/4109731-united-states-incom...


Thanks! I do note this only refers to "workers" (they give an exact definition), so does not include people who only have unearned income. I don't know how much that would distort the data at the top.


> The paper looks at corporate and income tax but completely ignores sales tax, so it treats Oregon as being the highest tax state (with a 9.1% income tax rate, 0 sales tax) while a state like Texas is treated as a 0% income tax state when it has an 8.25% sales tax rate for major urban areas.

Point taken, but, there's a practical limit to how much a high earner can buy and thereby pay sales tax, but there's relatively less limit to how much a high earner can earn and thereby pay income tax. Kind of like shorting stock.

To say it another way, you can choose to limit your purchases, but you're highly unlikely to forgo income.

Orthoganally, both taxes greatly impact low earners, since most or all such wages must be spent. Sales tax impacts high earners much less than income tax.


Some of these “low tax” states bury costs in other areas. Things like sales tax on food, user fees for basic services like fire protection, poor regulatory environment for things like ehalth care, etc.


My favorite is Texas with no income tax and a.)chronically underfunded public schools with the situation getting worse as tea party influence on state legislature increases over the years b.) maternity outcomes worse than almost every other country on earth.


You need to know a little more about Texas schools first.

1. They're independent, which means that they have their own taxing authority. Citizens in the ISD (Independent School District) get a bill yearly to pay the tax. So if you buy land here, you need to know what school, emergency services, water, etc. district you'll be in to get an idea of what your taxes will be (commonly in the 2-3% of your property value every year).

2. Because many ISDs are in low-population areas, or areas with low-value land (since the taxes are based on property values), ISDs in wealthy areas have what is called recapture. Which means that funds are taken from them and redistributed to the rest of the state. So people owning property in Austin, Houston, Dallas, etc. are funding schools in the rest of the state.

3. The tea party influence (deep conservatives) in the legislature is real. They do support better schools - only they support some subjects more than others. And some not at all.. But they only meet every 2 years so their influence is limited on the ISDs.


Everything you said is true except the last part has an omission.

The state helps fund local school districts, and has been doing several things to reduce that amount every session ever since White was governor if my memory serves. For just one example there was an expose in the Houston Chronicle how the Texas Education Agency imposed a set percent limit on special education students so no matter what, if a school reported a percentage above that, the school was penalized heavily so school adminstrators played the game, and never found more special ed students than the TEA limit. The budgetary reason for this is special ed students cost a lot more per student.

There was a public radio expose but the story was in print first. http://www.houstonchronicle.com/denied/


The only way to compare high vs low taxes is government spending / GDP. Anything else is just comparing how things are taxed not how much people pay.


One should also compare effective tax rate on overall wealth. Comparing the top 1%, 5%, and 10% to the bottom 50% would be interesting. Low income tax states generally get a higher proportion of their revenue from regressive taxes and fees.


> Looking at their scatter plots the correlations they present look weak

TU;DR (too uncorrelated, didn't read). When I scrolled down and saw those poor regressions, I immediately decided to close the page.


> "The paper looks at corporate and income tax but completely ignores sales tax..."

Good point. But I would take it a step further. It's all taxes. For example, tax on energy, gasoline, etc., or property taxes.

The state and local govs require revenue. To think these entities exist but don't require revenue is simplistic, if not naive. Regardless of what form it takes, a tax is a tax is a tax.

For the sake of transparency, simplicity and efficiency (i.e., each tax has some infrastructure to implement it, police it, etc.) we'd be best served with a single trackable tax. As it is, we get nickled and dimed without realizing it. What's worse we (thr people) can't "manage" and can't hold accountable that which we can't identify.


Oregon is an outlier, but with other states: TX vs. CA, you have no state income tax but you do have a sales tax while California has both. Property taxes also matter; but CA property taxes are also higher.

Really what needs to be done is a measure of effective tax rates.


Are CA property taxes higher than Texas? This article says the effective tax rate in Texas was almost double that of California’s two years ago:https://taxfoundation.org/how-high-are-property-taxes-your-s...


An interesting thought though.. if a 3 bedroom house in Cupertino costs $1.7 million and an equivalent house costs $250k in a Houston suburb. Then the property tax in Cupertino will be vastly higher for the same amount of space. While that higher housing cost is offset somewhat but a higher salary, that higher salary also lands you with a higher income tax bracket and overall tax bill, which means at the end of the day, your standard of living, your disposable income is vastly lower in the Bay Area vs. Houston. You could argue that Cupertino has better schools on average, but then again for the income differential, you could afford Houston’s best private schools and still come out ahead financially. The only objective reason to live in SV is because you are required too — you work at Apple, Google, etc. If you are a startup locating to locate somewhere, it’s a fallacy that you have to be in the Valley. If your business is worth funding, being near Sand Hill is unnecessary.

The ‘money’ is located in the Valley for sure, but remember those guys writing the checks aren’t paying the same kind of taxes the employees are. While the people that actually have to work at your startup are just scraping by despite the high salary. I ran the numbers and it would be cheaper for me to live 2 hours from Cupertino, buy a Cessna and commute via 20 minute flight each way than to live within 20 minutes drive — that cost difference includes buying and maintaining the airplane, hangar costs as well as a cheap car kept at te Reid airport in San Jose. Even having to drive occasionally due to weather makes that idea still more reasonable.

A standard engineer salary in the Valley isn’t enough to buy a house unless you win the equity lottery or save for 10 years for a down payment. Anywhere where housing is 40-50% of salary plus high taxes is a recipe for disaster — regardless of how many electric car charging stations or trains their might be.

Obviously this is a tangential issue to taxes, however it’s still a vital part of the calculus of being in SV.

I just wish I could commute to Nevada as that would be a huge savings, but a bit of a stretch for a twice daily Cessna flight.


California limits property tax annual appreciation however. Many of these "star scientists" in CA might have bought 30 years ago and have a property tax assessment on their $1.7M house of $250k.


From talking to coworkers there is such limitation but only to the point that it can only go up 2% per year. I am guessing there is no other type of cap? In Texas such cap exists for people 65 and older where their tax rates are frozen in perpetuity. Does such cap exist in CA?

So putting aside inflation and the like, assuming value held constant wouldn’t taxation property values “catch up” after many years? Or are there other factors at play besides the 2% limit?


Look up Prop 13 and Prop 8, it's essentially rent control for property owners. Property tax cannot exceed 1% of the assessed value, and the assessed value cannot increase more than 2% a year. Prop 8 ensures that property taxes go down when the assessed values decline.

The caps stay in place until the property changes hands. This applies to all property in California. Commercial? No problem. Second home? Same discounts. There's been talk about either abolishing or reforming Prop 13 via a split roll system, but it's a political third rail.


It’s common to see large disparities; from looking at a random street (Betlin Ave, Cupertino) there is a house with an assessed value of $67k while Zillow estimates its value at $1.7M. Assuming that inflation is ≥2% per year yet property values are held constant, it would take 164 years for the assessed value to “catch up” to market value. And the tax discount is not just for the elderly. Children inherit their parents’ low assessment. The tax only jumps up beyond 2% in a year when the owner renovates or sells the house.

The tax rate is around 1.25% (limited by Proposition 13 to 1%, plus slightly more for infrastructure bonds). The low tax rate contributes to higher market prices. Proposition 13 is a very good system for property investors.


Thanks for this. The deeper insight here helps to add more context to the story of why the property values are so much higher in California than other places.


California is one of the lower property tax states by rate.

https://taxfoundation.org/how-high-are-property-taxes-your-s...

Texas is higher, and Illinois + many north eastern states are much much higher.


I hear second hand from friends - tales of people buying houses in Texas (it's so cheap for so much!) then bailing after a couple of years because of the property tax burden. Even people in medicine overestimate what they can afford in property tax.


Income and Corporate Tax tax (and disincentivize) production, Sales Tax taxes consumption.

IMO the latter is preferable because of the above plus it doesn't accrue.


Sales tax is like inverting the income tax brackets.

Suppose the rich can buy everything they need with 1% of their income, sales tax is 10%. Then, the rich are paying 0.1% income tax. If the poor have to spend 20% of their income on the same necessities, they are paying twenty times more tax per amount of income.


Both things are true. Income taxes are less efficient, sales taxes are more regressive.


Possibly the most efficient and progressive tax is a land value tax: https://en.wikipedia.org/wiki/Land_value_tax


I’m not sure what your point is. That wealthy people would spend a lower percentage of their total income on necessities, and thus a lower tax per income dollar, sounds like a truism.


This argument assumes the the rich only spend 1% of their income.


Sales tax also puts a larger unavoidable burden on the poor, which I suppose is preferable to some. Then consumer spending among groups who have no room to save leads to them buying cheaper products, increasing numbers of which are cheap or shoddy alternatives, and in the case of non-food items, cheap imports.


In many cases, like Texas, necessities like food, medicine, baby products, etc are exempt from sales tax to reduce this burden.


Income tax may disincentivize earning at the margin but also incentivizes investment because businesses can deduct expenses. In any given tax period a business will have a strong incentive to re-invest earnings which is the easiest way to avoid paying income tax.

From a theoretical perspective it seems to me that income tax would favor economic growth while sales tax, which can't be avoided, would depress economic growth.


> but also incentivizes investment

No, it just doesn't disincentivize investment, same as sales tax.


Let's say I am in the last quarter of a tax period where I am going to have to pay taxes on a million in revenue.

Now let's say that I am planning to invest in some new equipment in the next 6 months.

If I buy the equipment now I don't have to pay taxes. If I wait I do.

How is that not an incentive to invest?

Every dollar in business expense is discounted by the amount of income tax not paid. The higher the income tax the greater the discount on business expenses and the greater the incentive to do something with cash other that hoarding it.


> If I buy the equipment now I don't have to pay taxes. If I wait I do.

If you wait until the next year, you'll don't have to pay the taxes next year instead of not paying them this year.

But my point is about income vs sales tax:

- 0 % sales tax, 10% income tax:

You don't pay taxes on what you invest (assuming deduction)

- 10 % sales tax, 0% income tax:

You don't pay taxes on what you invest (assuming deduction)

Result: You cannot claim that people are investing more just because you have income tax.

You could of course assume that putting a tax on everything except investment will encourage investment. But it's the same as if you raise the price of everything except water to encourage the consumption of water.

If you ask why this wouldn't work, think about how taxes affect disposable income.


Texas also has massive real estate taxes which would impact high earners more than sales taxes. Colorado has a 4.6% income tax but I pay 30% of the rate that my Texas colleagues do in taxes on my house.


Do you have any citations or figures to back up your claims?


>The paper looks at corporate and income tax but completely ignores sales tax,

Say I make $150k/yr.

Income tax will be taxed on almost all of it.

Sales tax will be taxed on what I spend. I'm pretty sure I'll spend a lot less than $150k/yr.

The effect of income tax for higher income folks vastly exceeds the effect of sales tax.


Your assessment is flawed. Because income taxes are generally progressive, you will have a lower rate applied to a portion of your income and higher rates towards the top end. Effectively paying a lower absolute rate. But you would pay the full sales tax on all your purchase, leading to a similar net effect.


> Because income taxes are generally progressive

Except that assumption is wrong. State taxes are typically not progressive, even though they should be. The lack of progressive brackets at the state level is one of the sources of biggest economic injustice in our tax system in this country.


https://taxfoundation.org/state-individual-income-tax-rates-...

Forty-three states levy individual income taxes. Forty-one tax wage and salary income, while two states—New Hampshire and Tennessee—exclusively tax dividend and interest income. Seven states levy no income tax at all.

Of those states taxing wages, eight have single-rate tax structures, with one rate applying to all taxable income. Conversely, 33 states levy graduated-rate income taxes, with the number of brackets varying widely by state. California and Missouri each have ten brackets, the most in the country.

Some states have absurdly narrow banding in their brackets, "Missouri taxpayers reach the state’s tenth and highest bracket at $9,072 in annual income". Alabama's highest bracket starts at $3,000 for single filers.


Thanks for adding those details. Some brackets are absurd not only in the top being very low income level but the difference in rates being negligible.

Oregon has 4 brackets but they are 5%, 7%, 9%, and 9.9%. But the 9% kicks in at $8400, so even those at poverty-level incomes are paying that high. The fact that the next bracket starts at $125,000 would be notable if the jump weren't so pathetic. But the extra less than 1%? Come on… absurd. This does NOT count as progressive taxation by any real measure.


>Your assessment is flawed. Because income taxes are generally progressive, you will have a lower rate applied to a portion of your income and higher rates towards the top end.

My assessment takes all of that into account. I live in one of the higher income tax states, and one of the lower sales tax rates. I have real numbers I can work with.

I pay over $6K/year in state income tax, and my income is nowhere near $150K.

Do the math: At a 10% sales tax, how much do I need to spend to match that? If I take all of my annual spending per year, a 10% sales tax would be less than $6K. And keep in mind we don't pay sales tax for housing and services. At 10%, my sales tax would likely be under $3K. Perhaps even under $2K.

>But you would pay the full sales tax on all your purchase, leading to a similar net effect.

Not true in many states. In one state I lived in the sales tax on groceries was 1%, and over 6% on everything else, for example. I believe I encountered a state with 0% sales tax on groceries, but high on everything else.


This.


Wow, the argument of the text is almost completely disconnected from the title: it's ostensibly about where star scientists choose to locate, but in reality talks only about where star scientists choose to migrate. These are very different things.

If a star scientist chooses to remain where they began being a star scientist, that itself is a decision. And it's an obviously important one: a high-tax state may well be able to offset receiving migrating scientists if it creates more in the first place.

The account for this, you'd want to look at where star scientists start out, or simply look at per capita star scientists for each state. To ignore this strongly biases the article in favor of low-tax states.

I suspect that in practice, high-tax states tend to have more star scientists per capita than low-tax ones, simply because part of what they've done with greater tax revenue is invest into higher education that helps create those scientists. This would be not unlike how so many tech companies are headquartered in high cost metro areas, because those areas are more successful at creating those businesses, not at convincing businesses started elsewhere to move there.


The conclusions seem explicitly misaligned with the title. The title implies agency and choice, the conclusions don't come anywhere near making that inference

Title: Where star scientists choose to locate: the impact of US state taxes Conclusion: Overall, we conclude that state taxes have a significant effect on the location of star scientists.

There is absolutely no causal link established in their data. They simply hypothesize one and then use it as a title.

Edit: Its also a summary (by the authors) of an actual research article that on a brief read is much more circumspect but still commits many of the same logical sins https://www.aeaweb.org/articles?id=10.1257/aer.20150508


This article leaves more questions than it answers.

Is top 5% of patents really a good measure of scientist productivity as opposed to published articles or research awards?

Is outmigration more relevant than absolute residency numbers? If "top scientists" choose to start their career in a high tax state and don't move, they will not show up in this study.

What percent of this rather arbitrary group are professors, working at large corporations, or some other affiliation?

What is the effect of outliers? No sample size is given and the individual tax rate effect doesn't look very robust in the scatterplots. Does that fact that a large amount of patent law gets set in Eastern Texas influence this more than taxes?

Even for the corporate tax rate effect that does look like a positive correlation, the effect could be due to one or a few large organizations moving their R&D department.

This just seems like some half baked regressions that aren't seriously trying to understand the issue.


> Is top 5% of patents really a good measure of scientist productivity as opposed to published articles or research awards?

In my field, one can have a long, storied, and impactful career that ends up with serious business national honors and never file a patent.


Ha! Top scientists choose to locate where they can get positions. If you've got some seriously fancy credentials, the Ivy Leagues come calling with big salaries, but you don't get to choose what tax jurisdiction that's in.


> The Ivy Leagues come calling with big salaries, but you don't get to choose what tax jurisdiction that's in.

If multiple Ivy League schools come calling, this does offer you the opportunity to pick the one with the most beneficial tax arrangement.


They ought to say "prolific inventors" rather than "star scientists." Science is only tangentially related to invention.

Also, a lot of star inventors deliberately choose not to patent their work. Focusing on patents ignores people like Tim Berners-Lee whose work would have pretty much been worthless if he had patented it.


I need no convincing that decisions get made based on tax rates.

However, I call "correlation is not causation" on this particular article. It is highly possible that tax increases and migration are both linked to some third variable ('local conditions') and the /Testing the validity of the results/ section of the article does not convince me that this possibility has been ruled out. Establishing why there is a link here requires more information.

I have observed that governments hate raising taxes because it upsets voters. In practice, resistance to raising debt takes a back seat to upsetting voters. If a government is raising taxes, in my experience, the economy is struggling and making up the government budget with debt isn't an option (for whatever reason). It is possible that rate of tax follows a similar principle where prosperous regions need a lower rate to achieve a better result.


The authors note that shifts in migration patterns follow, rather than precede, shifts in top-rate (but not middle-rate) taxation.


Yeah, but that isn't actually very strong causative evidence. If there is a confounding variable, maybe upending a scientists life and moving states takes longer than than the legislature raising the tax rate. Maybe the tax change compounds the impact of another variable and is followed by their relocation (ie, the tax change contributes to the decision, but isn't the root cause).

The point is, "A causes B" is a much stronger statement than "A and B are related". There is solid evidence of a relationship, because the core of their study is a statistical test of correlation. The causative elements here are of a speculative nature, because they didn't gather any data on cause. Cause is being extrapolated from a correlation, which is dangerous thinking. The conclusion here is a good one to be circumspect about.


Measuring the quality of scientist by how much money they make means that there are no star scientists a universities doing research that's publicly funded. (and making less while they do it.)

Libertarian studies be like: "Money makers making money want more money not less. That's just science. You can't argue with science."


Do those red lines actually correlate to the dots or did they just collect noisy data and draw totally random lines? Analyzing the "Destination state" and "Dest origin differential" it would appear that the drawn in red lines are actually facing the wrong way. In the "Destination state" graph the trend is strongly opposite to the line.


The third chart on each line shows the relationship between tax level and net-migration which what they are making their claims about.

The individual one (first line) looks like a shotgun blast to me but the business one (second line) shows some correlation.


If you trim the two outliers on the top right plot, it becomes a spherical blob: http://i2.wp.com/microeconomicinsights.org/wp-content/upload...

microeconomics.org is owned by the "Institute for Fiscal Studies," a British think-tank whose funders I haven't bothered to find. If you find the funders, I suspect you can predict the red lines.


if I had a student turn this in I would write them a note telling them its time to come in from their correlation safari.


Why are they only talking about income tax? Property tax is also a major factor.

If you had two job offers 10 years ago: either CA (good area) or TX (Austin) and got a mortgage. 10 years later Prop 13 would have severely limited the property tax growth of your house while property taxes on a house in TX would have ballooned.

Your taxes in TX would have gone up (and can go down) regardless of how much you make. You can max out your 401K to further reduce your taxes in CA while you really have no remedy in TX. CA encourages savers. TX often has to give tax breaks (property and sales) to large companies to locate there. Small businesses (grads starting tech startups that hire more grads) are left to pay the very regressive taxes that hurt them in regressive states like TX.

I would also argue that research clusters (both universities and companies who hire their grads) matter far more for grad job seekers than tax.

I really question if this article has any real value.


Conclusion: top 5% patent holders (== "star scientists" according to the authors) in the private sector can be tempted to relocate from higher tax jurisdictions to lower tax jurisdictions. It seems that a sustained multi-year tax differential helps increase migration, because the subject people don't respond right way.


At the top end, these differences in state income tax make a big difference. For instance, in a salary-capped league like the NBA, the difference between signing a max contract with a Texas or Florida team (where there are no income taxes), and teams in California, New York, or Toronto (to name some of the places that take the biggest cuts) is really significant.

Even for us plebs, it can be a hefty chunk of change. Staying at the same job, I moved across the border from Maine to New Hampshire, and instantly started netting another $500 or so dollars a month.

http://www.bankrate.com/finance/taxes/taxes-cost-professiona...


Did your job change locations with you? If not, do ME and NH have a tax reciprocity agreement? If not, you may be committing tax evasion.


I always worked in NH, which has no wage income tax, but if you live in Maine, you have to pay Maine income tax, whether or not that income was earned in Maine. At least that is how the accountants have explained it to me.


Ah, yeah, I read it as moving from Maine to NH to avoid paying taxes on a job in Maine.


People don't make decisions in a vacuum. Scientists also care about their children learning STEM, so they will want to live in places with good schools. It's partially why they stay in Minnesota over Wisconsin, Massachusetts over New Hampshire, and so on.


If you're almost able to afford private school the extra money from not being taxed into oblivion could allow you to justify the cost/benefit of private school.


There are so many things wrong with this analysis that I wouldn't know where to begin, except to say that I'm very disappointed to see this on the front page of HN. Looking forward to the meta-analysis of papers addressing the number of angels we can fit on the head of a pin.


yeah this is bad..bad bad bad...science


Having experienced the weakness of the scientific institutions in several low tax states, I wouldn't expect this effect to be very strong.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: