Friday, December 28, 2007

Higher State Taxes = Lower Population Growth

The chart above shows graphically the negative relationship between state population growth using Census data available from the WSJ here, and state tax burden data available from The Tax Foundation here, both for 2007.

The OLS regression results above further verify that the negative relationship between state population growth and state tax burden is statistically significant at the 1% level (highest level generally reported).

Implication of the OLS Results: Based on the negative regression coefficient of -0.232, we could say that as the state tax burden increases by 1%, population growth decreases by about .23%. Alternatively, we could also say that for every one percent decrease in state tax burden, state population growth would increase by .23%. (For an overview of OLS/linear regression, go here.)

Bottom Line: States with lower tax burdens attract more businesses, workers and people, and states with higher tax burdens lose businesses, workers and people. In other words, these results confirm the theory that if you tax something, you'll get less of it.

Thanks to Ben Cunningham for supplying the data. See First Trust Portfolio's similar analysis "Voting With Our Feet"
here using the same data.


At 12/28/2007 7:02 PM, Anonymous Anonymous said...

Is this relationship caused by migration, or can natality be affected by the tax rate as well?

At 12/28/2007 7:35 PM, Blogger thomasblair said...

Visually, the fit is awful - not everything varies linearly. In this case, at first glance (and without the raw data), a normal distribution appears accurate.

At 12/28/2007 9:11 PM, Anonymous Anonymous said...

I don't know about regression coefficients or any of that stuff, but I do know that my area of upstate NY is losing population. The younger people are leaving for greener pastures. Three of my four children left. Could that have anything to do with New York being the second highest in the lists?

I stay in New York because my business is here. When I retire in a few years, it won't be in New York.

At 12/29/2007 12:38 PM, Anonymous Anonymous said...


A couple years ago I had found a similar statistical correlation between net migration between countries and their level of economic repression. Same result (see here)

At 12/29/2007 2:06 PM, Anonymous Anonymous said...

The limitation of the analysis is that it only looks at only one economic factor, state taxation in relationship to migration. Upstate NY is also less attractive than Florida due to the weather. One might also consider economic opportunity as most of us would agree that a vibrant economy is more likely to attract residents.

There are many quality of life issues that people consider when relocating such as the quality of health care, education and drinking water. Many of these considerations cannot be quantified by an economist, anymore than an accountant considers opportunity costs when preparing financial statements but they are no less real.

On the other hand, many state economies that were dominated by manufacturing and primary industries have suffered in recent years. The reason that upstate NY, the rust belt and New Hampshire don't attract new residents also reflects their lack of strong diverse economies offering high paying jobs.

It is very tempting to want to find a simple answer to the problems of dwindling population and economic stagnation (ie. Low state taxes = higher population growth). Unfortunately, there is no simple recipe for making a viable economy.

Ed Glaeser did a good piece on Buffalo recently. It does help to put a some of the challenges into perspective.

At 12/29/2007 2:16 PM, Anonymous Anonymous said...

Clicking on logicalopinion's name takes you to the appropriate link, the "here" link brings you back to this comments page.

Nice posts on your blog, LO.


At 12/30/2007 10:24 AM, Anonymous Anonymous said...

This seems to be only income tax burden. What about the effects of adding sales tax and fees (phone, car tags, etc)?

At 6/30/2008 2:27 PM, Anonymous Anonymous said...


But you do realize that if you set the intercept at zero and run the regression you get the opposite slope.

At 3/14/2009 9:04 AM, Anonymous Anonymous said...

You state this correlation as a causation high taxes ==> low growth, but is this true? With emigration, fewer taxpayers remain to fund the inelastic public base, and thus taxes might have to rise. I think if your idea is correct, and you do indeed offer some logic, then you need to tighten your logic with some clever IV.

Lones Smith
Professor of Economics
University of Michigan


Post a Comment

<< Home