A Re-Examination of the State Comparison Matrix

by Jason S.

In August I wrote an "Analysis of the State Comparison Matrix" in which I argued that no matter how you weight the variables (within reason), four states continually appeared at the top: Alaska, Wyoming, Delaware, and New Hampshire. Tim Condon's article, "Our Most Important Decision", challenges this analysis and argues that North Dakota is the ideal candidate state for the Free State Project. George Cunningham's "Choosing Which State to Liberate" and Tennyson's "Analyzing the Freedom Orientation of Existing State Populations" also present food for thought, both coming down more or less on the side of Wyoming.

Why do the results differ? Of the four states at the top of my old list, Tim eliminated Alaska, Delaware, and New Hampshire outright, the former because it is isolated and the latter two due to voting population. There is one more important difference: he ignored quality-of-life variables (jobs, crime, income, livability) and de-emphasized federal dependence.

I actually agree with Tim's argument that quality of life is far less important than probability of success. (But if quality of life is not very important, why eliminate Alaska simply because it is cold and isolated? Perhaps because it might be difficult to get 20,000 people to commit to moving there.) I also agree that he has made a strong case for North Dakota, a hitherto unjustly ignored state. However, I strongly disagree with the de-emphasis of federal dependence, and I think that Tim's analysis would have been more rigorous had he used the quantitative tools available. In addition, he eliminates states outright simply because of voting population, even though other factors might make up for a poor showing on these variables. Are Delaware and South Dakota really totally nonviable candidates simply because they had 25-40,000 more votes cast in the 2000 presidential election than Vermont and North Dakota? Surely there might be factors that could overcome this gap: even for states higher on voting population, like Montana, Idaho, New Hampshire, and Maine. We've already eliminated states more populous than these because they are not doable.

The purpose of the following analysis is not to argue for a particular state; rather, it is to use the new data available and new techniques to make the case that what state you think is best depends on how you look at the data. If this analysis succeeds in its purpose, henceforth debates about the state the FSP should choose will be driven by certain regular arguments and considerations. What I mean by this will become clear.

For this analysis I've used the updated state comparisons spreadsheet (Mac-friendly version here), which has recently been updated with new data and new variables. The spreadsheet works by allowing you to assign positive numbers to each variable to indicate a sense of relative importance of each variable. States' scores have been "normalized" from zero to 10, so that the worst state on a particular measure is given 0, while the best state is given a 10, and the others are ranged between the two.

I agree with Tim that voting population is critical, so I've given the number of voters variable a "14" weighting. Campaign finance is given a "6" weighting. I've gone ahead and given total populationa "3" weighting since non-voting population represents future or potential voters. These variables make up the "Size" category, which now has an overall weighting of 23. The Viability category is next. Geography is treated differently in the spreadsheet from the [long gone] Javascript state ranking page. I have scored Maine, Delaware, and Alaska a "10" on geography because of their coastline, New Hampshire a "7.5" because of its small coastline and border, North Dakota and Montana a "5" because of their Canadian borders, Vermont and Idaho a "2.5" because of their small Canadian borders, and Wyoming and South Dakota "0" because they are landlocked within the U.S. You are free to fiddle around with these scores and see how the rankings change as a result: they don't really change much. I've given the geography variable a "5" weighting to start with since it probably does matter, but is not terribly important. Federal dependence I have given a "10" weighting because it is very important. My dissertation research shows that autonomist parties are consistently more powerful in regions that "lose out" economically from centralization. The federal government might be less willing to give autonomy to states that lose out, because they lose revenue, but ultimately they have no choice. They cannot stonewall our efforts forever. Furthermore, critics of the Free State Project in the mainstream media have already pointed out that, for example, Montana is heavily dependent on the federal government, and that an autonomy movement would make little sense there. I think they're wrong that an autonomy movement would make no sense in Montana, but the federal dependence issue will be difficult to overcome. The final viability variable is federal land, the assumption being that states with more federal land have a legitimate grievance against the federal government. I've given this variable a "3" weighting, for a total Viability weighting of 18.

The other two categories are "Culture" and "Quality of Life." For now I will give zero rankings to all the variables in these two categories, because political culture can be changed, and all the states we're considering are not too far apart, in the big picture, on political culture, and because quality of life can be sacrificed for, and later improved by, freedom.

The weighting that results from all this is as follows:

• Wyoming - 313.78
• Delaware - 302.77
• Alaska - 299.1
• Vermont - 282.02
• North Dakota - 226.40
• Idaho - 195.50
• Montana - 185.37
• New Hampshire - 170.92
• South Dakota - 169.05
• Maine - 133.2

This is very different from the original conclusions: Vermont has risen substantially, and New Hampshire has fallen substantially. So if you believe that number of voters, federal dependence, campaign spending, geography, federal land, and total population, in that order, are the only important variables, then you should favor Alaska, Vermont, Wyoming, and Delaware as your favorite four states. Which of these four states you favor can change depending on how you calibrate the four variables. If you think geography is not important, Wyoming moves well ahead of everyone, with Vermont, Delaware, and Alaska grouped together.

But surely this is too simple. Quality of life was glibly dismissed as something that can be forsaken. But jobs are very important: if our 20,000 cannot obtain jobs, they will not be able to undertake political activism, and they will eventually move out. Consider the jobs figures on the State Data page. Wyoming is expected to create just 27,000 jobs over the next 10 years - and the figure is probably lower now, because that prediction was created in 2000 before the recession. If we assume that job creation will be half that in the five years when we will be moving in, that means there will only be 13,500 jobs for our number, and we'll be competing with local job market entrants for all those jobs. Of course, we'll also be bringing some jobs with us, but probably not more than 7,000. There are also out-of-state jobs. But consider these figures from the Census: 232,682 of Wyoming's 239,809 working residents worked in-state. If that ratio continues to hold true over the next ten years, then in our period of migration, only about 400 of us will be able to get jobs outside Wyoming! The figures are little better for Alaska, North Dakota, and Vermont. Alaska will create about 24,000 jobs in our five-year period of migration, and almost nothing out of state. North Dakota will create about 17,000 jobs (optimistic figures) and only about 700 out of state, using the same Census-based calculations. Vermont will create about 17,000 jobs in-state and about 1,200 out-of-state using Census-based calculations. None of these states will be able to support our numbers if 20,000 people end up moving.

So that means Alaska, Wyoming, North Dakota, and Vermont have to be eliminated from consideration, leaving just Delaware out of the above top five. Delaware will likely create 31,000 jobs during our migration and about 3,700 jobs out-of-state using Census figures (40,000 Delaware residents currently work out-of-state). That's probably just enough for us.

Of course, many people would eliminate Delaware for a combination of lifestyle and subjective reasons; it's not a very romantic destination, and to a lot of people it just seems too close to the Washington, DC political establishment for comfort. If Delaware is eliminated, things get thrown up in the air. Idaho, New Hampshire, Montana, South Dakota, and Maine are all pretty close to each other. Of these, South Dakota is far and away the smallest.

What if we include political culture factors as well? Maybe they do not make a big difference, but they can be part of creating a "myth of separateness" that may put us in good stead in the future. So to the above weightings, I have given a 3 to (lack of) violent crime, 2 to (lack of) state & local government spending, 2 to (lack of) state and local taxation, 1 to conservative presidential vote in 2000, 1 to (lack of) gun control, 1 to homeschooling freedom, 1 to (lack of) residents born in state, 3 to lower city urbanization, 1 to lower clustered urbanization, 2 to lower NEA presence, 1 to lower government employment, 1 to more economic freedom, and 1 to less land planning, making for a total "Culture" weighting of 18, equal to Viability, and a total "Quality" weighting of 3 (crime is considered a Quality factor in the dataset). A word on political culture is in order here. Tennyson's analysis looked at political culture through the prism of "big-government" and "small-government" votes, assuming that votes for Bush in 2000 were small-government votes. But this is hardly the only, or even the best, way to look at political culture. In statistics we learn to get at an underlying concept it is best to use several different measures: put together they will all converge on what you want to get at. So New Hampshire, for example, has the smallest state government in the country and extensive gun freedom, but it nearly went for Gore. Apparently not all of those Gore voters, strange as it may seem, were partisans of big government.

So the above weightings yield the following ranking:

• Wyoming - 439.59
• Vermont - 393.20
• Delaware - 393.11
• Alaska - 388.30
• North Dakota - 348.39
• Idaho - 334.40
• New Hampshire - 327.67
• South Dakota - 320.36
• Montana - 293.05
• Maine - 233.10

Wyoming, Delaware, Vermont, and Alaska are again clearly the top four, with Wyoming putting a little space between itself and the opposition. If all these plus North Dakota are eliminated for lack of jobs & non-quantifiable reasons, then New Hampshire, Idaho, and South Dakota are fairly close to each other.

So what about intangible, non-quantifiable factors? Ben Irvin argues that lived freedom is greatest in Montana, and I am persuaded. So let's add "lived freedom" as another variable in which Montana scores 10 and all the other states 0. Then let's give it a "2" weighting, like some of the other political-cultural variables: taxes, spending, and NEA presence. That adds 20 points to Montana's score. Montana enters the race with South Dakota, New Hampshire, and Idaho.

Now let's stir in some quality of life variables, while keeping them at a low weighting. I'll add 1 to livability, 1 to income, 3 to jobs, and 3 to amount of private land. The total Quality weighting is now 11, still a lot less than all the others. Things don't change too much.

• Wyoming - 463.33
• Alaska - 411.50
• Delaware - 411.15
• Vermont - 405.71
• North Dakota - 376.73
• Idaho - 376.50
• New Hampshire - 366.30
• South Dakota - 368.51
• Montana - 337.92
• Maine - 257.00

Now let's say you didn't buy my arguments about jobs or federal dependence at all. In fact, let's go all the way and say that these variables are meaningless. Maybe you think that a lot fewer than 20,000 people will end up moving, so jobs are not an issue, and maybe you think that we should not try for autonomy, so federal dependence is not an issue. Then what happens? Wyoming and to a lesser extent Alaska become clear winners, followed by North Dakota; Vermont, Delaware, and South Dakota are next but not very close. What if you think, in addition to all that, that geography is not important? Then Wyoming wins by a long shot and North Dakota comes in second. So the proper way to argue for North Dakota is to make the case for including cultural factors, to argue that jobs, geography, and federal dependence are not at all important, and to say that some as-yet-to-be-identified subjective or non-quantifiable factors disqualify Wyoming. I think most people would say, however, that if anything, the non-quantifiable factors favor Wyoming, but my point here is simply to point out what needs to be argued to make the case for North Dakota.

So let's have a rundown. What do you need to argue in order to make the case for each of the candidate states?

To make the case for Wyoming - Size is important; jobs are not at all important. Political culture and quality of life variables do not have an impact: Wyoming does well whether they are included or not; however, Wyoming only beats Delaware (and Alaska) when jobs are not considered and coastal geography is not considered crucial, OR when jobs are not considered and lifestyle and non-quantifiable factors eliminate Delaware and Alaska but not Wyoming.

Alaska - Size is important; jobs are not at all important. Political culture and quality of life variables do not have an impact: Alaska does well either way; however, Alaska only beats Wyoming when geography is emphasized and only beats Delaware when federal dependence is minimized (and jobs are not considered at all), OR if lifestyle and non-quantifiable factors eliminate Delaware but not Alaska. You also need to make the case that we can get 20,000 commitments for Alaska.

North Dakota - Size is important; jobs are not at all important. Political culture and quality of life variables do not have a major impact: North Dakota does fairly well either way, though a little worse when political culture variables are included. However, North Dakota only beats Wyoming, Vermont, and Delaware if federal dependence and jobs are considered totally unimportant and geography is considered somewhat important - then it is still behind Alaska, so you also need to make the case that lifestyle or non-quantifiable factors eliminate Alaska but not North Dakota.

Vermont - Size is important; jobs are not at all important. Political culture and quality of life variables do not have a major impact: Vermont does well either way. If you think political culture is not at all important, then Vermont is slightly behind Wyoming and a little further behind Delaware. So you have to argue that geography is very important (to get it past Wyoming) and that Delaware is eliminated due to subjective factors. If political culture is important, then Vermont moves ahead of Delaware. You have to argue then that geography is really, really important, that the Canadian border Vermont has puts it ahead of Wyoming.

Delaware - Size is important; jobs are somewhat important. Even when jobs are not considered at all, and political culture variables are considered, Delaware does better than any other state except Wyoming and Vermont (with whom it is basically even). When jobs are considered, of course, Wyoming and Vermont drop out. If political culture variables are not important, Delaware does better, pulling ahead of Vermont but still behind Wyoming. If geography is not considered at all important, and political culture variables are considered, and jobs are not considered, Delaware falls significantly behind both Vermont and Wyoming, but remains slightly ahead of Alaska and well ahead of the rest. If federal dependence is not considered at all, then North Dakota and Alaska move well ahead of Delaware, and South Dakota slightly ahead, but Delaware remains ahead of Montana, Idaho, New Hampshire, and Maine. So if jobs are important, Delaware always wins - unless federal dependence is totally unimportant, geography is totally unimportant, and political culture is important. Then South Dakota, Montana, and Idaho are slightly ahead.

South Dakota - Size is important; jobs are somewhat important. About the only way to make a case for South Dakota is to argue that jobs are important, that federal dependence is totally unimportant, and that geography is totally unimportant. Under these parameters, South Dakota competes well with Delaware, and beats Delaware handily when political culture is considered. South Dakota might be the right choice for anti-Easterners concerned about jobs but not federal dependence. If jobs are important and federal dependence remains important, but geography is unimportant, South Dakota is well behind Delaware but slightly ahead of Idaho and New Hampshire. So you could also argue that if Delaware should be eliminated for subjective reasons, if jobs are important, and if geography is unimportant, whether federal dependence is important or not, then South Dakota is the (marginal) winner.

Montana - Size is important; jobs are somewhat important; federal dependence is unimportant. In addition to eliminating states that fall below the jobs threshold, you have to eliminate the weighting on federal dependence, keep the geography weighting, and rely on subjective factors to make Montana the winner. When all this is done, Delaware is still well ahead of Montana and South Dakota is slightly ahead. So you have to employ subjective factors to eliminate Delaware and put Montana over South Dakota. (When political culture is considered, Idaho is also ahead of Montana.) Montana might be the right choice for anti-Easterners concerned about jobs but not federal dependence and who believe Montana beats South Dakota on non-quantifiable factors or geography.

Idaho - Size is not important; jobs are important. Believe it or not, the case for Idaho is about as easy as the case for South Dakota or Montana, despite its size. If you eliminate the smaller states due to jobs issues and Delaware due to lifestyle or subjective factors, but keep the federal dependence weighting, even if you keep the full Size weightings, Idaho pulls ahead of South Dakota and Montana, but remains behind New Hampshire. To make the case for Idaho over New Hampshire, you also have to argue that jobs are really important above and beyond the minimum required to sustain us, or you can include political culture variables, which depending on how you tweak them can put Idaho slightly ahead of New Hampshire. If you put a large positive weighting on jobs, Idaho pulls ahead of New Hampshire. If size is not very important, the weightings on number of voters and campaign finance being reduced dramatically, then Idaho still falls behind Wyoming, Alaska, North Dakota, Vermont, Delaware, and New Hampshire. Only if size is totally unimportant, and jobs are important, does Idaho pass most of these - again, except New Hampshire.

New Hampshire - Size is not important; jobs are important; federal dependence and political culture are important. New Hampshire's main opponent (besides Delaware) when size is de-emphasized and jobs are emphasized is Idaho. New Hampshire gets the edge over Idaho when Delaware is eliminated for subjective factors and federal dependence is included. Political culture variables can favor either Idaho or New Hampshire: the economic and gun variables favor New Hampshire, but schooling variables, amount of private land, and such help Idaho. Despite its significant fall in the rankings since last time, New Hampshire is the appropriate choice for those who do not care about East vs. West but are concerned about jobs and think Delaware should be eliminated for subjective reasons. This holds true even when Size variables are at full weighting. Only a heavy reliance on subjective factors and/or a total de-emphasis on federal dependence can get Montana past Idaho and New Hampshire. A de-emphasis on federal dependence and a heavy emphasis on jobs or an emphasis on quality of life variables like amount of land are necessary to put Idaho past New Hampshire.

Maine - There's no nice way to say this: the case for Maine is very difficult. The only way Maine wins is if geography is the only variable you consider, and Alaska is eliminated for lack of jobs or difficulty in obtaining commitments, and Delaware is eliminated for lifestyle and non-quantifiable reasons. As soon as you take into consideration anything other than geography, Maine starts dropping like a rock. It does well only on geography, lack of urbanization, and low crime rate. It is mediocre on livability, federal dependence, government employment, amount of private land, and gun control, and does poorly on everything else.

January 9, 2003

The views expressed in this essay do not necessarily represent those of Free State Project, Inc.