Real State Per capita Income
The Bureau of Economic Analysis has been working on creating state and metro cost of living indices for several years and they have just published a new set of them that can be used to create real per capita income comparisons between states.
In their press release they show a map comparing real growth in 2012 that ranged from plus 12.7% in North Dakota and minus -2.3% in South Dakota.
http://www.bea.gov/newsreleases/regional/rpp/rpp_newsrelease.htm
But I found the ranking of the states real per capita income much more interesting.
Technically, the District of Columbia has the highest real per capita income but I suspect that measure is distorted the same way the D.C. real GDP is distorted. Many people work in D.C., but commute in from other states. Because they live elsewhere they are counted in those states population, but their income or output generated in D.C. is included in the D.C data. Interestingly, Luxemburg has the same distortion that gives it the highest real per capita GDP in Europe.
BEA had published some preliminary results from this for 2005. The biggest change from that data appears to be New York moving from near the bottom of the ranking to number 15 and Texas falling from number 15th in rank to number 24th. Deleware fell from number 1 to number 28th in rank. I suspect that the major reasons for these rank changes stemmed from improvements in the data and methodology since BEA did the 2005 calculations.
Differences in rent is the major factor driving the differences in the cost of living by state. For example, the Massachusetts price index is index is 107.2 and within that rent is 121.4. In Mississippi, in contrast the overall price index is 86.4 and the rent index is 62.1. So the difference in the overall index is about 14 while the rent index spread is about 60. Rent in Massachusetts is approximately double rent in Mississippi. Mississippi has the lowest rent in the country while Hawaii is the highest at 159. Hawaii has the second highest overall cost of living at 117.2 compared to 118.2 in Washington, D.C.. Only 14 states regional price indices are greater than 100.
Real per capita income in the top ten states is 135% of that in the bottom ten states.
I stick with my long time comments that Massachusetts is 120% of the national average and Mississippi is 89% of the national average, but the republicans/libertarians keep trying to tell us that we would be better off is we were more like Mississippi and less like Massachusetts.
Real Per Capita Personal Income
Chained (2008) dollars
…………………………………………rank…….. % of US
$41,282 United States
$59,759 District of Columbia………………….145
$57,367 North Dakota…………. 1…………… 139
$51,559 Connecticut……………. 2……………125
$49,587 Wyoming……………….. 3………….. 120
$49,355 Massachusetts………… 4 …………..120
$48,627 South Dakota…………. 5 ……………118
$47,188 Nebraska………………. 6……………. 114
$46,376 Iowa …………………….7……………. 112
$45,702 Maryland……………… 8……………. 111
$45,552 New Jersey……………. 9……………. 110
$45,494 Minnesota…………… 10……………. 110 ……..117.9 average of first ten states
$45,217 Kansas………………… 11……………. 110
$44,314 Virginia………………. 12 …………… 107
$43,905 Rhode Island………. 13……………. 106
$43,722 New Hampshire…… 14……………. 106
$43,603 New York…………… 15……………. 106
$43,601 Alaska……………….. 16……………. 106
$43,173 Pennsylvania………. 17…………… 105
$43,063 Illinois………………. 18……………. 104
$42,846 Wisconsin………….. 19 …………….104
$42,701 Oklahoma………….. 20…………… 103………..105.7 average of second ten states
$42,559 Colorado…………… 21……………. 103
$42,427 Ohio…………………. 22…………… 103
$42,164 Washington……….. 23…………… 102
$41,961 Missouri ……………24 ……………..102
$41,733 Texas………………. 25…………….. 101
$41,726 Vermont………….. 26 ………………101
$41,432 Louisiana………… 27……………… 100
$41,282 United States………………………… 100
$40,848 Delaware………… 28………………… 99
$40,371 Tennessee………. 29………………… 98
$39,553 Indiana…………… 30………………… 96………101.5 average of middle ten states
$39,225 Florida…………… 31…………………. 95
$39,103 North Carolina…. 32………………… 95
$38,888 California……….. 33 ………………….94
$38,665 Montana………… 34…………………. 94
$38,530 Alabama……….. 35…………………. 93
$38,516 Maine…………… 36 …………………..93
$38,479 Georgia………… 37………………….. 93
$38,317 Michigan ………..38…………………. 93
$38,224 Arkansas ……….. 39………………… 93
$37,909 Kentucky………. 40…………………. 92……………93.5 average of fourth set of ten states
$37,451 Oregon………….. 41 …………………..91
$37,425 West Virginia….. 42 ………………….91
$36,803 Mississippi……… 43………………… 89
$36,769 Nevada………….. 44………………… 89
$36,507 South Carolina… 45………………… 88
$36,087 Hawaii…………… 46………………… 87
$35,553 New Mexico…… 47…………………. 86
$34,905 Arizona………… 48………………….. 85
$34,819 Idaho ……………49………………….. 84
$34,580 Utah……………. 50………………….. 84……………..87.4 average of bottom ten states
$59,759 Maximum
$34,580 Minimum
$25,179 Range
$41,282 United States
I’m kind of surprised New Hampshire is so tight with Massachusetts.
I wonder if there is some process factor that is tying up Boston and Southern New Hampshire rents.
Is this the average for each state?
If so, the median would be a better metric.
Interesting either way.
Southern New Hampshire is essentially a suburb of Boston.
Have you ever seen the rush hour traffic between Mass & New Hampshire?
Spencer, Please address the following question. Given that the “real” per capita personal income measure is a calculation based on personal income and some measure of regional cost of living (RPP), and that personal income is described as follows, “Personal income is the income received by all persons from all sources. Personal income is the sum of net earnings by place of residence, property income, and personal current transfer receipts,” what is the error measurement for RPI that may result from distributional distortions of personal income?
I realize that that question is a bit convoluted. What I’m getting at is my discomfort with all comparative statements of personal income for very large populations. I’ve seen no reference to how the measure of central tendency of personal income is corrected for distributional distortions. In a nut shell, is it realistic to suggest that any large population on income earners can be summarized in a single personal income measure? No one seems to say what measure of central tendency is being used. No one notes standard deviations from the measure being used. If one percent of a population earns 35% of the personal income is it practical, or meaningful, to speak of an average measure of personal income?
You have a valid point.
Why don’t you call the BEA and suggest they publish the factors you are discussing.
Eric Figueroa at 202-606-5620 or
Jeannine Aversa at 202-606-2649
But I suspect income distribution in the individual states is very similar to the national distribution. But seeing how it differs would be interesting. I suspect the wealthier state have more wealthy individuals but the poorer states would still have a more unequal distribution.
I would be interested to see how farm incomes fit in with the overall picture, given that South Dakota, Nebraska, and Iowa, three states not generally perceived as “high income” states come in at #s 5, 6, and 7.
Spencer, it is and it isn’t.
People live up there because it does cost less to some degree, and they’re willing to tolerate a long commute. Personally I commute northward on 93, so I see it every day, but from a position where I don’t have to actually live in it.
If it was treated like a single market, that’s still incorrect, even if it is a bedroom community for Boston.
The Bureau of the Census includes the New Hampshire counties of
Stafford, Rockingham and Manchester-Nashua in the greater
Boston-Worcester-Manchester Metropolitan Statistical Area.
I go north on route 3 to Nashua from Woburn to shop at least once a month.
” What I’m getting at is my discomfort with all comparative statements of personal income for very large populations. I’ve seen no reference to how the measure of central tendency of personal income is corrected for distributional distortions.”
Jack ,
Anwar Shaikh has come up with a method to correct per capita income measures for the effects of inequality , arriving at a figure he calls “VMI” , or Vast Majority Income , which represents the income per capita of the lowest 80% of the distribution. You need to know the Gini coefficient for the selected population , and can then calculate VMI :
VMI = 1.1 ( 1-gini ) ( per-capita income )
http://www.ipc-undp.org/pub/IPCPolicyResearchBrief7.pdf
Worth noting is that Shaikh is one of many who have helped to discredit the neoclassical production function , and in a particularly clever way , in his 1970s “Humbug” exercise :
http://ideas.repec.org/a/mes/postke/v34y2011i2p255-274.html
Marko,
Thanks for the reference. It seems to begin to get at the problem of reporting a mean data point when the distribution is skewed and the population is large. Why no reference to the standard deviation from the mean, which would be the more typical measure to describe the spread of the data? The mean score by any other name would still be only the mean score and it needs greater definition.
Shaikh’s VMI measure would imply a bi-modal distribution of income, or at least force the data into a bi-modal shape. I’ve not seen income data referred to as having a bi-modal distribution though I suspect that that may be a fact and the reason why so little is done to correct the disparate distribution of wealth and income. In effect, are we really one big economy or two over lapping economies whereby the healthier of the two subjugates the other to its own benefit?
I don’t see how any bi-modality is necessarily implied. Changes in the Gini index can do a fairly good job of detecting increased concentrations of income , but it’s not perfect by any means , especially if it’s survey-based.
What Shaikh shows in the paper is that when countries are ranked based on mean per-capita incomes , and then re-ranked based on inequality-corrected data , the rankings don’t change a whole lot , but there are exceptions , like Chile , where inequality makes a difference. ( However , if your Gini is survey-based , and missing significant top-end income concentration , the results could be suspect.)
Still , for re-ranking the state listing above , the Shaikh method can give you some measure of how well the low-to-middle-income classes ( i.e. the average of the bottom 80% ) are doing in each state.
Take New Hampshire and New York in the list above. They have nearly equal mean per-capita incomes ( $43,722 and $43,603 , respectively – a difference of ~ 0.3% ), yet the calculation for per-capita income of the bottom 80% , per Shaikh , yields values of ~$27654 and $24029 per-capita , respectively.
That’s a 15% difference , and in a re-ranking of the states by this measure, would no doubt separate NH and NY by a fair amount.
One way to think about those results for NH and NY. Let NH represent the U.S. in the 1960s , and NY the U.S. of today.Assuming the same overall U.S. per-capita income , the paychecks of the bottom 80% would be 15% higher today ( and for all the days since the 60s ) if we had evolved to be more like NH , and less like NY.
BTW , I got the state Gini data here :
http://en.wikipedia.org/wiki/List_of_U.S._states_by_Gini_coefficient