USA Today published a piece from 24/7 Wall St. on the best and worst state economies. The Drudge Report linked to the piece, likely multiplying its readership manifold. Unfortunately, the ranking is fundamentally flawed for two key reasons which will be explained below.

Colorado, Utah, Massachusetts, New Hampshire and Washington captured the top-5 in the ranking. Among the five most-populous and diverse states, California came in at #8 overall, followed by Florida (#13), Texas (#21), New York (#31) and Illinois (#32). Bringing up the bottom of the roster were New Mexico (#46), Mississippi (#47), Louisiana (#48), Alaska (#49), and West Virginia (#50).

The ranking, by 24/7 Wall St. senior editor Michael B. Sauter and analysts Samuel Stebbins and Evan Comen, used five factors to score  states’ economic performance: economic growth from 2012 to 2017, the official poverty rate in 2016, the official unemployment rate in June 2018, employment growth from 2012 to 2017, and the share of adults with bachelor’s degrees.

Of their survey, the authors’ write:

The best ranked states tend to have fast-growing economies, low poverty and unemployment rates, high job growth, and a relatively well-educated workforce, while the opposite is generally the case among states with the worst ranked economies.

The challenge with any rankings is the data included, the data left out, and the weighting of the data used.

In their ranking the authors’ use economic growth from the last five years. But the issue with that data is that the Great Recession hit some states far harder than others. For instance, California’s housing bubble was particularly large compared to Texas’ due to a variety of factors, including state law and the effect of California’s restrictive regulations which artificially reduced new supply thus inflating the value of existing housing. As a result, from 2007 at the onset of the last recession to 2009, California’s real economic output declined 4.4% vs. 3.2% for the U.S. as a whole while the Texas economy contracted an imperceptible 0.01%. Thus, as the economic recovery took hold, California was initially slow to recover, but then it outpaced most of its rivals for a few years with its economic growth only recently slowing.

A better gauge of consistent, sustained growth is a 10-year window, beginning in 2007 before the recession and continuing until 2017, the last year for which data from the U.S. Bureau of Economic Analysis is available.

Over the past decade, the top states by GDP growth are: North Dakota, Texas, Nebraska, Washington, and Oregon. The slowest growing states are: New Jersey, Louisiana, Wyoming, Nevada, and Connecticut with the latter two seeing a decline in real GDP of 2.2% and 9.1%, respectively.

The other factors used in the USA Today ranking have issues as well. The greatest determiner of poverty is employment, not government benefits. People with jobs are far less likely to be in poverty. Further, job growth is closely linked to economic growth and is inversely related to unemployment—the more job growth, the lower unemployment generally is. Thus, of the five factors used in the survey, four are largely redundant with economic growth having a high correlation to both the unemployment and job growth while poverty is closely correlated with the unemployment rate. The share of adults with bachelor degrees has little statistical correlation to economic growth or poverty but is predictive of demographics as well as the kind of industry in a state.

Lastly, the use of the U.S. Census Bureau’s Official Poverty Measure is itself highly problematic because the official poverty measure, in use for more than 50 years, does not account for the large variations in states’ cost of living—it assumes rents in New York are the same as rents in Mississippi. As a result, the Official Poverty Measure significantly understates real poverty in high cost of living states such as California and New York while overstating poverty in low cost states such as Texas and most of the Midwest and South.

To remedy the shortfalls of the old Official Poverty Measure, the U.S. Census Bureau developed the Supplemental Poverty Measure, which does look at the cost of housing from state-to-state as well as a wider array of benefits for the poor and costs incurred by the poor. Further, unlike the Official Poverty Measure, the Supplemental Poverty Measure accounts for the value of noncash government assistance such as food stamps (now called Supplemental Nutrition Assistance Program or SNAP) or housing vouchers, as well as other common expenses, such as out-of-pocket medical costs. Thus, states that have a more robust safety net have more of that assistance measured, though data would suggest that employment has a stronger effect on reducing poverty than welfare alone.

When using Supplemental Poverty Measure, the states with the highest poverty as averaged from 2014 to 2016, are: California (20.4%); Florida (18.8%); Louisiana (18.4%), Arizona (17.8%) and Mississippi (16.9%). The national average Supplemental Poverty rate over the last three years reported was 14.7%. Texas’ poverty rate was at the national average.

The states with the lowest Supplemental Poverty rate were: Minnesota (8.0%); Vermont (8.6%); New Hampshire (8.8%), Iowa (8.8%) and Utah (9.4%). These states are among the least diverse in the nation, however, with the minority population of Vermont (6.5%) the second-lowest in the nation, followed by the 4th-least diverse, New Hampshire (8.7%); #5 Iowa (12.9%); #13 Minnesota (18.6%); and #18 Utah (20.7%). The most populous states are very diverse, with just under 40% of Illinois’ population being minority while both California and Texas are majority-minority.

Combining two key factors, economic growth from 2007 to 2017 and the Supplemental Poverty Measure from 2014 to 2016, provides a better look at a state’s economic well-being. By these two measures, each equally weighted using the percentage departure above or below the national average for both measurements, we see: North Dakota (#1), Nebraska (#2), Texas (#3), Utah (#4) and Washington (#5). The bottom five are: New Jersey (#46); Arizona (#47); Louisiana (#48); Nevada (#49); and Connecticut (#50).

Among the five big states, the results show:

State 10-year Real GDP 2007-2017 Supplemental Poverty Rate
Texas 30.4% 14.7%
California 19.4% 20.4%
New York 12.7% 16.0%
Illinois 5.1% 13.4%
Florida 4.6% 18.8%

The 50-state ranking for economic vitality follows:

1 North Dakota
2 Nebraska
3 Texas
4 Utah
5 Washington
6 Minnesota
7 Colorado
8 Oregon
9 Iowa
10 South Dakota
11 Oklahoma
12 New Hampshire
13 Idaho
14 Montana
15 Pennsylvania
16 Tennessee
17 Vermont
18 Massachusetts
19 Wisconsin
20 Maryland
21 California
22 Kansas
23 South Carolina
24 Ohio
25 Indiana
26 Hawaii
27 North Carolina
28 New York
29 Arkansas
30 Delaware
31 West Virginia
32 Georgia
33 Missouri
34 Michigan
35 Virginia
36 Rhode Island
37 Kentucky
38 Wyoming
39 New Mexico
40 Illinois
41 Alaska
42 Maine
43 Alabama
44 Florida
45 Mississippi
46 New Jersey
47 Arizona
48 Louisiana
49 Nevada
50 Connecticut

The most effective policy remedy to reduce poverty is robust job growth and a strong economy. Nationally, we see that in work with the cutting and reforming of federal taxes along with the reduction of regulations by the Trump administration and the Republican-controlled Congress, under which America is now seeing record low unemployment for minorities. As the new U.S. Census income surveys come in this year and next, America should see a long-welcome decline in its poverty rate.

This commentary was originally featured in Forbes on August 27, 2018.