Education Measures Not Business Indices Correlate with a Better Economy: Ranking the 50 States

August 22, 2016

Education Measures Not Business Indices Correlate with a Better Economy:

Ranking the 50 States

Dave Wells, Ph.D.

Research Director, Grand Canyon Institute


Executive Summary

What’s Better Economic Performance?

On many social welfare and education indices Arizona ranks low.  On many tax and business policy indices Arizona often ranks quite well.  Arizona has for decades experienced much stronger job growth than the nation as a whole, though not so much recently.  This report looks at state economic performance using four measures:

  • Growth in Real State Gross Domestic Product Per Job

o   a measure of productivity growth

o   2001-2014 (longer-term, the last two expansions)

o   2009-2014 (this expansion only),

  • Growth in Per Capita Personal Income Growth

o   a measure of income growth

o   2001-2015 (longer-term, the last two expansions)

o   2009-2015 (this expansion only),

  • Unemployment rate (average of the U3 and U6 rate as captured over the last year), and
  • Gini Coefficient for Personal Income Growth from 2009-2014 (a measure of inequality)



Certainly, we measure the economy in many ways, so why these four measures?

First, if we just measure aggregate “growth” as is often the case, we capture the impact of two different things, growth in individual well-being and growth in the population.  If a population grows substantially, the economy will grow, but that doesn’t necessarily mean that individual well-being improves.

If a population grows substantially, the economy will grow, but that doesn’t necessarily mean that individual well-being improves.

Real Gross State Domestic Product captures the added value of all new goods and services produced in the state measured in inflation-adjusted dollars, so increases in output, not higher prices increase it.  Those goods and services are produced by individual proprietors and those employed by others.  So dividing it by total employment gives us an indicator of the economy’s productivity.  However, the benefits of real GDP need not stay in the state.  Corporate profits, for instance, may go to entities and individuals outside the state and much of the growth early in the expansion went to profits.

Personal Income by contrast measures all incomes accruing to people who reside in the state. However, not all of it is wages and salaries.  About one-third of it is dividends, interest and rent (DIR) as well as transfer payments (e.g., Social Security).  These passive income earning activities may not originate in the state. However, DIR and transfer payments are an important component of individual well-being. To correct for population changes, however, we divide personal income by population to get state per capita personal income.

For both of these measures the focus is not on the absolute value as some states start at much higher levels than others, but on its growth.  Economic growth is best measured over a span of years, so these measures includes the last two business cycles, starting from 2001 and the last business cycle starting from 2009.

To evaluate how well the state labor market is functioning, the unemployment rate is considered for the last year. U3 is the unemployment rate that gets the most publicity, measuring those actively looking for work in the past month divided by those currently employed plus those actively looking for work.  This is the conventional unemployment rate.  U6 is the broadest measure of labor underutilization.  U6 includes unemployment and underemployment—including those actively looking for work, and those who have been employed or sought work in the last 12 months but haven’t looked in the last month due to obstacles or becoming discouraged.  In addition, it includes those working part-time that are not able to find full-time work due to a lack of full-time work or obstacles (e.g., affordable child care) that prevent it.  U6 is roughly twice as large as U3.

Taking the average of U3 and U6 gives a priority to those most actively seeking work, but still includes a measure of the broader labor market.

Finally, growth in real GDP per job and growth in per capita personal income ignores who receives that income.  The Gini coefficient ranges from 0 to 1 and is based on how much the distribution of income deviates from perfect equality. A Gini coefficient of 0 indicates everyone had the same income and a Gini coefficient of 1 means all income was held by one household. In this case, the analysis looks at the recent change in personal income 2009-2014 and how that change in income has been distributed.  If in State A income increased 10 percent and all households had their incomes go up by 10 percent, then the Gini coefficient would not change.  So the Gini coefficient for added income would be the same as the state’s Gini coefficient.  But if in State B upper income families grew by 20 percent and lower income families did not grow at all, then the Gini coefficient would rise, and the Gini coefficient for added income would be higher than the state’s prior Gini coefficient to capture that higher level of inequality. Likewise, if in State C, lower income families incomes rose by 20 percent and higher income families did not change, then the Gini would move in the direction of greater income equality, and the Gini coefficient for additional income would capture that.  With respect to these three cases, the last case, State C would rank best, the state with equal gains would rank in the middle and the state where only upper income people gained would rank worst.

Collectively, these four measures provide an inclusive measure of the economic performance of a state.  Each of the 50 states were ranked on each measure, and then ranked again from the state that had the best average ranking to the state which had the worst one.

Table 1 looks at half the states for Economic Performance: the top 15 and bottom 10.  Unfortunately, Arizona is the only state to rank in the bottom ten for all four criteria.


Table 1 

State Rank by Economic Performance Variable (Top 15, Bottom 10)

State GDP per Job2001/2009-2014 PC PI2001/2009-2015 UnemploymentRate Gini2009-2014 Avg. Rank ECONOMIC PERFORMANCE
North Dakota 1 1 1 14 4.25 1
Nebraska 5 10 3 2 5 2
Oklahoma 4 3 15 11 8.25 3
South Dakota 6 23 2 3 8.5 4
Iowa 2 18 5 17 10.5 5
New Hampshire 16 12 4 12 11 6
Texas 10 8 13 21 13 7
Vermont 17 9 9 22 14.25 8
Hawaii 26 15 11 6 14.5 9
Arkansas 7 7 21 28 15.75 10
Minnesota 9 21 8 27 16.25 11
Wyoming 46 2 16 5 17.25 12
Wisconsin 13 33 14 10 17.5 13
Montana 8 6 17 40 17.75 14
Utah 30 25 7 9 17.75 14
Illinois 35 38 41 30 36 41
Connecticut 44 22 33 48 36.75 42
Delaware 43 49 24 35 37.75 43
North Carolina 37 46 38 31 38 44
Maine 41 43 19 50 38.25 45
New Mexico 27 30 49 47 38.25 45
Florida 48 45 32 38 40.75 47
Nevada 50 50 50 15 41.25 48
Georgia 47 48 39 32 41.5 49
Arizona 42 47 44 42 43.75 50



What best correlates with Better Economic Performance?  Business or Education Indices?


Policymakers frequently express a desire to improve economic prosperity. The best business climate lists typically infer that higher ranking states do better economically. Some argue that tax measures, regulatory indices, and/or labor cost indices are the most important indicators.  Collectively, these form the main basis of many business indices.

On the opposite end would be to focus on labor quality, emphasizing educational outcomes and attainment.

This study looks at state outlook rankings put together in 2009 which purported to predict how states would perform moving forward as well as the most current version of those rankings to see how well they correlate with the Economic Performance Index.

On the opposite side, the study examines how well education measures correlate with better economic performance.

These are the indices evaluated.


  • American Legislative Exchange Council (2009 and 2016 rankings)

o   ALEC’s Economic Outlook rankings are based on having top personal or corporate tax rates that are low, regressive income tax structures (if income goes up, tax burden going down is preferred), requiring that unionized workplaces be open (no one is required to join the union), low taxes overall, recent changes that reduce taxes, a low minimum wage, fewer state employees per 10,000 population, and expenditure controls on state government.

o   Arizona has historically ranked very highly with ALEC. In 2009 Arizona ranked 3rd. In 2016 Arizona ranked 5th.

  • Chief Executive Magazine (2009 and 2016 rankings)

o   Chief Executive Magazine completes an annual survey of 513 CEO’s to rank states based on the “friendliness of their tax and regulatory regime, workforce quality and living environment (This latter category includes not just the cost of living but the education system and the state and local attitudes toward business).”

o   Arizona has historically been ranked highly by Chief Executive Magazine.  In 2009 Arizona ranked 8th. In 2016 Arizona ranked 6th.

  • Forbes Best States for Business (2009 and 2015 rankings)

o   Forbes factors in 40 data points across six areas to determine their rankings.

  • Forbes most heavily weighs their business cost index which relies heavily but not exclusively on Moody’s Analytics Cost of Doing Business.
  • Forbes’ other categories include labor supply which looks at educational attainment and net migration.
  • The regulatory score is largely impacted by the “Freedom in the 50 States” Report of the Mercatus Center at George Mason University.
  • Their economic climate portion focuses on changes in jobs, income (not clear if per capita or just growth in personal income generally), Gross State Product, and unemployment over the past five years.
  • Growth prospects is based on Moody Analytics expectations for jobs, income and gross state product going forward, and
  • Finally, Quality of Life includes cost of living, public education test scores (may not be demographically adjusted), crime rates, average temperature, and cultural and recreational amenities.

o   Arizona ranked 36th in 2009 and had risen to 23rd by 2015.  In 2015 Arizona’s best areas were growth prospects (7th) and labor supply (15th).  Its worst areas were quality of life (40th) and economic climate (34th).

  • Tax Foundation (2010 and 2016 Business Tax Climate Rankings)

o   The tax foundation looks at over 100 tax variables across five areas: corporate, individual income, sales, property and unemployment insurance to determine their ranking.

o   Arizona has done fair with the Tax Foundation.  In 2010 Arizona ranked 28th and by 2016 Arizona was 24th.



To these business rankings four measures of educational outcomes plus combined educational outcome measures were examined.

  • 2005-2009 NAEP 8th grade free and reduced lunch-adjusted math and reading scores

o   Includes 2005, 2007 and 2009 with a random selection of schools selected from each state designed to be representative.

o   In 2009, 2005 cohort would be graduating from high school.  In 2015, the 2007 cohort would include some graduating from college.  This measure provides one indicator of school quality.

o   Free and Reduced Lunch-adjusted students means that each state’s scores were split evenly between students on free and reduced lunch and those not qualifying for the program, so on that level all states were demographically identical.  NAEP on line data does not enable more sophisticated controls for income and other aspects. The two sets of math and reading scores were then added for each year, then added across the three years of testing and the states ranked.  To qualify for free and reduced lunch a student’s family must be no higher than 185 percent of the federal poverty line.

o   Arizona ranks 41st.

  • 2003-2009 High School Graduation Rates

o   Uses Average Freshman Graduation Rate (AFGR). Takes the number of students receiving high school diplomas divided by the size of its freshman class four years earlier.

o   These students would be 18 to 24 in 2009 and 24 to 30 years old in 2015

o   This is based on the percent of freshman graduating four years later as collected from the states and published by the National Center for Educational Statistics

o   Arizona ranks 37th.

  • 2012-2013 Economically-adjusted High School Graduation Rates

o   Uses Adjusted Cohort Graduation Rate (ACGR).  Under the guidelines of No Child Left Behind, the National Center for Educational Statistics has developed a new student tracking based system that tracks actual students who were freshman and what percent of them actually graduate in four years.  The 2011 graduating year was the first year data became available.

o   NCES data includes “economically disadvantaged” students but No Child Left Behind gave the states the ability to define what that constituted.  The Everybody Graduates Center in the School of Education at Johns Hopkins University beginning with its annual report in 2014 for the class of 2012 breaks down for each state the percent of students considered “low income” and “nonlow income.”  It also uses NCES data to indicate what the graduation rate for each state was for “low income” and “nonlow income” students.  The economically-adjusted graduation rates averages the low income and nonlow income graduation rates for each state across the two years and then ranks them to create an economically-adjusted high school graduation rate that gives each state equivalent demographics across these two categories.

o   Arizona ranked 40th.

  • 2005-2012 percent change in adults 25-64 with an Associate degree or higher (years limited due to data availability)

o   Certainly more people with higher levels of educational attainment correlate with higher incomes.  This measure looks at data from the American Community Survey at the percentage change of prime working age adults 25-64 who have  earned an Associate degree or higher.  This could reflect the impact of in-migration for higher quality jobs as well as improvement longer term residents’ educational credentials.

o   Arizona ranks 42nd.

  • Combined Education Measures

o   Average ranking for states across the four education measures( then re-ranked)

o   Average rankings for states across three of the education measures using only one of the two high school graduation measures, so that is not weighted stronger.

o   Arizona ranked 45th in two and 49th in the other.


The Results

Rankings were correlated with economic performance, which examines how well they coincide with each other.  For instance, in baseball the more strikeouts per 9 innings that a pitcher gets has been found to negatively correlate with batting average, meaning more strikeouts generally results in fewer hits.  At the Olympics we find that a country’s per capita Gross Domestic Product correlates with the number of athletes they send to the games.  If very small population countries are eliminated, this correlation increases.  Essentially wealthier countries send more athletes to the Olympic Games.

Correlations range from -1 (perfectly negatively correlated) to +1 (perfectly positively correlated).  To illustrate, among countries with a population of 10 million or more, there is a 0.72 correlation between the number of athletes sent to the Rio Olympics and the country’s Gross Domestic Product per person (or per capita).  For instance, the Netherlands (pop. 14 million) has a per capita GDP of $49,200 and sent 237 athletes to Rio.  Chile (pop. 17.5 million) has a per capita GDP of half as much $23,500 and sent 41 athletes, while Zambia (pop. 15 million) has a per capita GDP of $3,900 and sent only 7 athletes.

In this study’s case we have two sets of rankings from best to worst, 1 to 50, and the question is whether a particular ranking correlates positively with economic performance.

Figure 1 shows the scatterplot for the 2009 ALEC Economic Outlook rankings and Economic Performance along with the 2003-2009 High School Graduation Rate and Economic Performance.


Figure 1

Figure1 Scatterplot Correlations


The ALEC 2009 graph shows no discernable relationship at all.  In fact, the mathematical correlation comes out as 0, technically .003.

The High School Graduation Rate is scattered, but also shows a trend. None of the worst economic performance states had top high school graduation rates (blank space in upper left) and none of the best economic performing states had low high school graduation rates (blank space in lower right).  The two indices correlate—as one goes up (performance decreasing), so does the other.  In this case, the mathematical correlation was 0.571 and due to the greater concentration of the dots, we’re more than 99 percent sure that in fact there is a positive relationship here.

All of the business measures fail to correlate with economic performance with the exception of the Forbes Best Business Climate Index.  However, even Forbes’ best states for business index comes nowhere close to how well each of the education measures correlate with economic performance. 

In Table 2 below you’ll see a bolded number for the correlation for each of the indices and Economic Performance (including the ALEC index).  That indicates how strongly the two move in tandem.  The number in parenthesis below the bold number indicates how likely the correlation in bold could in fact be zero or not aligned with economic performance.  The .573 result for the high school graduation has essentially no likelihood of being zero or in a negative direction.

In Table 2 * indicates that the probability is 10 percent or less that the relationship could be zero or negative. ** indicates that the probability is 5 percent or less that the relationship could be zero or negative, and *** indicates a probability is 1 percent or less that the relationship could be zero or negative.

All of the business measures fail to correlate with economic performance with the exception of the Forbes Best Business Climate Index.   The correlation for Forbes is low to modest .234 in 2009 and .284 in 2015.  And the probabilities that these could be zero or negative relationships are 10.3 percent in 2009 and 4.6 percent in 2015. This particular index is most sensitive to including changes in GDP per job as part of Economic Performance.  When that portion is weighted more, Forbes’ index improves. When it is omitted, Forbes’ index drops off.  The other business indices failed no matter what components and weights were included in Economic Performance.

However, even Forbes’ best states for business index comes nowhere close to how well each of the education measures correlate with economic performance.

The correlation of economic performance with 2005-2009 free and reduced lunch-adjusted 8th grade NAEP scores was .409.  The correlation with 2003-2009 High School Graduation rates was, as already noted, .571.  The 2012-2013 Economically-adjusted High School Graduation rates correlate with economic performance at .480. Finally, the correlation with the percent change in adults age 25-64 with an Associate Degree or higher was .406.  The probabilities for all four of these education measures being positively correlated with Economic Performance was better than 99 percent.  In addition, the NAEP result correlates very strong with the High School Graduation rate.  The Economic Performance Index’s findings for education were robust through seven different formulations of the index which adjusted factors included and the weight of certain growth factors.  No matter what the specification, the correlation with education was strongly positively correlated with economic performance.

When the education rankings are averaged across the states and re-ranked, the combined education indicators correlate 0.6 with economic performance, more than twice the correlation of the best business index.

Collectively, when the education rankings are averaged across the states and re-ranked, the combined education indicators correlate 0.6 with economic performance, more than twice the correlation of the best business index, Forbes, and six times greater than the typical business climate index’s correlation with economic performance.

The full study also finds that all the business indices correlated better with a political index (state percentage vote in 2012 for Romney) or in some cases something not related to economic performance at all, the final NCAA Men’s public state university basketball rankings (a.k.a. March Madness), than they did with economic performance.

Findings relative to Economic Performance are detailed in Table 2 (below).




Table 2

Measure ECONOMIC PERFORMANCECorrelation with 4-Factor Ranking of 14 and 6 year growth in both Real GDP per Job and Per Capita Personal Income, Unemployment with Underemployment, and how equally growth is distributed,  1 to 50.  growth_chart_smallCORRELATES withECONOMIC PERFORMANCE?
ALEC 2009 Economic Outlook Ranking .003(.983)  remove-151678_960_720
ALEC 2016 Economic Outlook Ranking .098(.497)  remove-151678_960_720
Chief Exec. Magazine CEO Survey 2009 -.035(.810)  remove-151678_960_720
Chief Exec. Magazine CEO Survey 2016 .016(.911)  remove-151678_960_720
Forbes Best Business Climate 2009 .234*(.103)  approved-151676_960_720
Forbes Best Business Climate 2015 .284**(.046)  approved-151676_960_720
Tax Foundation Business Tax Climate 2010 .030(.834)  remove-151678_960_720
Tax Foundation Business Tax Climate 2016 .060(.681)  remove-151678_960_720
2005-2009 NAEP free and reduced lunch-adjusted 8th grade reading and math scores .409***(.003)  approved-151676_960_720
2003-2009 Average High School Graduation Rate .571***(.000)  approved-151676_960_720
2012-2013Economically-adjusted High School Graduation Rate .480***(.000)  approved-151676_960_720
2005-2012 Percent Change in adults 25-64 with Associate degree or higher .406***(.003)  approved-151676_960_720



Measure growth_chart_smallECONOMIC PERFORMANCECorrelation with 4-Factor Ranking of 14 and 6 year growth in both Real GDP per Job and Per Capita Personal Income, Unemployment with Underemployment, and how equally growth is distributed,  1 to 50.  CORRELATES withECONOMIC PERFORMANCE?
All FourThree (High School, NAEP, Assoc. Degree)Three (High School Adj., NAEP, Assoc. Degree) .610***(.000).598***(.000)  .579***(.000)   approved-151676_960_720approved-151676_960_720approved-151676_960_720




Business Climate indices may get a lot of press and emphasis from policymakers, but they do not correlate with Economic Performance.  Education measures do.  The educational attainment of its labor force is probably the most important area for a state to focus on, if it wishes to improve its economic performance.  Improvements take investments, not simply policy moves.

Figure 2 summarizes the results of Table 2.  Educational outcomes have a dramatically stronger correlation with economic performance than any of the business measures.




Figure 2

Figure2 Correlation with Economic Performance

The educational attainment of its labor force is probably the most important area for a state to focus on, if it wishes to improve its economic performance.


Educational outcomes will continue to impact Arizona, as the K-12 system prepares a future workforce.  Arizona’s 8th grade NAEP free and reduced lunch-adjusted improvement in 2013-2015 relative to 2005-2009 is the fifth highest in the nation. This outcome suggests a good initial trajectory and worth a more careful examination.

High School graduation rates though look less promising.  The 2012 and 2013 graduation rates for Arizona were only modestly higher than the average of 2003 to 2009.  Many other states improved more and Arizona’s relative ranking dropped to 42nd from 37th. , using the Average Freshman Graduation Rate.  Looking at the Adjusted Cohort Graduation Rate, Arizona is one of only three states that had has seen its graduation rate drop from the class of 2011 to the class of 2014.  The drop in Arizona from 78 percent to 75.7 percent (-2.3 percent) is the largest drop in the country. Arizona’s graduation rate is the sixth lowest in the country.

Arizona has been a leader in cutting taxes including a so-called jobs bill that reduces taxes by more than half a billion dollars annually during this decade.  Arizona has also been a leader in “school choice,” while at the same time falling significantly behind other states in its level of educational investment.  Prop. 123 was a good initial step to rectify the failure of Arizona’s lawmakers to meet the legal minimum funding for K-12 education, but real improvement will take a more significant targeted investment in education,