Finding Patterns in Counties’ Statistics (Michael Pakko Commentary)

In preparation for a recent presentation to the Association of Arkansas Development Organizations, I gathered a set of statistics on demographics and economic development for Arkansas’ 75 counties. Mapping the results yielded some interesting patterns, although in many cases not terribly surprising. In most cases, the counties that are part of the state’s metropolitan areas have higher growth in both population and incomes. They are also the areas that tend to have lower rates of unemployment and poverty and higher rates of educational attainment.

Even more interesting was the pattern of correlations among the various statistics. Some variables correlated with others for obvious reasons — for example, those counties with the highest employment growth rates tended to have lower unemployment rates.

Other relationships were more surprising. The higher a county’s share of manufacturing as a percent of total employment, for example, the lower both the per capita income and poverty rate. Evidently a strong presence of manufacturing tends to level out the distribution of income. However, with manufacturing employment declining across the state, counties with high concentrations of manufacturing employment also tended to have the lowest rates of job growth in recent years.

One variable that seemed surprisingly unrelated to any others was the child dependency ratio — the number of residents under age 18 relative to the total number of 18 to 64 year-olds. On the other hand, the similarly constructed old-age dependency ratio (measuring the population 65 and over) was positively related to unemployment and negatively related to personal income, particularly income net of transfer payments (which include Social Security and Medicare). Counties with a high share of older residents also tended to have a low share of bachelor’s degree holders.

Two sets of demographic variables were significantly related to economic outcomes for Arkansas’ counties. The first was educational attainment. Whether measured by high school diplomas or bachelor’s degrees, the number of educated residents of a county is positively related to per capita income and employment growth, and negatively related to unemployment and poverty rates. Counties with a high share of educated residents have a low proportion of their personal income coming from government transfer payments, and have a low concentration of manufacturing employment.

The other variable that is consistently correlated with economic activity is population growth. From the decennial census in 2010 through 2014, only 22 of Arkansas’ 75 counties experienced positive population growth. In addition to being negatively correlated with the old-age dependency ratio, counties with high population growth rates are strongly associated with lower unemployment and poverty rates and also tend to be associated with positive employment growth since the onset of the 2008-09 Great Recession. (Only 18 Arkansas counties experienced positive job growth from December 2007 through December 2014.)

Population growth rates were also very highly correlated with the educational attainment measures. The counties with the highest levels of educational attainment are also those that are experiencing the most rapid growth.

Although all of the relationships cited here are statistically significant, they represent correlation, not causality. It could be the case that counties with weaker economies have fewer resources and lower priorities for education. More likely, however, is a scenario where young educated Arkansans are moving to where economic opportunities are greater — creating clusters of more highly educated and rapidly growing regions of the state.

The causality clearly matters when it comes to formulating potential public policy responses to divergent economic trends, but an excellent first step is identifying some of the characteristics of different regions around the state.

Michael Pakko is chief economist and state economic forecaster with the University of Arkansas at Little Rock’s Institute for Economic Advancement. Copies of the color-coded maps prepared for the Association of Arkansas Development Organizations meeting are available at his blog, ArkansasEconomist.com.