North Carolina and, indeed, the entire South have made enormous economic progress in recent decades. Broad measures of economic improvement like business sales, employment and income have increased faster than in the nation over the last 30 years. This, of course, is a track record to be proud of.
But the gains in prosperity have not been equally distributed, especially geographically. In general, improvement in urban areas has outpaced that in rural areas. Some rural regions have actually slipped backward.
A decade ago there was excitement that information technology (IT) would change this. By increasing access to people, information and markets ó wherever oneís location ó it was thought IT would reduce the disadvantages of businesses being located in more remote, less-dense areas. In short, IT would negate the costs of distance.
But the opposite appears to have happened. Metropolitan areas like the Triangle and Charlotte have been some of the fastest-growing regions in the country. In contrast, some rural counties in North Carolina have been losing population.
Harvard economist Ed Glaeser has an explanation for these results. IT allows more contacts to be made between people and businesses. But, Glaeser argues, IT is often not a substitute for face-to-face meetings. So more IT contacts lead to a desire for more face-to-face meetings, which is exactly what urban areas (cities) provide.
Does this automatically mean rural areas have fallen further behind their urban counterparts in North Carolina? Not necessarily, because public policy efforts could counteract the impacts of IT. For example, the state wields considerable power to influence economic development through spending on public schools and highways.
In North Carolina, the Leandro court ruling in 1994 mandated that the state provide equal educational opportunities for students in all public schools across the state. In addition, federal educational funding in the last two decades has focused more resources on under-performing schools. Significant road money has also been directed to rural counties in recent years.
To compare the economic progress of urban and rural areas, the obvious first step is to define urban and rural. The standard definition is based on density (persons per square mile). Usually some cutoff is used to separate urban and rural, meaning areas (counties) with densities above the cutoff are considered urban, while those with densities below the cut-off are classified as rural.
One problem with this method is that any cutoff is necessarily arbitrary, and different cutoffs will give different numbers of urban and rural counties. Also, if density is the major feature defining urban and rural, then collapsing the full range of density into two groups loses considerable information about the effects of density.
So instead of separating North Carolina counties into urban and rural, I looked at correlations between county density and various economic measures over time. A correlation is simply a statistical measure showing how closely two measures are related. For example, a correlation of 1 means two measures move exactly in the same direction, a correlation of -1 means two measures move exactly in opposite directions, and a correlation of 0 means there is no relationship between the measures.
I found some interesting results. First, letís look at two policy measures: school spending and roads. Public school (K-12) spending from all sources (federal, state, local) was slightly positively related to density 25 years ago, meaning spending was higher in more dense (urban) counties. Today it is actually modestly negatively related to density, meaning spending per pupil increases as density decreases; that is, as counties become more rural.
Paved highway mileage per square mile in the county was positively related to density both in the past and now, indicating the availability of roads is higher in more dense (more urban) counties. But the correlation is lower today than 25 years ago, so highway availability has become closer for counties of all density levels.
Now what about economic outcomes? SAT math scores, business establishments per capita (per person), and average worker earnings all are positively related to density. But for SAT math scores and business establishments per capita, the positive relationship is lower today than in the past. So density appears to be less of a factor for school performance and business starts.
However, the positive correlation between density and average worker earnings is higher today, implying good-paying jobs are even more readily found in more dense counties now than in past years. This finding supports the theory that IT has favored urban counties for jobs and pay.
My analysis suggests that North Carolina, through its public school and highway policies, has tried to bring urban and rural counties together ó and with some success. But technology seems to be working in the other direction. Which will win in the future? You decide!

Dr. Mike Walden is a professor and N.C. Cooperative Extension economist at N.C. State University.