With the Fall season comes football and favorite holidays, but it’s also a time when high school seniors begin to think about college applications.
One likely target for those applications? The University of Georgia, often jokingly referred to as The University of Cobb County given how many students apply and are accepted from Cobb each year. But as nicknames go it really doesn’t apply. Not any more.
When it comes to how many undergraduate students they supply to UGA, some counties are moving up and some are moving down the rankings. The table below ranks the top 10 counties in 1998 and 2008 (the last year data was available).
The University of Cobb County? The county has dropped from second to third place among all Georgia counties. Others slipping down the rankings are Bibb, Houston, and Clayton. The up-and-comers? Oconee, Forsyth, and Henry counties.
So what’s it all mean?
When it comes to explaining why a county moves up and down the rankings, you have to look at population and poverty and plain old geography.
Let’s start with the assumption that feeding the UGA educational machine is a good thing. As I teach at UGA, I’m clearly biased (for or against, you decide). Now that I’ve practiced semi-full disclosure, let’s move to some of the interesting tidbits found in the data.
A glance at the map to the right may not be worth a thousand words, but it sums up the matter nicely. The most populous counties, obviously, send the most students to UGA. The core Atlanta area in white dominates, followed by the surrounding suburban counties in gray and then a bunch of other colors.
There are a few exceptions. Clarke County is geographically the smallest in the state, but it’s also in white. The reason is simple. That’s where you find UGA and neighborhoods full of overeducated faculty and their overachieving kids. Next door to Clarke is Oconee, basically an Athens suburb with good schools and less poverty – where every kid is above average – so you’d expect to see its numbers on the rise.
In the South, you never quite escape the problem of poverty.
If you examine the counties with the lowest poverty rates, they also coincidentally happen to be the ones with the biggest increases in students sent to Athens.
In the data I used, Forsyth County boasted the lowest poverty rate of all 159 Georgia counties (5.2 percent – your mileage may vary depending on the data used). Between 1998 and 2008, Forsyth had a remarkable 206 percent increase in students attending UGA, the largest percentage increase of any county. Combine that with the county’s 67 percent population increase and you easily understand how it improved from 20th place in 1998 to 8th in 2008.
Bibb County, however, has a relatively high poverty rate combined with less population growth, at least compared to many Georgia counties. So it’s not surprising that the number of UGA undergrads from Bibb dropped 15.9 percent between 1998 and 2008.
Simply put, the 10 Georgia counties with the lowest poverty rates all have positive increases in the number of students enrolled at UGA during this period.
The counties with high poverty rates tend to send so few students to UGA, examining the percentage change becomes mathematically challenging. Clarke County is the exception. With a poverty rate of nearly 29 percent, it continues to hold on to 5th place in the rankings, a demonstration of the power of geography and the odd splits found there in education and income.
And then there’s Echols County.
Located just east of Valdosta, it has but a few thousand residents. Other fun facts:
- 94 percent of the county’s land is covered by trees
- The Tax Assessor’s office is closed for lunch
- Statenville, the county seat, was once known as Troublesome
- The county has only one public school – from kindergarten through high school
Final fun fact – rarely is an Echols County resident a student at UGA.
Between 1999 and 2002, the county had no undergraduates listed as enrolled at UGA. In 2003 and 2004 one student was enrolled. In 2005, no one. Between 2006 and 2008, one student from Echols County was listed as enrolled.
From my analysis, Echols has the lowest enrollment of any of the state’s 159 counties over the time I examined, at least in terms of UGA undergraduate admissions.
You’d get awfully lonely searching campus for a fellow Echols County Wildcat.
Why so few students? Not lousy grades. A quick scan of test scores suggests the county does a fine job and performs well in the acronym salad that makes education such a treat (CRCT, AYP, etc.).
Instead of grades, it’s three things: population, poverty, and geography.
Echols is among the least populated counties in Georgia, one with a 27 percent poverty rate, so put the two together and you’d hardly expect it to Sanford Stadium. And it’s a long drive from Statenville to Athens, while only a few miles away the school is a perfectly good Valdosta State University.
Warning: Stats Ahead
Which is the better explanation, population growth or poverty?
I built various different regression models to explore this. The great thing about multiple regression is that each of your predictor variables (here, population size and poverty) statistically control for each other.
Maybe I lost you there. Think of it this way. Age is related to income. So is education. But if I control for education, it’s possible age is no longer a factor. Education may overwhelm age in predicting income. Social science geeks like me would say that education explains unique variance, but age does not.
Okay, so I tried a number of ways to do this – brute empiricism at its best, the details best left to a footnote in an academic journal read by tens of people worldwide. I’ve buried the lede. The result is simple: population is a more powerful predictor than poverty. When you control for population in various ways (I tried several), poverty no longer matters. Or flip it around. If you enter poverty first in the model, it has a negative relationship with how many kids a county sends to UGA. But if you add population, that poverty relationship disappears.
Do I buy this? Not completely. In fact, I’m fairly sure my analysis is messed up somewhere or not as developed as it should be, but best I can tell, adding lots and lots of people to your county will offset the effects of poverty in how many kids attend UGA.
A Case Study: Clarke and Oconee Counties
As we saw earlier, Clarke County ranks high among Georgia counties but its next door neighbor, Oconee, is racing up the charts with a bullet.
This is largely a function of demographics, white flight, and population growth on the part of Oconee, which sucks many middle class and upper middle class folks out of Clarke. The result is seen to the right, with the blue line representing the number of UGA students enrolled from Clarke and the red line from Oconee (yes, I chose those colors on purpose, but that’s a different story). The big year was 2004 when Clarke dropped below 1,000 students for the first time while Oconee rose above 500.
Will Oconee ever catch Clarke?
Not for a while. The trendlines appear to be flattening for both Clarke and Oconee, but here’s a telling statistic: if you create a proportion of the people who attended UGA in 1998 and 2008, one county in Georgia has the biggest increase.
The county with the biggest decrease?
As a case study you can’t do much better than Clarke County with its high poverty numbers compared to Oconee County, which may not even allow poor people to live there. Clarke used to have about 11 UGA students were 1,000 population. Now it’s 8. Oconee used to have 15 students per 1,000. Now it’s 19.
So sheer population size isn’t enough. Population growth matters, but so does poverty.
To Finally Finish
I’ve subjected you to a lot of number crunching. And I skipped a lot of factors, such as how well a county scores on SATs, etc., that might explain enrollment patterns at UGA. But since poverty is highly related to standardized test scores, I’m willing to risk it.
Want to look for yourself? Here’s an example of process journalism, or self-correcting journalism. I’ve parked the data here on my server so you can call up the Excel file and fiddle at your own leisure. One worksheet, cleverly named The Data, has — obviously — the data. The other worksheet, called What Stuff Means, explains what is in each column. Add your own data if you like, run your own analyses. Give in to the dark side and be a numbers geek. But best of all, tell me where I screwed up or what you found and maybe we can drive the editors crazy and change this story on the fly, day to day, or at least as long as it lasts on the front page.