Skip to content

KFF’s national analysis matches TIGC’s Georgia findings on the Red-Blue Covid-19 divide

Research spotlighting the differences in how Red and Blue America are responding to virtually every aspect of Covid-19 continues to pile up: the Kaiser Family Foundation (KFF) went up yesterday with a report declaring that its polling had found “that political partisanship is a stronger national predictor of vaccination than other demographic factors.”

I haven’t attempted to do an analysis that looks at the full spectrum of demographic factors — race, gender, age, etc. — but I certainly don’t doubt KFF’s findings. Its national polling and findings are very much in line with what TIGF has been watching take shape here in Georgia for well over a year, although there are a couple of minor differences.

KFF reports that, nationally, the vaccination gap between counties that voted for Democrat Joe Biden over Republican Donald Trump has widened slightly from about 12 percent toward the end of last year to 13.2 percent as of January 11th. Nationally, KFF found, the Biden counties were 65 percent vaccinated as of that date versus 52 percent for the Trump counties.

Here in Georgia, as of data published yesterday (January 19th) by the state Department of Public Health (DPH), the split was right at 10 points — 51.9 percent in the Biden counties to 42.0 percent in the Trump counties — and that’s about what it’s been for the past few months.

In terms of raw numbers, however, the Biden counties continue to grow their advantage of vaccinated and virus-resistant residents. As of yesterday, the Biden counties had fully vaccinated just shy of 875,000 more people than the Trump counties — 2.97 million to 2.09 million. Lately the gap has been widening by an average of just over 850 people a day. If that pace continues, the difference will hit one million in mid-June.

Another difference involves booster shots. KFF found that nationally “the share of fully vaccinated individuals who have received a booster dose is the same (37%)” in the Biden and Trump counties. Here in Georgia, the Biden counties are doing better in this category as well: 40 percent of the fully-vaxxed residents of the Biden counties have gotten boosters versus 37.4 percent in the Trump counties.

As regular readers of TIGF know, I’ve been watching a broad range of Covid data through a political prism for more than a year now (see stories here, here, and most recently here). The obvious question is whether the differences between Red and Blue Georgia in vaccination and death rates — which increasingly favor Blue Georgia — will be sufficient to have an impact on this fall’s election outcomes.

Watch this space.

(c) Copyright Trouble in God’s Country 2022

TIGC takes an early look at the Georgia GOP’s gubernatorial death cage match

Ordinarily the Georgia General Assembly is a shoo-in for top honors in the Best Free Show in Town competition. This year it’ll have stiff competition from the Republican-on-Republican death cage match between incumbent GOP Governor Brian Kemp and former President Donald J. Trump’s handpicked lapdog, ex-U.S. Senator David Perdue.

I wouldn’t place a bet on this race right now if my life depended on it, but I would wager that it’ll bring the schism between Republicans in blood-red rural Georgia and Metro Atlanta’s purplish suburbs and exurbs into sharper focus than ever before.

Picture the Georgia GOP as Humpty Dumpty. The one thing we know for sure is that the Kemp-Perdue match will pull him off the wall and bust him into at least two big pieces. The question is whether either candidate can put him back together.

The differences in these two wings of the party are profound. Rural Georgia Republicans are among the poorest and least well-educated voters on the planet. Their suburban and exurban GOP cousins are pretty much the exact opposite: highly-educated, economically productive, and very affluent. It was among this latter group that Trump arguably lost Georgia in the 2020 presidential race.

Trying to parse those voting blocs right now strikes me as an exercise in futility. My first impulse would be to give Perdue the edge, thanks almost entirely to the Trump endorsement. It was, after all, a Trump endorsement in 2018 that doomed former Lieutenant Governor Casey Cagle’s then-frontrunning gubernatorial bid and all but handed the Republican nomination to Kemp. How can Kemp expect to do without that Trumpian support the second time around?

That line of thinking might hold true in rural Georgia, but the ‘burbs are different. I write this without the benefit of any polling data, but I have to wonder if the stink of Trump still clings to Perdue in those climes — and whether Kemp might have the advantage there. I am no Kemp fan, but I think a fair assessment of his first term has to be that it hasn’t been a total disaster (hey, my expectations are pretty low). He’s chalked up some impressive economic development wins and has somehow managed to avoid embarrassing the state on any kind of regular basis.

Okay, okay, he signed S.B. 202 surrounded by a group of mostly over-fed old white guys while sitting under a painting of a former slave plantation, but — let’s face it — that won’t hurt him with most Republicans. I wouldn’t be surprised to see him use it in a campaign ad — especially in the aforementioned rural regions of the state.

*****

One presumed advantage for Kemp is that he’s built up a $12 million campaign war chest. I say “presumed” because recent history tells us a fat bank account is no guarantee of political success in Georgia (See Barnes, Roy, 2002). On top of that, he’s now apparently sitting on a multi-billion dollar state surplus and wants to spend about $1.6 billion in “tax refunds” to all Georgians.

It’s unclear whether he’ll have to report those refunds as campaign expenditures, but, frankly, it’s also unclear whether they’ll do him much political good. Trump’s name was printed on hundreds of billions of dollars in Covid stimulus checks issued in the spring of 2020 — and he promptly went on to lose re-election a few months later.

His successor, Joe Biden, seems to have fared little better with his own stimulus checks (although he did not have his name printed on the checks); based on recent polling data, it’s far from clear that his stimulus program did him much political good.

If the Trump and Biden experience is any guide, Kemp’s taxpayer refunds will be largely forgotten within a few weeks after the checks go out.

For what it’s worth, I think Kemp missed a Nixon-to-China moment. With a few billion spare bucks lying around, why not put it to strategic use and plow it into hardwiring rural Georgia for broadband internet service? Broadband has, after all, been held out by many Republicans as key to rural Georgia’s salvation, and that kind of initiative would have created hundreds if not thousands of jobs and helped build a foundation for economic development in the parts of the state that need it the most.

(c) Copyright Trouble in God’s Country 2022

Another blogger weighs in on Covid’s political impact

In my last post, I wondered aloud if Covid would kill so many more Republicans than Democrats that it might actually influence Georgia’s election results this fall. Since then, a couple of things have happened.

One is that I’ve gotten a couple of pretty thoughtful notes suggesting my projected body count was low. As a result I’ve been fiddling around with various chunks of data to see if I could come up with a credible way of fleshing out my last estimate.

The second thing that happened is that I got scooped.

Yesterday, Donald G. McNeil, Jr., a former New York Times health and science writer who now blogs about the pandemic, went up with a terrific post that basically did what I was working toward.

McNeil’s whole piece is well worth reading, but here are just a few of the money grafs:

“As of this week, about 1,800 Americans a day are dying of Covid; the C.D.C. expects that number to rise above 2,600.

“Virtually all are adults. If 95 percent were unvaccinated and we assume that 75 percent of those were Trump supporters, that’s 1,300 to 1,900 of his voters being subtracted from the rolls every single day.

Donald Trump lost Arizona by a mere 10,000 votes. He lost Georgia by 12,000, He lost Wisconsin by 21,000. He lost Nevada by 33,000.

Right now, about 60 Arizonans, 36 Georgians, 34 Wisconsinites and 14 Nevadans are dying of Covid each day. Seventy five percent of 95 percent of that would be minus 103 Trump voters per day — just in those four swing states. Week after week. That adds up.”

Obviously, these kinds of projections can get to be a little dicey. There are a lot of moving parts and the data is obviously very fluid. But your humble scribe here at TIGC would wager that data-crunchers and strategists in Democratic and GOP campaigns alike are paying attention to it. It’s going to be a significant part of this fall’s political story.

Stay tuned.

(c) Copyright Trouble in God’s Country 2022

TIGC tackles the big political question of 2022

This is the time of year when most journalists look back at the previous year and recap its major stories. Here at Trouble in God’s Country, I’ve decided to look to the future and take on the major question that will probably hang over Georgia politics for most of the rest of the year.

Specifically: Will Covid kill so many more Republicans than Democrats that it might actually influence the election results in November?

I know, I know. You’re thinking it’s impossible to know whether Covid victims voted red or blue. You’re probably also thinking the question is rude, insensitive and in poor taste. You may be right on both counts. But bear with me.

I took a first pass at this question back in September. At the time, I was looking at the laissez-faire approach Governor Brian Kemp was taking on Covid and linking that to the differing death and vaccination rates that were already taking shape between the state’s red and blue counties.

My thinking then was that the numbers were interesting but that the possibility that they might actually impact future election results was a little far-fetched.

Now, I think I can report that the possibility is a good bit less far-fetched.

First, one data point I used in that initial report probably understated the difference in the Covid death rates in red and blue Georgia. Back then — on September 10th — the Georgia Department of Public Health’s daily Covid report revealed that the Trump counties had suffered 10,545 deaths from the virus versus 9,468 for the Biden counties.

In that analysis, however, I ignored one column in the Georgia Department of Public Health’s daily reports: “Probable Deaths.” I did that in the interest of being cautious and conservative in the way I analyzed the data. I’ve since decided that was unnecessary and, frankly, wrong. Whatever the final cause of death is ruled to be, those “probable” Covid victims are still dead and, presumably, won’t be able to vote.

Add those “probables” to the tally and the body count in the GOP counties jumped, as of last September 10th, to 12,597 versus 10,361 for the 30 Democratic counties — a difference of 2,236. More interesting, I thought, but probably still not a big enough number to get worked up about.

So, what’s happened since then? Well, as of December 31st, the total Covid death toll in the Trump counties — for confirmed and probable deaths — was 17,119 versus 13,157 in the Biden counties, a difference of 3,962.

The bottom line arithmetic on this is that, for the 112 days between September 10th and the end of the year, the Republican counties, on average, lost an average of just over 40 people (virtually all of them voting age) to Covid versus just under 25 people in the Biden counties — a difference of 15.4 deaths per day.

Extrapolating from December 31st until the November 8th General Election would obviously be a risky exercise, but if — big if, I know — the current trend holds, the gap between the Republican and Democratic counties would swell to more than 8,700.

In a state where former President Trump got himself tape-recorded pleading with Georgia Secretary of State Brad Raffensperger to “find 11,780 votes” so he could reverse Biden’s Georgia victory, that’s probably a big enough number to merit a little attention.

And, yes, I know: I may be on shaky ground in suggesting that the geographic differences are a proxy for the political split. But at this point there’s enough data available that I’m comfortable doing just that: I’d wager the law of large numbers is kicking in and that, overall, the geographic and political splits are pretty close.

I’d bet that’s especially true once we factor in the vaccination differences. As of September 10th, the Democratic counties had already given two Covid shots to nearly 800,000 more of their residents than had the GOP counties. As of the end of the year, the vaccination advantage in the Biden counties had grown by another 60,000.

This picture comes into much sharper focus when you look at political universes that are overwhelmingly red or blue. Twenty-five largely rural or exurban counties gave Trump at least 80 percent of their 2020 vote; collectively they hit 83.6 percent for the incumbent president. As a point of comparison, urban DeKalb County gave Biden 84.1 percent of its vote.

This table summarizes the key data points.

With a much smaller population, the 25 Trump counties had nonetheless posted 1,129 more Covid deaths than DeKalb County at year’s end; indeed, the collective Covid death rate for those counties is substantially worse than Mississippi’s, which is currently the worst in the nation.

DeKalb, meanwhile, had fully vaccinated 52.6 percent of its population and gotten boosters in the arms of 19 percent. The 25 Trump counties lag badly in both categories.

Will these trends really ripple into Georgia’s political waters and influence the electoral tides this fall? We won’t know until the night of November 8th, but I think the numbers have gotten big enough that they’re worth watching.

And I’ll add this: If the former president has to come back to Georgia this winter in search of more supposedly missing votes, I’ll have a suggestion about where he should look. I’ll also offer one other piece of advice: bring shovels.

GSU Urban Studies Institute urban-rural symposium now available online

Back in November, Georgia State University’s Urban Studies Institute sponsored a day-long symposium on the state’s widening urban-rural divide, and the video is now available online here. Yours truly was the lead-off speaker, but don’t let that scare you off. There were a number of good presentations and a lot of useful information shared.

Georgia’s 40-year PCI rollercoaster ride

In the final 20 years of the last century, Georgia made remarkable economic progress on at least one front: the state’s per capita income (PCI) gained more than 10 percentage points against the national average and came within five points of that important benchmark. In the process, the state’s PCI rank among the 50 states and the District of Columbia climbed from a low of 41st to a high of 25th.

In the first two decades of this century, however, Georgia has surrendered nearly all those gains and fallen back into the nation’s bottom ranks for per capita income. Its 2020 rank was 37th. Only three states gained more ground against the national PCI average than Georgia between 1980 and 2000, and only four lost more ground between 2000 and 2020.

In the process, hundreds of thousands of Georgians were first lifted out of the bottom national quartile for per capita income but have since fallen back into it. As TIGC reported in its last post, Georgia finished 2020 with more of its citizens living in bottom-quartile PCI counties than any other state in the union, including states like Texas and Florida with significantly larger populations.

While that last post focused exclusively on new 2020 data released in November by the U.S. Bureau of Economic Analysis (BEA), this one takes a deeper look at a half-century’s worth of PCI data with an eye toward trying to answer a question we posed in the last post: Why is such a large portion of Georgians apparently stuck at the bottom of the nation’s income ladder?

The answer to that question remains elusive, but the lookback at the last 50 years reveals that Georgia has been on a PCI rollercoaster that is all but unique among the 50 states and D.C. Through the 1970s, Georgia’s PCI ranking was basically flat at between 38th and 40th place among the 50 states and D.C. But from 1983 through the end of the century, the state’s ranking climbed steadily, if unevenly, to a peak of 25th place in 1999 before plateauing at 26th place for the next three years. Since then, Georgia has dropped precipitously in the national PCI rankings. It bottomed out at 40th and 41st from 2008 through 2015 before rebounding to 37th by the end of 2020.

These are among the major findings and observations from an ongoing TIGC review of 50 years of personal income data produced by the U.S. Bureau of Economic Analysis (BEA).

And while identifying the exact causes of the state’s rollercoaster ride will require further research, it’s difficult to ignore the overlap between the rise and fall of the state’s PCI fortunes with the transition in the state’s political leadership: all the gains occurred under Democratic governors and all the losses followed under Republicans.

The most dramatic progress came under Governor Joe Frank Harris, who served from 1983 through 1990. During that period, Georgia moved up in the PCI rankings from 37th to 30th. The progress continued under Governor Zell Miller, who succeeded Harris and served the next eight years. After plateauing at 30th or 31st for three years, the state resumed its climb and reached 26th place by the time Miller left office at the end of 1998. Under Governor Roy Barnes, the state’s last Democratic governor, the state’s PCI ranking peaked at 25th in 1999 and then plateaued at 26th for the final three years of his one term in office.

Barnes lost his 2002 reelection bid to Sonny Perdue, who took office in January 2003 as the state’s first Republican governor in modern times and went on to handily win reelection in 2006. By the time he left office in 2010, Georgia’s national PCI ranking had plunged 15 spots to 41st, tying an all-time low for the last half-century.

Under Nathan Deal, Perdue’s successor and the state’s second GOP governor in more than a century, the state’s national PCI ranking floated along at 41st and 40th for the first five years of Deal’s two terms before rebounding slightly to 38th by the time he left office. Now two years into the term of the state’s third Republican leader in modern times, Governor Brian Kemp, the state stands 37th in the nation’s PCI ranking — the same ranking it had when Governor Harris took office.

Whether the various governors deserved all the credit or blame for the ups and downs in the PCI rankings is a matter for debate, but it seems a fair question. Were there changes in economic development, education or other policies and practices that drove the rankings and, more importantly, the personal fortunes of Georgians impacted by the 40-year seesaw effect? Or was it all just a huge coincidence?

As part of this analysis, I have used the BEA data to rank all the counties in the country by per capita income, sort them into national quartiles, and then pull the 159 Georgia counties out of the national list. As a result, I can determine how many Georgia counties — and Georgians — lived in each quartile in any given year. So far, I’ve done this part of the analysis for each year preceding a gubernatorial transition.

At the end of 1982, just before Harris took office, some 21.8 percent of the state’s 5.65 million residents — about 1.23 million people — lived in 91 counties in the bottom quartile of the nation’s PCI rankings. By the time Barnes left office in 20 years later, those numbers were down dramatically: less than 10 percent of the state’s 8.51 million citizens — fewer than 830,000 people — lived in 53 counties in that bottom national quartile.

When Perdue left office in 2010, the percentage of Georgians living in bottom quartile counties had exploded to more than 30.3 percent of the state’s estimated population of 9.7 million people — some 2.9 million people in 104 counties. At the end of 2020, the picture was no better: more than 3.05 million Georgians were living in 104 bottom quartile counties — the most of any state in the country, as TIGC reported in its last post.

To be clear, Georgia’s PCI has continued to rise throughout this period, but for the past 18 years it has lost ground against the national average. When Barnes left office at the end of 2002, the state’s average PCI of $30,133 was 94.6 percent of the national average of $31,859. When Perdue left office eight years later, the state’s average PCI stood at $34,830, but that was only 85.6 percent of the national average for 2002: $40,690.

The two graphs below should help tell this story. The first one illustrates the rise and fall of Georgia’s per capita income as a percentage of the national average. The straight blue line across the upper part of the graph represents the national average across the 40-year timescale. The orange line below it illustrates the rise and fall of Georgia’s per capita income as a percentage of that national average.

This second graph shows how Georgia’s actual per capita income tracks against the national average over the past 40 years. The gap narrowed, sometimes unevenly, through the 1980s and ’90s before beginning to widen at the turn of the century.

This is the second of at least three posts I’m developing based on the BEA’s latest economic data. In the next one, I’ll return to my usual focus on the state’s urban-rural divide, including trends in the different regions of the state.

(c) Copyright Trouble in God’s Country 2021

Rural Georgia leads race to the bottom in per capita income. The question is, why?

The week before Thanksgiving, I served as the lead-off speaker for a day-long symposium, sponsored by Georgia State University’s Urban Studies Institute, on Georgia’s urban-rural divide. About an hour before I started my presentation, the U.S. Bureau of Economic Analysis (BEA) put out its annual report on county-level per capita income. It’s a shame I couldn’t have gotten an advance look at the data; it would have provided a great addition to my presentation.

I’ve now spent two or three days rolling around in the data and can already see that I’ll be able to milk several solid posts out of the BEA spreadsheet. For starters, though, I’ll focus on Georgia’s at least mildly surprising showing at the bottom of the nation’s per capita income pile.

One useful thing about the BEA report is that it includes data on more than 3,100 counties and comparable governmental jurisdictions. That makes it possible to compare Georgia to its neighbors and, indeed, the entire country. It also makes it possible to document the extent of the divide between Georgia’s haves and have-nots.

The first unhappy headline out of this data dive is that Georgia counties occupy the bottom two places on the national list. Wheeler County finished 3,114th out of 3,114 counties with a 2020 PCI of $21,087, just behind Telfair County at 3,113th with a PCI of $22,644. As a frame of reference, those figures are less than one-fourth of Fulton County’s state-leading per capita income of $95,683 and about one-tenth the PCI of $220,645 in Teton County, Wyoming, which ranks No. 1 nationally.

Perhaps even more troubling, Georgia is home to 10 of the bottom 30 counties nationally. The only other states with more than two counties in the bottom 30 are Florida with six and South Dakota with four. Because Georgia has so many more counties than most states, it might be possible to argue that the number of counties on any such list isn’t all that important. So, let’s look at population.

Of the 10 states with counties in the Bottom 30, Georgia had a larger share of its population living in those counties than any other state except South Dakota, whose four counties in the Bottom 30 were made up largely or entirely of impoverished Indian reservations. As the table at right shows, some 1.2 percent of Georgia’s overall population resides in a Bottom 30 county; except for South Dakota, all the other states’ Bottom 30 populations were below one-half of one percent.

Still untroubled? Okay, let’s broaden the focus.

As I’ve already suggested, the BEA data allows you to sort and rank all 3,114 counties (and comparable jurisdictions) nationally. Having done that, I’ve also sliced the nation, and the state, into quartiles. Of Georgia’s 159 counties, 104 counties posted 2020 PCIs in the bottom national quartile.

Those 104 counties are home to 28.5 percent of Georgia’s 10.7 million residents — a higher percentage of people living in the bottom quartile than any of its adjoining states except Alabama, where the number is 29.6 percent. This table shows the total populations and quartile splits for Georgia and all its contiguous neighbors.

I’ll have more to say about this in a subsequent post, but one initial takeaway (in my view) is that it’s pretty good illustration of extent of the chasm between Georgia’s haves and have-nots.

To widen the lens even further, Georgia has more people living in the bottom quartile than any other state in the nation, including Texas, Florida and all the other states with larger populations. Some 3.05 million Georgians live in the bottom PCI quartile.

Texas, with nearly three times Georgia’s population, has only 2.75 million residents living in the bottom quartile. In Florida, which has double Georgia’s population, the number of residents in the bottom quartile is 2.01 million. North Carolina, with essentially the same population as Georgia, has nearly 1.3 million fewer people in its bottom tier counties.

Of the 779 counties in TIGC’s bottom quartile, 104 are in Georgia; only four other states — Arkansas (54 counties), Kentucky (65), Mississippi (55) and Missouri (54) — had more than 50 counties in the bottom quartile.

That rural Georgia’s 2020 per capita income is so low is not in and of itself all that surprising. But that the state performs so much worse than neighboring states like Florida and North Carolina is frankly more than a little disconcerting and a bit of a mystery. How those states have been able to do a better job of moving their populations out of the bottom PCI tier and up the economic ladder is a question that needs to be answered.

Watch this space.

The interactive map below highlights Georgia’s 159 counties based on their National PCI Quartile. The lighter the shade, the higher the quartile.

The interactive table below shows 2020 per capita income data for all 159 Georgia counties, along with their state and national rank and the national quartile into which each county falls.

Gauging the gap in educational attainment in Georgia’s urban-rural divide

This post is the second in a series I’m developing as I rework my TIGC presentation for a mid-November symposium on Georgia’s urban-rural divide. In the first of this series, I took a look at the widening gap in premature death rates between Metro Atlanta and Georgia’s Other 147 Counties. Today’s topic is educational attainment.

I’ve covered some of this before but I don’t feel like I’ve ever stitched together the whole story, which I’ve come to view as the single most important driver in the division between my TIGC 12-county Metro Atlanta region and the Other 147 Counties.

I’ll start with this graph, which I’ve just developed.

What this shows is the number of college graduates by region over the last half-century, as documented by the U.S. Census Bureau and compiled by the U.S. Department of Agriculture’s Economic Research Service (ERS).

The first takeaway from this data is that the Metro Atlanta/Other 147 divide is a relatively recent development. In 1970, according to the Census data, there were fewer than a quarter of a million college graduates in the entire state, and slightly more than half of them lived outside today’s Metro Atlanta region (much of which was also rural back then).

That balance had tipped in favor of my TIGC 12-county Metro Atlanta region by 1980, but just barely. It was only as the state moved into the 1980s that the gap in the number of college graduates really began to widen. Based on the Census Bureau’s American Community Survey (ACS) data for the five-year period 2015-2019, more than 63 percent of the state’s nearly 2.2 million college graduates reside in the 12 Metro Atlanta counties.

While I regard the educational attainment data as the most important metric in this area, the truth is it’s a lagging indicator — and there’s a leading indicator that forecasts even more bad news. Over time, I’ve stitched together 14 years of fall enrollment data from the University System of Georgia (USG). It’s reflected in this graph.

What this shows is that for the first four years of my study period, the Other 147 Counties were still sending more freshmen to USG colleges than the 12 Metro Atlanta counties. Look at the orange line and you’ll see that the Other 147 Counties peaked in 2009 when they sent a little more than 23,000 freshmen to USG institutions and then started a pretty dramatic six-year slide.

This is arguably the most significant of the Great Recession “aftershocks” I’ve written about in other pieces. Metro Atlanta continued to send growing numbers of freshmen to USG institutions for another two years before being hit by a comparable aftershock and suffering a two-year tumble before beginning to recover.

As the chart shows, Metro Atlanta continued to send more freshmen to USG institutions for the remainder of the current decade. One takeaway from the data is that — between 2006 and 2020 — the Other 147 Counties went from sending just over 15 percent more freshman to USG colleges and universities than Metro Atlanta to sending nearly 16 percent fewer — a swing of more than 30 points in 14 years.

But there’s a hint of good news at the end of the decade. After an odd, statewide downturn in freshman enrollment in 2019, both Metro Atlanta and the Other 147 Counties posted large and highly comparable increases in the fall of 2020. The Other 147 Counties may not have gained any ground on Metro Atlanta, but at least they didn’t lose any more.

Overall, the 12 Metro Atlanta counties sent about 54 percent of the total Fall 2020 in-state enrollees to the University System’s 28 institutions.

But the picture is much starker when you look at the four major research universities — Augusta University, Georgia State University, Georgia Tech and the University of Georgia. At those four institutions combined, just over 70 percent of the Fall 2020 in-state enrollees came from the 12 Metro Atlanta counties.

At Georgia Tech, the percentage of Fall 2020 enrollees from Metro Atlanta hit 74 percent, and the 83 counties highlighted on the map at right didn’t send a single freshman to Tech that year. At the University of Georgia in Athens, the percentage of Fall 2020 in-state enrollees from Metro Atlanta topped 62 percent, and it pulled students from more of the state; only 20 counties failed to send a single freshman to UGA that Fall.

That this concentration of educational muscle in Metro Atlanta holds economic implications for the entire state should be obvious. But it also foreshadows major cultural and political shifts, which I’ll get into in a future post.

(c) Trouble in God’s Country 2021

Gauging the Gap between Metro Atlanta and the rest of the state on premature death rates

I’ve spent a good part of the summer taking a deep dive into several pots of economic and education data with an eye toward fleshing out a couple of chapters in the book version of TIGC. As a result, I’ve been neglecting the blog.

Then, in mid-August, I got a call asking me to make a presentation to this year’s opening session of the Georgia House Rural Development Council (HRDC) on September 1st. Over the years, I’ve given more than 50 presentations on my TIGC work, and generally I’ll update or amend my PowerPoint with whatever new data I’ve been working on. The result has been that the presentation has devolved, in my view, into a bit of mishmash.

Now I’ve been asked to speak in mid-November to a symposium on the state’s urban-rural divide and, with about six weeks to work, I’ve decided to do a major overhaul on my presentation — and to write an occasional post as I do so.

This is the first of those posts. My plan right now is to create a series of slides built around the theme of “gauging the gap” between Metro Atlanta and the rest of the state.

I started this process yesterday by updating my research and analysis of the state’s county-level premature death rates. I use premature death rates — formally known as “Years of Potential Life Lost before Age 75,” or YPLL 75 — as a proxy for community health status.

I’m sure I can get a fair argument that this approach is insufficient or an oversimplification, but, as an old public health friend once explained to me, “Premature death is the Dow Jones Industrial Average of population health.

“If you want to look at one number and get a feel for a community’s health status, look at premature death,” he said. Backing up that assessment is the fact that County Health Rankings & Roadmaps uses premature death rates as a key component in evaluating health outcomes for more than 3,000 U.S. counties.

Another reason I like using YPLL 75 rates is that they’re simple to calculate. Working with age data mined from death certificates submitted to the state, the Georgia Department of Public Health (DPH) keeps track of all the years of potential life lost before age 75 for all the counties in the state on an annual basis and posts this data to its excellent public health database.

If a 50-year-old Macon man dies, he contributes 25 years to Bibb County’s bucket of years of potential life lost before age 75. A 74-year-old woman up in Clayton, Ga.? That’s a year on Rabun County’s tally. A six-month-old infant in Decatur? That goes down as 74-and-a-half years for DeKalb County. And so on.

The DPH database also includes annual population estimates generated by the U.S. Census Bureau, and that’s really all you need to calculate YPLL 75 (or Premature Death) Rates. The formula is:

(YPLL 75/YPLL 75 Population) X 100,000 = YPLL 75 Rate

As a couple of examples, Forsyth County had Georgia’s best YPLL 75 rate in 2020 and Clay County had the worst. In fast-growing Forsyth County, the total number of years of potential life lost before age 75 was 9,970. Divide that by the county’s YPLL 75 population (the number of people 75 or younger) of 238,043 and multiply the result by 100,000 and you get a YPLL 75 rate of 4,188.3. I haven’t double-checked this, but I’ll guarantee you that premature death rate is among the best in the country.

Tiny, poverty-stricken Clay County, located hard on the Alabama line in southwest Georgia, turned in true third-world numbers for 2020 — a YPLL 75 Rate of 22,180.6 (basically five times worse than Forsyth County).

(Its data also generated something of a riddle. With a YPLL 75 population of just over 2,500 people, its YPLL 75 total more than doubled from 2019 to 2020 — jumping from 253 to 562.5.

(When I saw that big increase in years of potential life lost, my first reaction was that Clay County had gotten clobbered by Covid-19. Adding to that line of speculation was the fact that the total number of deaths in the county actually dropped to 45 in 2020 from 57 in 2019, and I immediately figured the virus had claimed a number of relatively young people in the county. That wasn’t a crazy notion. Clay County has little if any healthcare infrastructure and isn’t that far from Albany, which at one point last year was the global ground zero for the virus.

(But no. A quick check of DPH’s Covid-19 data reveals that Clay County suffered only four Covid-19 deaths in 2020 (and none so far this year) — and only two of those victims were under the age of 75. Together they contributed only 28 years to the county’s total YPLL 75 pot of 562.5 years.

(So what was killing younger people in Clay County? A review of the cause-of-death data in DPH’s OASIS database didn’t turn up many significant differences over 2019 — until I got down to the External Causes category. In 2019, the county reported no suicides or poisoning deaths. In 2020, there were three poisoning deaths and four suicides, all under the age of 75. Indeed, four were under 50; that’s at least 100 years of potential life lost before the age of 75 right there.

(Were the suicides or poisoning deaths some sort of collateral damage from Covid-19? I don’t know, but I’m curious enough to check and see if similar patterns turn up other counties.)

I have, however, meandered. To tie off this incredibly laborious math explanation I started about a half-dozen paragraphs up, here are the numbers for Georgia’s best and worst YPLL 75 counties.

Now back to my original objective — a comparison of premature death rates between my TIGC 12-county Metro Atlanta region and the state’s other 147 counties. (I should probably acknowledge that I’ve waffled on the question of how best to analyze the data I’ve collected. I’ve looked at using a basic North Georgia/South Georgia split as well the five regions I created early on in this process. At times I’ve analyzed county data based on their political leanings. For the purposes of this upcoming presentation (and these blog posts) I’ve decided to try to keep it simple and just compare Metro Atlanta to the rest of the state.)

This chart shows the YPLL 75 performance for TIGC’s 12-county Metro Atlanta region versus the state’s Other 147 Counties from 1994 (the earliest year for which DPH has data) through 2020. With YPLL 75 Rates, the lower the number, the better, so Metro Atlanta’s blue line at the bottom of this chart signifies the better peformance.

There are several important takeaways from this chart. The first is that the gap between Metro Atlanta the rest of Georgia has been widening throughout the 26-year period. In 1994, Metro Atlanta had what might be described as a 15.8 percent premature death “advantage” over the rest of the state; by 2020, that “advantage” was up to 41 percent. As I’ll get into in future posts, these differences have implications both for healthcare costs and for economic productivity.

A second takeaway has to do with what began happening in about 2011. Up until then, both Metro Atlanta and the rest of the state had generally been gaining ground, if somewhat unevenly. But that pretty much came to a halt in 2010, and in 2011 the entire state’s YPLL 75 Rate ticked up (that is, got worse). Metro Atlanta regained some ground in 2012, but then saw its YPLL 75 Rate deteriorate a little or flatline every year for the next five years before finally posting some improvement in 2018 and 2019.

The Other 147 Counties fared worse over that same period. If Metro Atlanta’s YPLL 75 Rate performance produced a five-year plateau from 2013 through 2017, the rest of the state generated a couple of uphill climbs before finally finding a downhill slope and producing three straight years of (slightly) improving numbers. But then came 2020 and Covid-19 — and the YPLL 75 Rates took off like moonshots.

Back to the takeaways. Back in 2011 and ’12, as I was starting work on Trouble in God’s Country and just beginning to dig into the premature death data, I remember noticing the sudden uptick in YPLL 75 Rates but didn’t know what to make of them. Over time — and in combination with similar adverse trends in other data — I’ve come to believe that what I was seeing amounted to aftershocks from the Great Recession. It showed up in economic, education and even birth and death data, and in every case, the Other 147 Counties suffered a harder blow than Metro Atlanta.

What to make, then, of the modest gains that began showing up in 2017 and ’18? As it happens, I’m seeing similar patterns in other datasets. I discussed those in my HRDC presentation a few weeks ago, but hadn’t yet started this review of YPLL 75 Rates. In isolation, I wouldn’t attach much significance to any one instance of improvement, but in combination I think they signal that my Great Recession aftershocks may finally be playing themselves out.

Now, though, we have Covid-19, and its impact across a range of societal sectors — health, economics, education — is already profound. The question is whether it will generate its own wave of aftershocks that will be with us for years to come.

TIGC takes a fresh look at the political arithmetic of Covid-19 and poses a rude question

With less than 14 months to go before Georgia’s 2022 statewide elections, TIGC has decided it’s time to tackle the obvious political question that other observers and commentators are too polite and high-minded to address, namely: Are Republicans killing their own voters?

This is admittedly tough to prove. But it’s difficult — nay, impossible — to compare the state’s Covid-19 performance with recent election results and not at least wonder. As of this past Friday, the 129 counties that sided with Donald Trump in the 2020 presidential election had significantly higher Covid-19 case rates and death rates — and much lower vaccination rates — than the 30 counties that went for Joe Biden.

Some raw numbers: Covid-19 data published Friday, September 10th, by the Georgia Department of Public Health (DPH) tells us that the Trump counties had suffered 1,077 more deaths than the Biden counties while vaccinating nearly 800,000 fewer people. Perhaps even worse for the Trump counties, their combined 14-day case rate — a measure of current rather than long-term trends — is a solid 41 percent higher than the rate in the Biden counties.

Of course, these numbers alone don’t prove anything. The virus is, as far as we know, politically agnostic, and neither death certificates nor vaccination records list political party preference. Further, it’s probably mathematically possible that an actual body count (an audit, perhaps, that compares death certificates with primary voting histories) would tell a different story. But a look at various bits and pieces of anecdotal data makes it difficult to conclude that Democrats are suffering a bigger Covid-19 hit.

Take, for instance, Brantley County. Located in deep southeast Georgia, Brantley gave Trump his biggest Georgia margin — 90.9 percent of the vote — but, as of Friday’s DPH report, it had the fourth-worst vaccination rate in the state at 20.8 percent. This could be purely coincidental, but your humble scribe here at TIGC is skeptical of that. Of the 51 Covid-19 deaths Brantley had reported by this past Friday, 43 were white, and all but three were 50 or older.

It is, of course, possible to find counternarratives in county-specific data. As an example, dirt-poor and heavily-black Hancock County, which gave Biden one of his biggest margins (72.1 percent) also had the state’s worst Covid-19 death rate as of Friday. That said, Hancock Countians seem to be taking the hint: 42.2 percent had been fully vaccinated as of last Friday, according to DPH data, one of the state’s highest rates, especially among rural counties.

Indeed, any attempt to find county-level correlations between Trump-Biden vote splits and, say, case or death rates is doomed to failure — thanks to a host of other variables that come into play, including race, poverty and educational levels, probably among others.

But at a macro level, fairly clear patterns begin to emerge, as this table shows.

Against the backdrop of those kinds of numbers, you’d think Georgia’s GOP leaders would be doing more to promote vaccinations and other Covid-19 mitigation measures, including masking. While Governor Brian Kemp, a Republican, has gotten vaccinated and publicly encouraged others to do so, it seems fair to say his support for anti-Covid policies has been less than full-throated. He has overridden attempts by local governments to impose masking mandates and other mitigation measures, and he’s up on Twitter today with (so far) three tweets attacking President Biden’s plan to require all businesses with more than 100 employees to ensure they’re vaccinated or at least tested weekly for the virus.

Kemp’s lack of enthusiasm on the anti-Covid front may have trickled down and infected the state’s bureaucracy. DPH has made a good bit of Covid material available to the media on its website and produced at least one television ad earlier this year, but it’s not clear how much play that ad got — or how effective it was. It does seem fair to suggest that the state-level effort left a vacuum that at least some local governments and health departments have felt compelled to try to fill.

As an example, Gwinnett County earlier this year launched a campaign built around “listening to moms” to encourage Covid-19 mitigation measures, including vaccinations, and has reportedly spent more than a half-million dollars on the campaign. It may be getting a decent return on that investment. While Gwinnett went through a Covid “hot spot” phase several months ago and has one of the state’s higher overall case rates, its Covid death rate is one of the lowest in the state and its vaccination rate, at 48.5 percent, ranked 6th best in the state as of last Friday. Its also the largest of the Metro Atlanta counties that was solidly Republican a decade ago but has shifted from red to blue since then: it went nearly 60-40 for Biden in the 2020 election.

TIGC won’t attempt to use these numbers to extrapolate over the next 14 months and estimate an impact on the 2022 elections, but it’s difficult to imagine that any of the state’s Republican politicians or operatives would find much good news — or comfort — in them. If the current 1,077-death difference between the Trump and Biden counties just happens to parallel the difference in voters lost by each party to Covid so far, that alone probably won’t spell the difference in next year’s elections.

But then you have to figure out how to factor in the difference in vaccination rates and recent Delta variant case rates — and layer that onto that the fundamental health differences between the state’s overwhelmingly rural Republican areas and its largely Democratic urban climes, including, specifically, higher rates of lethal comorbidities such as obesity and diabetes. Will those conditions, in combination with Covid-19, compound the premature death rates that are already higher in predominantly Republican rural Georgia?

Governor Kemp’s management of the state’s Covid plague may not quite rise to the level of criminal negligence or manslaughter. But it might yet turn out to be political suicide.

(Couple of notes on my methodology in this piece. In crunching the presidential votes, I’ve ignored Libertarian votes, as I usually do. In analyzing various pieces of DPH data, I’ve found that different units of the department use different population estimates to calculate the various case, death and vaccination rates. The vaccination rates published by DPH are pegged to 2018 population estimates, according to its own “Data Descriptions” published with the daily reports. It’s not clear to me what population estimates DPH uses to calculate daily case and death rates; the numbers don’t quite match any of the annual estimates I can find. In the interest of consistency, I have used 2020 population estimates pulled directly from the Department’s OASIS database (I haven’t had time to get into the actual county-level census counts yet). My use of the 2020 estimates produces slightly different case, death and vaccination rates than those shown on the various DPH reports. Also, many thanks to several Facebook friends who helped me crowdsource information about state and local Covid communications programs, especially old friend Terry L. Wells.)