The Continental Congress renamed their new nation the United States of America, from the previously used “United Colonies.”
On September 9, 1933, WSB Radio in Atlanta was upgraded to broadcasting via 50,000 watt transmitter. The first broadcast included Will Rogers and a letter from President Roosevelt.
On September 9, 1939, an audience at the Fox Theater in Riverside, California watched a preview of Gone With the Wind.
On September 9, 1954, Marvin Griffin won the Democratic Primary election over Melvin Thompson.
Happy 72d birthday to former Congressman John Linder. Linder served in the State House from 1974-1980 and 1982-90. In 1990 he ran unsuccessfully for Congress against incumbent Democrat Ben Jones; in 1992, after redistricting, Linder was elected to Congress from the 7th District and served until his retirement after the 2010 election.
Georgia Politics, Campaigns, and Elections
Today’s edition will be very short, but dense, as we have some numbers to crunch. From 6 AM to 9 AM, I’ll be sitting in with Tim Bryant on Georgia’s Morning News, on 1340 AM WGAU. If you’re in the Athens area, please listen in on 1340 AM, or click here to tune in via the magic of the internet. That link opens a page and you’ll need to click the “Listen Live” button at upper center.
Yesterday, the New York Times, YouGov, and CBS News released their latest online polling on the Senate race. Here are their numbers:
David Perdue (R)……..44%
Michelle Nunn (D)…….38%
Amanda Swafford (L)……7%
The AJC reported different numbers, showing Perdue 47, Nunn 41. I think they’re reporting the NYT’s prediction, which includes the second poll, whereas the numbers above are the second poll alone.
The most important number is that the NYT weighted African-Americans to 25%, while GaPundit’s latest poll weighted African-Americans to 30%. I’m right, they’re wrong, here’s why.
In 2010, African-American turnout was 28.26% of the electorate, in 2012 it climbed to 29.89%. Those are actual turnout numbers as reported by the Georgia Secretary of State’s office. Indisputable.
Yesterday, Michelle Obama spoke to a Democratic event in Atlanta, and according the AJC Insider:
If Democratic voter rolls and turnout increase just 3 percent — which amounts to 50 more voters at each precinct turning out than did in 2010 — Nunn and Carter can be elected, the first lady said.
Overall 2010 turnout was roughly 3.9 million voters and 3 percent of that is 117,000 votes.
So, according to Michelle Obama, 3 percent of the vote wins the race this year. If we take the 2010 turnout percentage of 28.26 and assume that half of the 3 percent Mrs. Obama refers to will come from African-American voters, we can say that if 30% of voters are African-American in November, the Democrats will be halfway home.
I don’t think there can be any question that 30% African-American electorate is the goal of the Democrats this year, and it is achievable.
In the 2012 general election, which saw President Barack Obama re-elected – though he fell short in Georgia — DeKalb County churned out 307,474 ballots.
But in the 2010 gubernatorial contest, when former Gov. Roy Barnes, a Democrat, faced Republican Nathan Deal, the DeKalb vote amounted only to 208,732. The 100,000-vote difference in DeKalb is significant, given that Barnes lost by just over 250,000 votes.
So the majority of the gain in African-American voter turnout could come from DeKalb County alone. I think that’s unlikely, but if DeKalb contributes significantly more votes than in 2010, that’s a good start on gaining that 117,000 needed votes for the Democrats to win.
Finally, I did this: I re-weighted the NYT poll to see what would happen if they used the same data but boosted African-American turnout from their 25% to the 30% I’m predicting. That change alone, without assuming any change in the White and Other vote closes the gap from Perdue +6 points to Perdue +1.1 points.
After reweighting the NYT results to 30% African-American turnout, the race looks like this:
David Perdue (R)………..41%
Michelle Nunn (D)………40%
Amanda Swafford (L)……3%
That’s Perdue +1 point. My poll last week showed Nunn +1.6 points for a difference between the NYT (reweighted) and mine of 2.6 points, well within the margin of error.
In summary, the two polls are substantially identical with respect to the positions of the leading candidates, and nearly all the difference can be explained as a function of different predictions of the African-American percentage of the electorate. (NYT also show a higher percentage of undecideds, and I suspect the difference is a result of differing survey sampling and screening techniques).
That demonstrates how important correctly estimating African-American turnout is in polling accuracy and how important actual turnout in November is. (It’s a complicated procedure and I’ll post the Excel Spreadsheet later today after the radio show.)
It’s closer than we thought it was, and we have to get to work.