Biden Picks Chief of Staff

Joe Biden is operating under the understanding that he is going to become president on Jan. 20, something that he is absolutely right about. And that means that Biden needs to start putting an administration together. On Wednesday, he took one of the first big steps in that direction, and named Ron Klain as his White House chief of staff. The chief of staff does not need Senate confirmation.

The selection of Klain was not much of a surprise. He and Biden have known each other since the 1980s, when Klain served in Biden's Senate office. He also served as Biden's chief of staff while Biden was VP. Klain is popular with both progressive and moderate Democrats, is also able to reach across the aisle and work with Republicans, and has experience dealing with public health and economic crises. Klain ruffled the feathers of some folks in Biden's orbit in 2016, when he guessed Biden would not run and jumped on board with the Hillary Clinton campaign. However, that's all water under the bridge now, and he served as a key adviser to the Biden campaign this year. It's a tough job he's signing up for; in the last half-century, only H.R. Haldeman (Nixon), James Baker (Reagan), Andrew Card (George W. Bush), and Denis McDonough (Obama) have managed to hold on for four full years.

The Biden campaign thought about announcing several key picks yesterday, with an emphasis on the diversity of the incoming administration, but they backed off and just announced Klain. This is surely a wise choice; if Biden announces another member of his administration every day or two, then pretty much every news cycle will have a reminder of his victory, and thus pushback against Donald Trump's narrative of the election. Undoubtedly, more announcements will come later this week. (Z)

Republicans Win in Alaska

It took a long time to get results from Alaska. After all, those sled dogs can only mush so fast. However, about three-quarters of the vote has now been counted. And the results are lopsided enough that it's possible to call the state for the Republicans, with Donald Trump claiming the three electoral votes, and Sen. Dan Sullivan (R) and Rep. Don Young (R) winning reelection.

Inasmuch as it's only three electoral votes, and inasmuch as Young had won 24 consecutive elections before this, two-thirds of those results did not generate much drama. However, with challenger Al Gross running an effective and well-funded Senate campaign, many Democrats hoped that he might just pull off a miracle. Not so much; he's down 57.9% to 37.1%, so even if the remaining (all-mail-in) ballots break pretty heavily for him, he's still going to take a drubbing. And with 33 of 35 Senate races now called, it means that the fate of the Senate (at least for two years) will be decided by the two runoffs in Georgia. (Z)

Exit Polls

The results are largely in now, so now it is time to try to figure out what happened. In this item we look at the exit polls; in the next one we look at the pre-election polls. That said, an exit poll is still a poll and may be subject to the same sampling problems as the pre-election polls. The exit polls are sponsored by ABC, CBS, NBC, and CNN and are conducted by Edison Research. Edison hires thousands of workers to stand at the exits of polling places around the country and stop one out of every N voters as they leave to ask them how they voted, as well as some demographic questions. Some of them refuse and some might lie. In addition, since over 100 million people voted early this year, Edison also called thousands of people and asked them: "Did you vote early, either by mail or in person?" If the answer was yes, they continued with the questions; otherwise, they thanked the respondent and hung up.

The data from the exit polls are still being crunched and may change, so these results are preliminary. These numbers are from The New York Times. Here are the answers to some of the basic questions by demographic group. The numbers in parentheses show what fraction that group is of the electorate.

Demographic (% of electorate)	Biden %	Trump %	Winner's margin
Men (48%)	45%	53%	Trump +8%
Women (52%)	57%	42%	Biden +15%
Whites (67%)	41%	58%	Trump +17%
Blacks (13%)	87%	12%	Biden +75%
Latinos (13%)	65%	32%	Biden +33%
Asians (4%)	61%	34%	Biden +27%
Age 18-29 (17%)	60%	36%	Biden +24%
Age 30-44 (23%)	52%	46%	Biden +6%
Age 45-64 (38%)	49%	50%	Trump +1%
Age 65+ (22%)	47%	52%	Trump +5%
College graduate (41%)	55%	43%	Biden +12%
No college degree (59%)	48%	50%	Trump +2%
Evangelical (28%)	24%	76%	Trump +52%
Not evangelical (72%)	62%	36%	Biden +26%
Income < $50,000 (35%)	55%	44%	Biden +11%
Income $50K - %100K (39%)	57%	42%	Biden +15%
Income > $100K (26%)	42%	54%	Trump +12%
Served in military (15%)	44%	54%	Trump +10%
Never served in military (85%)	53%	45%	Biden +8%
First time voters (14%)	64%	32%	Biden +32%
Voted before (86%)	49%	49%	Even
Political liberals (24%)	89%	10%	Biden +79%
Political moderates (38%)	64%	34%	Biden +30%
Political conservatives (38%)	14%	85%	Trump +71%
Democrats (37%)	94%	5%	Biden +89%
Independents (26%)	54%	41%	Biden +13%
Republicans (36%)	6%	94%	Trump +88%
Live in city of >50K (29%)	60%	38%	Biden +22%
Suburb (51%)	50%	48%	Biden +2%
Small city or rural (19%)	42%	57%	Trump +15%

For example, women made up 52% of all voters and they went for Biden 57% to 42%, a 15-point gap. So what's the take? Trump won old, white noncollege men and Biden won young ethnic college-educated women. This is not a huge surprise, of course. Democrats are banking on the "rising electorate," but the reality is that white men, especially noncollege white men, are still a potent force. That is gradually changing, but the change is slow.

The Brookings Institution took a look at how the demographics have changed since 2016. They grabbed the data at a different moment than the Times so the (evolving) numbers are not quite the same (in part due to the corrections needed in the telephone sample of the 100 million early voters, which is more like a regular poll than an exit poll). Also note that the 2016 polls were not originally weighted by education but they have been retroactively corrected for that. This adjustment is a tricky business and certainly introduces some error. Still, even a first cut like this is worth something. Here is Brookings' graph of change by race:

First the bad news for Biden. He didn't do as well as Hillary Clinton with Black, Latino or Asian voters. So much for the rising electorate story. Now the good news. Whites did not find him as offensive as they did Hillary Clinton. Given the dominance of whites in the electorate, much of Biden's win can be ascribed to not disgusting whites as much as Clinton did rather than getting extra love from minorities. If this continues to hold when the final numbers are in, it has implications for 2022 and 2024. It says that Democrats should focus more on winning whites, since that seems to be doable, rather than on winning minorities. However, this has to be taken with a bucket of salt because there is also a supposition (more on this tomorrow) that macho young Black men and young Latino men liked Trump due to his aggressive take-no-prisoners style and possibly their envy of him being able to grab women by the you-know-what and get away with that. A Tucker Carlson, Tom Cotton, or Josh Hawley doesn't have that at all, and Nikki Haley certainly doesn't have that (although being a minority herself might compensate for the style issue).

Now let's break down white voters by education and gender:

Not surprisingly, Hillary Clinton did well with white women with a college education. That was her base. Biden barely lost them. On the other hand, Biden more than made up for that by winning college-educated men, a group that didn't like Clinton at all. Biden also did better with noncollege men and women. Some of this is likely to do with Clinton's style, which many people felt was arrogant and entitled. Men didn't like this at all and neither did noncollege women. It is hard to project this into the future, though, because the next time the Democrats run a woman, she might not be so grating. Veep-elect Kamala Harris has 4 years to take this lesson to heart. She will have to project leadership without projecting entitlement and arrogance. It's possible to be humble about the voters giving you their trust and still be strong, but it may require walking a line finer than any that a male politician ever has to walk.

The third Brookings graph slices the comparison between 2016 and 2020 by age:

These results are surprising. Old Joe Biden did much better than Clinton with young voters. It's also possible that age really wasn't the issue, though. It could be that young voters so hated Donald Trump that they would have voted for a yellow dog had the Democrats nominated one. Biden improved a little bit with seniors, but not as much as he did with younger voters. Maybe the seniors know how hard it is to function at 77 and thought he wasn't fully up to the toughest job in the world. Who knows?

If you want yet another take, here is CNN's story on the exit polls. The Hill summarized the exit polls with these takeaways:

• Men moved away from Trump as much as women did
• Independents moved heavily toward Biden
• Trump, much to the Democrats' dismay, did better than Clinton with minorities
• White evangelicals stuck with their Christ-like hero, St. Donald
• Trump won the "a plague on both their houses" voters in 2016, but that group was only 1/5 as big this time

Undoubtedly more analyses will be forthcoming as statisticians dive into the details in the coming weeks and months, but the main points will probably survive. (V)

What's Going on with the Polls?

Nate Cohn, The New York Times' replacement for Nate Silver (in physics this would be called the Law of Conservation of Nates), has been looking at the election data and has a couple of ideas of why the polls were so far off. He notes that doing a post-mortem now is like asking the coroner for the cause of death while the body is still at the crime scene, since the numbers aren't final yet. Still, here is his first take.

• The president hurt the polls: Donald Trump attacked the media and the polls mercilessly for years. There probably weren't a lot of "shy" Trump voters in 2016, but there might have been more in 2020 due to his supporters being instructed to distrust the media and pollsters. However, this probably doesn't explain everything. There has to be more. One very real possibility is that Trump convinced many of his supporters to refuse to talk to pollsters at all. This is different from lying to them. When a pollster calls and identifies as such, the Trump voter just hangs up the phone. With response rates below 5%, pollsters have no way of distinguishing between busy mothers trying to put their kids to bed and Trump-loving single women with plenty of free time who simply don't trust the pollsters.
  
  
• The resistance hurt the polls: This one is the inverse of the "shy" Trump voters and Trump refuseniks. With many Democrats livid with Trump and actively working to defeat him, when a pollster calls one of them, she instantly mutes Rachel Maddow and very carefully answers every question to make sure her response is counted. This raises the response rate among Democrats at the same moment Trump has reduced it among Republicans. Social science research has shown that Democrats tend to trust strangers and Republicans tend not to trust strangers. This is known as "social trust" and the effect amplifies the differential response rate. This explanation could really be getting at the problem. Unfortunately, it is hard to fix.
  
  
• The turnout hurt the polls: The U.S. ranks 30th out of 35 developed countries in voter turnout. This year there was pretty good turnout. A mere 75 million eligible voters didn't bother to vote, even though it was easier than ever with no-excuse absentee voting possible in most states. Even among registered voters, not everyone turns out, so pollsters have screens and models for likely voters. If Republicans turned out at a higher rate than pollsters expected, that could explain some of the error. In Florida, for example, registered Democrats outnumber registered Republicans by 1.5%. Yet among actual voters there, Republicans outnumbered Democrats by 2%, a shift of 3.5%, which was slightly larger than Trump's 3.3% margin in Florida. Such a big shift could account for much of the polling error in Florida. The same thing may have happened in other states, but the data aren't in elsewhere yet.
  
  
• The pandemic hurt the polls: The Times polls of Oct. 2019, 13 months before the election, were surprisingly accurate. The polls in Feb. 2020 showed the same thing. After the pandemic hit, Biden started doing better. This could have been due to people working from home and thus being easier to reach. But since Democrats have more social trust, they were willing to be polled whereas home-bound Republicans were not. Or, it could also be that the folks who voted Democratic this time (e.g. people with college educations) were more likely to have the sorts of jobs that allow them to work from home. Bankers can talk to their customers from home; bus drivers can't. Either way, pollsters may have talked to a disproportionate number of Biden voters without realizing it. We know there is some truth here because pollsters were giddy with increased response rates once the pandemic hit. What they didn't know is that they probably weren't getting anything close to a representative sample.
  
  
• The Latino vote hurt the polls: One state where the polls were especially bad was Florida. A fair amount of the error came from Trump's massive and unexpected gains among Latinos in Miami-Dade county. Clinton won the heavily Latino county by 29 points whereas Biden took it by a mere 7 points. In fact, Trump did well among Latinos nationwide, from Pennsylvania to the Rio Grande Valley. Why Trump did so well among Latinos is another story, but that will have to wait to be told some other day. Like tomorrow, so watch for an item on this subject.

Cohn did an interview with his friend Isaac Chotiner of the New Yorker. Here is a brief summary of their Q & A:

• Q: What surprised you the most about the results?:
  A: The swing of Latinos to Trump and Biden's not doing much better than Clinton among white voters in the Midwest while the polls said he would.
  
  
• Q: How big was the swing among Latinos?:
  A: My gut sense is that it is in the double digits.
  
  
• Q: How come?:
  A: The election wasn't about immigration as it was in 2016. Trump wasn't hitting on Latinos, as he did in 2016, so with that factor removed, working-class Latinos voted more like working-class whites and wanted to send the country a message.
  
  
• Q: Was it only Latino men or also Latinas?:
  A: The Latino gender gap was enormous.
  
  
• Q: In which election did immigration play the biggest role?:
  A: 2018, when Trump put immigrant kids in cages and talked endlessly about the caravan from Central America. And the Republicans were wiped out. Now they know that talking about immigration is a two-edged sword.
  
  
• Q: Why did downballot Republicans do better than Trump?:
  A: I don't know yet.
  
  
• Q: This time the state polls were weighted by education and were still way off. Why?:
  A: First, the nonresponse bias correlated very strongly with partisan preference, maybe for the first time. Second, Democrats were much more motivated to talk to the pollsters, increasing the effect. Third, pollsters assume that high turnout helps the Democrats, but maybe it doesn't, so their models of the electorate were all wrong.
  
  
• Q: What about the virus?:
  A: Maybe Democrats took it seriously and stayed home on Election Day and Republicans saw it as fake news and voted.
  
  
• Should we get rid of the polls?:
  A: Do you want stories about how some reporter talked to 20 people in a diner and 12 of them liked Trump so Trump is ahead? Is that better than a flawed poll?
  

There were a few more Qs & As, but these are the major ones.

Dylan Matthews of Vox talked to independent data analyst David Shor. Here is a brief summary of their Q & A.

• Q: What the hell happened with the polls?
  A: It was the partisan nonresponse. When pollsters called, excited Democrats all took the surveys and Republicans hung up the phone. They didn't lie. They just refused to talk. We learned that the percentage of survey takers who had donated to ActBlue skyrocketed. Liberals began taking surveys at an astonishing rate and Republicans refused to talk to the pollsters.
  
  
• Q: What is this social trust thing?
  A: The way polling is done just doesn't work any more. The top pollsters use random-digit dialing to call people. About 1% answer and do the survey. If the other 99% are pretty much like the 1% it works fine. But they aren't. The 1% who are willing to talk are weird. They are much more politically engaged. Before 2016, people who trusted their neighbors voted the same as people who didn't trust their neighbors. In 2016, it changed. Only the high-trust people did the surveys and they skewed Democratic.
  
  
• Q: Was 2020 like 2016?
  A: Yes. In 2020, literally 95% of the people who took the surveys voted (vs. 66% of eligible voters). Before 2016, survey takers and refuseniks voted the same way. This year, liberals began answering the pollsters' calls at a very high rate. They were not representative of the refuseniks. The model that once worked suddenly broke.
  
  
• Q: Can you fix that by weighting for religion or sexual orientation?
  A: No. My point is that the model of calling random people, applying classical adjustments for gender, race, etc. doesn't work any more. You need voter files, proprietary data, and teams of machine learning engineers. It's a whole new world.
  
  
• Q: Should pollsters incorporate ethnographic techniques and focus groups?
  A: Qualitative work can help win elections but doesn't explain why the polls were wrong. People don't lie to the pollsters. The polls were wrong because the pollsters were not getting a random sample of the electorate. If you call people at random and say: "Do you want to join a focus group?" Democrats will say: "When can I start?" and Republicans will say: "No." How does that help? People who take phone surveys are weird and people who do focus groups (which takes hours instead of minutes) are even more weird. We are going to need something else.
  

We are inclined to agree with Shor. It used to be that nonrespondents voted the same way as respondents. It doesn't matter if 95% or even 99% of the people called refuse to do the survey as long as their voting preferences are the same as those who take the survey. In the past that was true. It isn't any more. Now there is a huge disconnect. Many Democrats are willing to be polled but far fewer Republicans are willing to be polled.

Something radically different needs to be done. One approach is to ask people how they will vote but also how they think people in their neighborhood/area/city will vote. That could be more predictive because even an enthused Democrat might notice all the yard signs for Republicans around. This has to be tested, of course. Internet and text-message polling get much higher response rates and might have less nonresponse bias. It's worth experimenting with it. A problem here is demographic bias: A lot of well-off young white men use the Internet but not so many poor old Black women. Longitudinal studies might also help. If a pollster sends a link to a survey to exactly the same 100,000 people every month, changes from month to month are likely to be less affected by nonresponse bias. If the Republican is doing 4% better in September than in June with exactly the same group of potential respondents, something probably really did change. This doesn't give an absolute calibration, but it could reflect changes accurately.

Another approach is to use lists of voters in states where partisan registration is public. If you call 10,000 registered Democrats and 10,000 registered Republicans and 5,000 Democrats answer but only 1,000 Republicans answer, you can count each Republican as five people and that might be an unbiased sample. In any event, something needs to be done since the current model of dialing random phone numbers and adding up the scores doesn't work any more. (V)

Biden's Coalition May Not Be Stable

Joe Biden's voters were motivated primarily by getting rid of Trump. There was no outpouring of love for Biden himself, as there was for Barack Obama in 2008 and Bill Clinton in 1992. He's a kindly grandfather and not Trump. That's it. Trump won't be on the ballot on 2022 and probably not in 2024. Will Biden's coalition then fall apart?

The reality is that there are not enough minorities, college-educated suburban women, and young voters (as opposed to young citizens) to win elections. Are the Democrats due for a reality check? Affluent suburban voters may like the Green New Deal but white working class men may say: "I don't care about this talk about saving the planet; what I care about is saving my job as an oil worker." Likewise, they say: "This free college stuff is of no interest to me because my kids aspire to being oil workers like me, not going to college." On the other hand, he might be fine with raising the minimum wage to $15/hr because that also pushes up all wages above the minimum. Biden has to be careful, but there are progressive issues that do resonate with white working class men. Providing broadband Internet to rural areas is another winner. If Biden focuses like a laser on progressive issues that white working class men also like, he could make real inroads with them.

Biden also got substantial support from disaffected Republicans, but downballot they voted for Republicans for the Senate, the House, and state legislatures. There is no reason to think they will become Democrats in 2022 or 2024. So will his coalition last?

Democratic pollster Paul Maslin said: "Biden's coalition is very broad and it's very shallow. The 'Demographics is Destiny' kind of new Democratic coalition is ... not getting there anywhere as quick as people would hope."

Democrats have achieved a patchwork coalition but not, as yet, a fundamental realignment. The election did not show any sign of a permanent movement to the left. It was simply a referendum on Trump. There is no guarantee that the disparate groups will remain together when Trump has exited stage right. Chuck Rocha, a senior adviser to Sen. Bernie Sanders (I-VT), said: "I think we have to get better at talking to people...at starting earlier to talk to Black and brown voters."

One model for the Democrats going forward is to base the party on minorities in urban areas and white voters in the surrounding suburbs. That is what worked in Arizona and Georgia, but it needs to be consolidated. For example, in the diverse suburbs of Gwinnett and Henry Counties outside Atlanta, Biden greatly outperformed Hillary Clinton. If the Democrats can cement that, they can make Arizona, Georgia, and maybe North Carolina permanent purple states.

The instability in Biden's coalition was evident in the Latino vote. In Arizona, where Democrats actively courted Mexican Americans, Biden did well. But in Florida and South Texas, Republicans gained. One problem that Democrats have to deal with is that "Latino" is an artificial construct. If you ask a Puerto Rican to self identify, he'll say "I'm a Puerto Rican." Similarly, a Cuban American or an immigrant from Mexico or Venezuela will identify with his or her country of origin. Few of them will say: "I am a Latino." Lumping them all together as Latinos is a bad idea. Democrats have to see them as separate groups, each with their own concerns. As one example, Puerto Ricans don't care much about immigration policies because they are all Americans. Moving from Puerto Rico to Florida is simply a matter of buying a plane ticket.

Maslin again on this year's downballot losses: "We're not exactly dealing from a position of strength after what happened this year." Democrats who think they are now a permanent majority and get cocky could be in for a rude surprise. That said, realignments generally take several election cycles to fully crystallize. Nobody in 1932 fully appreciated that a sea change had just taken place; it wasn't obvious until 1936 or 1940. And right now, the Democrats are clearly making inroads with suburban voters. It could be that by 2024 or 2028, a new coalition is evident, just one that looks a bit different than expected. (V)

Democrats Can't Win Senate Seats in Trump States

The only state that split its ticket on Senate races in 2020 was Maine. In all the others so far (i.e., all but Georgia), the presidential race outcome completely predicted the outcome of the Senate race. Democrats are up against the reality that more states lean Republican than Democratic in presidential elections and that will increasingly mean a Republican Senate, which will block all Democratic plans unless they can crack that nut. If they could get a Senate majority once, they could admit D.C. and Puerto Rico as states, but if they can't get a majority they can't admit new states, so they are stuck in a catch-22 situation.

Look at the two Trump elections. In the 20 states that voted against Trump twice, Democrats have 39 of the 40 Senate seats. The only Republican there is Sen. Susan Collins (R-ME). In the 25 states that voted for Trump twice, Republicans will soon have 47 of their 50 Senate seats, the only exceptions being Sens. Jon Tester (D-MT), Sherrod Brown (D-OH), and Joe Manchin (D-WV). It looks like 87 of these 90 Senate seats are nailed down. The other five states (Arizona, Georgia, Michigan, Pennsylvania, and Wisconsin) hold the balance of power, but it is a narrow tightrope. To gain power, the Democrats pretty much have to run the table in the close states. The Senate overrepresents the thinly populated rural states that consistently vote Republican. And this reality is not likely to change any time soon.

It wasn't always like this. In the early 20th century, Republicans controlled 96% of the Senate seats that went for William McKinley in 1900 and then Theodore Roosevelt in 1904. Similar, the Democrats held 90% of the seats in the states that FDR won in 1932 and 1936. But after 1960, the connection between the electoral votes and the Senate loosened. After Richard Nixon won twice in 1968 and 1972, Republicans had only half the Senate. It stayed like that until the 1990s, when party-line voting took over again.

The intense polarization of the Trump era has pushed the trend close to its theoretical limit. A candidate like Al Gross, who was crushed in Alaska, would have won in a landslide in Oregon. There is nothing wrong with Gross. He had more than enough money and ran a fine campaign—just in the wrong state. In 2016, for the first time since direct election of senators began in 1918, every Senate race ended the same way as the presidential race. If the Democrats win the two Georgia seats, then Collins will be the only senator who bucked the trend in two cycles. Of course, if the Republicans win the two Georgia races, there will be three exceptions. But even three is not a lot out of 69 Senate races.

This situation shows the enormous difficulty for Senate candidates to swim against the stream in their state. Attractive, unbelievably well-funded candidates in Montana, Kentucky, Iowa, North Carolina, and South Carolina all lost despite this year outspending their opponents by huge amounts.

The fact that two Georgia candidates must win to get to even a 50-50 Senate in a year when Joe Biden got at least 5 million more votes than Donald Trump underscores the nature of the problem. If the Democrats lose one or both Georgia seats, they can do an autopsy if they want. But they will discover that the problem is not that they had bad candidates (except Cal Cunningham, who really wasn't a bad candidate until he...well, you know). Money certainly wasn't a problem. They had more than Scrooge McDuck. Bad campaigns? Not really. All the Democrats ran good campaigns. Better polling? Yes, that would be a real plus, but as we pointed out above, that isn't so easy to fix. The problem is the Constitution's malapportionment, and that is not easy to overcome.

One way out—maybe the only way out—is to turn more states blue and then win those Senate seats. Arizona will soon have two Democratic senators. Georgia conceivably could in January. North Carolina and Florida are possible targets for the future. (V)

Georgia on My Mind--Until Jan. 5, 2021 at 7 p.m.

Why is Donald Trump insisting that he won when he almost certainly knows he lost? But more important, why is the number of Republicans who have congratulated Joe Biden something you can count on the fingers of one hand without even using all the available fingers? It's not that Republican senators are stupid or willfully blind. You can get to be a senator without being able to compute square roots in your head, but counting votes is something that successful senators are generally pretty good at. The cat is out of the bag now. It's because Georgia is on everybody's mind.

The whole charade is about the two Georgia Senate runoffs on Jan. 5. Republicans know very well that if Jon Ossoff and Raphael Warnock prevail, that President-of-the-Senate-elect Kamala Harris will have a full-time job breaking ties in the Senate and Joe Biden will be able to enact much of his program. Trump is grasping at straws because he is aware that Manhattan D.A. Cyrus Vance and New York AG Letitia James are closing in on him and he was planning to move to Palm Beach, FL, not to Attica, NY. So he is not giving up for personal reasons.

But Senate Republicans have a different logic. Most of them probably privately prefer that Trump does move to Attica so they can take their party back. However, they are afraid that if they cross him and say: "It's over. Joe Biden won," then Trump's supporters will become angry and demotivated and won't vote in the two Georgia runoffs. If Ossoff and Warnock win, then the Republicans are really in deep doodoo, so some are saying things they very well know to be false (Trump won, voting irregularities, etc.) to help keep Georgia Republicans energized with a tiny bit of hope. The Senate Republicans fear that if they were to congratulate Biden, it would so deflate Georgia Republicans that many of them would skip voting in the runoffs. So it is all about Georgia, not Trump, and no, they are not stupid. On Jan. 6, most of them will suddenly realize that Biden won and will congratulate him and say they are looking forward to working with him. That's not true either, but that is a different story.

The Georgia runoffs are also the reason Sens. Kelly Loeffler (R-GA) and David Perdue (R-GA) are calling for Georgia Secretary of State Brad Raffensperger (R) to resign. In reality, they have nothing against Raffensperger. He ran an honest election and they know that. They are dissing him because Trump ordered them to do that in order to try to flip Georgia's 16 electoral votes. If they were to say: "Hey, I know Brad. He's a Republican and is doing his job well," Trump would have a hissy fit. Without thinking about the consequences, Trump would then send out a tweet of death for them, thus sealing their fates. They had no choice. The whole show is about not poking the bear.

Probably the highest-profile person who claimed election fraud was Trump's TV lawyer, Rudy Giuliani, who spoke from a parking lot in Philadelphia last weekend and said the election was rife with fraud. Theoretically weakening the case that there was fraud are the statements from multiple state officials that said there was no fraud. A spokewoman for Pennsylvania's AG, Josh Shapiro (D), said: "No active lawsuit even alleges, and no evidence presented so far has shown, widespread problems." Ohio Secretary of State Frank Rose (R) said: "The conspiracy theories and rumors and all those things run rampant. For some reason, elections breed that type of mythology." Kansas Secretary of State Scott Schwab (R) had his spokeswoman say: "Kansas did not experience any widespread, systematic issues with voter fraud, intimidation, irregularities or voting problems." Montana's secretary of state, Corey Stapleton, tweeted: "I have supported you, Mr. President, we (Montana) have supported you—and @realDonaldTrump accomplished some incredible things during your time in office! But that time is now over. Tip your hat, bite your lip, and congratulate @JoeBiden." In practice, few, if any Republicans will change their minds any time soon. However, the electoral votes have to be certified by Dec. 8 and the electors vote on Dec.14. Will Republican senators and other Republican officials continue to support Trump after the electoral votes have actually been cast? In particular, will officials up for reeelection in 2022 do that, knowing that Democrats have their cameras rolling? (V)

Stacey Abrams Raises $6 Million for the Georgia Runoffs

Democrats raised half a billion dollars for the top dozen Senate races and lost nearly all of them. The model behind modern campaigns is that if a voter sees 25 ads per evening calling some Republican a scoundrel but only 10 ads about his Democratic opponent that call him a low-life, the voter will pull the lever for the Democrat. It doesn't work. Stacey Abrams tried something else in Georgia: She registered thousands of new voters. The result is that Georgia went blue for the first time since 1992. Maybe she is onto something: It's the ground game, stupid.

She has now raised $6 million for the two January Senate runoffs. Money people give to the candidates will just air more negative ads, whereas money given to Abrams' group will result in more voter contact. If she registers a lot of voters between now and Dec. 7 (when registration closes), and the Democrats win due to an influx of new voters, that will send a very powerful signal to all politicians (and maybe to donors) that running more negative ads just doesn't cut it any more. (V)

Michael Cohen: Trump Will Go to Florida for Christmas--and Stay There

Michael Cohen, formerly Donald Trump's fixer and currently a felon who is staying at home under house arrest, knows Trump on a personal level as well as anyone outside Trump's family. Yesterday, he made a bold prediction. Cohen said that Trump will go to Mar-a-Lago in Florida, his new home, for the Christmas holidays and just stay there until after Jan. 20. He won't come back to the White House at all, even though he will remain president until noon on Jan. 20. He won't hide out in the White House and have to be removed by the Secret Service on Jan. 20. But he also won't go to the inauguration and sit still in a chair while Joe Biden is giving a speech, with half the cameras focused on him and a chyron reading "Former president Donald Trump" reminding the entire world that he is a loser.

Cohen also was explicit about Jan. 20, saying: "He's not going to stay in the White House past Jan. 20. They will remove him. He knows that." Trump doesn't want video footage of him being dragged from the White House by force (and possibly him being charged with trespassing) to go viral. He also doesn't want footage of him slinking out of the White House on Jan. 19 to go worldwide. Nor does he want footage of him leaving the White House at 11:55 a.m. on Jan. 20 while raving like King Lear on the heath to become the hallmark of his presidency. So what better way to end it than go to Florida in December and just stay there? After all, he can spend his time tweeting and pardoning his friends quite well from there. Plus, think of the final bill he will send to the Secret Service. (V)

---

If you wish to contact us, please use one of these addresses. For the first two, please include your initials and city.

• questions@electoral-vote.com    For questions about politics, civics, history, etc. to be answered on a Saturday
• comments@electoral-vote.com   For "letters to the editor" for possible publication on a Sunday
• corrections@electoral-vote.com To tell us about typos or factual errors we should fix
• items@electoral-vote.com            For general suggestions, ideas, etc.

To download a poster about the site to hang up, please click here.

---

Email a link to a friend or share:

---The Votemaster and Zenger