News from the Votemaster

History of the Early States

It's finally 2016. Let the caucuses and primaries begin! There is a tremendous amount of attention (probably too much) on the four early states so perhaps it is a good idea to see how predictive the Republican caucuses and primaries have been in past open-seat years. Here are the winners since 1980. Note that Nevada switched from a primary in the middle of election season to an early caucus in 2008.

Year	Iowa	New Hampshire	South Carolina	Nevada	Nominee	President
1980	Bush	Reagan	Reagan	Reagan	Reagan	Reagan
1988	Dole	Bush	Bush	Robertson	Bush	Bush
1996	Dole	Buchanan	Dole	Dole	Dole	Clinton
2000	Bush	McCain	Bush	Bush	Bush	Bush
2008	Huckabee	McCain	McCain	Romney	McCain	Obama
2012	Santorum	Romney	Gingrich	Romney	Romney	Obama

One simple indicator of how good the early states are is to look at all 24 caucuses and primaries listed above and count how many times the eventual nominee won them. It is 15 of the 24 times, for a score of 62.5%, decent, but not overwhelming. Iowa, which gets so much attention, is batting .333, New Hampshire is batting .667, South Carolina is hitting .833, and Nevada is at .667. So South Carolina is the most predictive of the lot. It is possible that by the time South Carolina steps up to the plate, the field has been winnowed enough that it is more serious. It is also possible that the South Carolina Republican electorate is more like the national Republican electorate than the other states. In any event, keep in mind that Iowa gets it wrong most of the time. (V)

Bettors Think It's Rubio by a Nose

Another way to look at the Republican race is through the eyes of bettors. People who have their own money on the line are more likely to mean it, whereas people can tell pollsters anything, and some of them won't bother voting. Political betting is illegal in the U.S. but legal in Ireland and the U.K. Paddy Power is the largest Irish betting site. Here are its current odds on the various candidates winning the Republican nomination. For people more used to traditional probability theory than gamblers' odds, the fourth column below gives the implicit probabilities the bettors are assigning to each candidate.

Rank	Candidate	Odds	Probability
1	Marco Rubio	7/4	.364
2	Donald Trump	2/1	.333
3	Ted Cruz	5/2	.286
4	Jeb Bush	9/1	.100
5	Chris Christie	14/1	.067
6	Ben Carson	66/1	.015
7	John Kasich	100/1	.010
7	Carly Fiorina	100/1	.010
7	Mike Huckabee	100/1	.010
7	Rand Paul	100/1	.010
11	Jim Gilmore	200/1	.005
11	Rick Santorum	200/1	.005

So bettors see Rubio, Trump, and Cruz bunched up near the top and the rest as longshots.

Paddy Power also takes bets on the general election. The probability of Hillary Clinton winning is .579, with Donald Trump at .181, Marco Rubio third at .167, and Ted Cruz at .125. No one else is close. (V)

Trump's Supporters May Be Prevented from Voting for Him

A new analysis from Nate Cohn turns up several surprising items. One of them is that a substantial number of Donald Trump's supporters are actually registered Democrats. For example, one of his best states is West Virginia, which generally votes Republican in presidential elections, but where a majority of the voters are registered as Democrats. The problem for Trump is that in a number of states, only registered Republicans are allowed to vote in the Republican primary, so some of his supporters may show up to vote and when given a ballot listing Hillary Clinton, Bernie Sanders, and Martin O'Malley may write in: "None of the above."

Another interesting observation that Cohn found is that a map of the U.S. showing where Trump's support lies is very similar to a map showing Google searches for racially-charged terms. Now, correlation is not causation, but it is still interesting. More likely is that Trump supporters are older, poorer, less-educated white people, and those people are also more likely to do politically incorrect searches. (V)

Polls May Be Underestimating Trump Support

There are many complications when it comes to guessing what Donald Trump's fate will be when people actually go to the polls to vote. In addition to the one discussed above is the fact that online polls consistently show higher support for the billionaire than telephone polls. Polling firm Morning Consult has examined the phenomenon, and believes it has an answer: White, college-educated males are embarrassed to admit their support for The Donald.

The methodology of the study is solid—Morning Consult conducted an online poll and then called some of the same respondents. In this way, they were able to identify individuals who gave one answer online and then a different answer when speaking to a real person. Assuming that the conclusions are correct, then Trump support may be 5% or 10% higher than is generally reported. That's the good news for the billionaire. The bad news is that, one way or another, he is likely to have an unusually large "spoil" rate—supporters who do not show up to the polls, or who show up but can't actually vote for him. So there is every chance that the Donald's election returns end up way off his polling numbers—the question is in what direction. (Z)

Trump's Attacks on Bill May Help Hillary

Donald Trump thinks that by bringing back the Bill & Monica show, he can remind the voters of Bill's bad past behavior and thus hurt Hillary. The data suggest that he is wrong. Polls about her favorability have been taken almost continuously since 1992. Her all-time high was 66%, which she achieved twice. The first time was at the height of the Lewinsky scandal; the second was during her tenure as Secretary of State. What is especially notable is that as the Lewinsky scandal got more and more news coverage, her favorability went up, not down. While the reasons can't be derived from the data, most likely some people felt her "stand-by-your-man" position was the honorable thing to do, despite the humiliation. But Trump probably doesn't follow polls, certainly not old ones, so he will undoubtedly continue to blast away, even if it reminds people that she was the victim and suffered in silence. (V)

O'Malley Fails to Qualify for Ohio Ballot

In 2012, 2.7 million Democrats voted in the presidential election in Ohio. To get on the primary ballot, a candidate needs to get 1,000 registered voters to sign a nominating petition, which is about 1/25 of 1% of this total. Democratic candidate Martin O'Malley failed to make the cut. He did submit 1,175 signatures, but only 772 were valid. O'Malley is contemplating what to do next. Dropping out might be a good idea, for example. (V)

More Clinton E-mails Released

The State Department has released another 5,500 pages of Hillary Clinton's damn emails. Did it do so on a holiday in order to bury the story? Probably. Nonetheless, several journalists have given up their day off to comb through the pile, and so far nothing too significant has been found. Most of the correspondence is fairly pedestrian—following up on phone calls, planning meetings, a bit of humor. The most salacious messages uncovered so far reveal that billionaire Democratic donor George Soros and German Chancellor Angela Merkel both had some unkind things to say about President Obama. Not exactly a smoking gun, but maybe the GOP will find something when the final batch of messages—another 12,000 pages or so—is released later this month. (Z)

No Matter Who Retires From the Supreme Court, Liberals Might Win

The Washington Post has a fascinating but arguable, analysis of the politics of the next Supreme Court appointment. And the conclusion is that just about every scenario imaginable to the author favors liberals.

The article begins with an assumption, one that has already become common among legal scholars, that any appointment that would change the makeup of the court would be blocked or filibustered by the opposing party, very possibly for years. In that scenario, the Supreme Court would be down to eight justices thanks to the vacant seat. The Constitution does not specify any particular number of justices, so this is entirely legal. Historically it has varied from 5 to 10. The result would be a large number of tie votes, and well-established precedent is that when the Court has a tie vote the ruling of the lower courts stands (although without establishing a binding national precedent). Since the lower courts are dominated by liberal justices, it means a lot of left-leaning rulings would survive rather than being struck down by the higher court.

However, there are some scenarios that the author missed. Suppose Ruth Ginsberg, who is just shy of 83 now and who has had both colon cancer and pancreatic cancer dies early in the next President's term. That would leave the Court divided with 4½ conservatives and 3½ liberals (Anthony Kennedy tends to swing both ways). The Senate Republicans might be content to keep it that way for 4, maybe 8 years, no matter who is President and no matter which party has a majority in the Senate. After all, in this situation, best case is conservatives win, worst case is a tie and no national precedent is set.

Another possible—in fact, likely—scenario is that the same party controls both the White House and the Senate. If a vacancy occurs, that party might want to grab its chance and either abolish or water down the filibuster. An easy first step would be to change the Senate rules requiring that filibusters all be "Mr. Smith Goes to Washington" filibusters. After all, Senate rule XXII was originally adopted to make sure no senator was interrupted while speaking. The new rule could require a senator who has the floor to stand (not sit) in the well of the Senate and talk. Very few senators could keep that up for even 12 hours. So a minority of 49 senators might be able to keep talking for 24-25 days continuously, but the majority could just wait them out and approve the nominee a month after he or she was nominated. Given how important the stakes are for lifetime appointments to the Court, both parties might be willing to go this route.

Supreme Court Nominations have not always been supremely controversial. Franklin D. Roosevelt nominated 8 justices to the Court. None of them faced lengthy battles in the Senate. Here is the list of his appointees and how much time elapsed between Roosevelt's nomination and the Senate confirmation vote.

Justice	Nominated	In Senate
Hugo Black	Aug. 12, 1937	5 days
Stanley Reed	Jan. 15, 1938	10 days
Felix Frankfurter	Jan. 5, 1939	12 days
William Douglas	Mar. 20, 1939	15 days
Frank Murphy	Jan. 4, 1940	12 days
Harlan Stone	Jun. 12, 1941	15 days
James Byrnes	Jun. 12, 1941	0 days
Robert Jackson	Jun. 12, 1941	25 days

The most noteworthy in terms of confirmation was James Byrnes, then a senator from South Carolina who was never a judge and who never even went to law school. He was approved by a voice vote of the Senate the same day as Roosevelt nominated him. (Z & V)

Email a link to a friend or share:

---The Votemaster