November 2014, Part 1
Jim Miller on Politics
How Accurate Were The Polls On The Senate Races? Not very. And they were off, systematically.
Based on results as reported through early Wednesday morning — I’ll detail our method for calculating this in a moment — the average Senate poll conducted in the final three weeks of this year’s campaign overestimated the Democrat’s performance by 4 percentage points. The average gubernatorial poll was nearly as bad, overestimating the Democrat’s performance by 3.4 points.Partisans often claim that the polls are biased against their party, and they are — sometimes — right.
This evidence suggests that polling bias has been largely unpredictable from election to election. Beyond the shadow of a doubt, the polling was biased against Democrats in 1998, 2006 and 2012. However, just as certainly, it was biased against Republicans in 1994, 2002 and now 2014. It can be dangerous to apply the “lessons” from one election cycle to the next one.You will, naturally, wonder whether the "bias" in those in those cycles is a result, not of systematic polling errors, but pollsters missing out on late shifts. Nate Silver doesn't discuss that possibility in this post, so I suppose I should look to see if he has, earlier.
- 10:59 AM, 8 November 2014 [link]
How Accurate Was The Washington Post's Prediction On The Senate Races? Quite, whether you measure by a simple count, or by a measure that takes into consideration how confident the predictions were.
Moreover, our predictions in individual races were almost entirely correct. Assuming that Warner in Virginia, Sullivan in Alaska, and Cassidy in Louisiana are ultimately going to win, we will have called 35 of 36 races correctly.(They missed North Carolina, which was close.)
How did they do it? By using, John Sides says, the "fundamentals established by the political science literature". That's a bit vague, so I'll mention two that I am nearly certain they used, estimates of the numbers of Republican and Democratic voters in each state, and the likelihood that each party's potential voters would actually vote.
For example, one reason I was not surprised by the Cory Gardner win in Colorado is that Republicans — in spite of Democratic wins there in recent years — have a small edge in registrations. Everything else being equal, a Republican should win the state. Especially in off-year elections, when Republicans are more likely than Democrats to vote.
(There are some complexities to using registrations. For example, many southerners have not changed their party registrations, even though they have been voting Republican in national elections, for years. But in states like Colorado and Iowa, just looking at registration numbers for the two parties will tell you a lot.)
That's a pretty good prediction, I'd say, although they almost missed on Virginia, which would have been a shocker.
(Stat folks may want to look at his description of the "Brier scores" he used to measure the accuracy of his predictions.)
- 10:35 AM, 8 November 2014 [link]
Argon As A Pollutant? The Environmental Protection Agency is proposing to ban the use of argon, as an additive in pesticides.
U.S. Environmental Protection Agency’s (EPA) Office of Pesticide Programs has published a list of 72 chemicals it proposes to remove from the “Approved Pesticide Inert Ingredient List” (docket # EPA-HQ-OPP-2014-0558) and comments are accepted till Nov. 21, 2014.Sometimes I get the feeling that the people running the EPA don't understand high school chemistry (or high school biology, for that matter).
By way of Anthony Watts.
(How hard is it to get a noble gas to combine with other elements? Linus Pauling predicted that it was possible in 1933, but it wasn't until 1962 that first researcher, chemist Neil Bartlett, succeeded in doing so.)
- 8:25 AM, 7 November 2014 [link]
Tuesday's New Yorker Cartoon is timely, or perhaps just little ahead of time, since the new Senate will be asking questions that some witnesses (many of them working for the Obma administration) will prefer not to answer, loudly and clearly.
- 7:41 AM, 7 November 2014 [link]
This 2014 Election Result Map is deceptive.
For instance, although Washington state is mostly red, Democrats actually control 6 of the 10 House seats here.
And, from another perspective, the map is not deceptive. What it shows is that the Democratic Party is almost nowhere competitive in rural districts. And that wasn't true forty years ago.
What has happened is that the two parties have traded, with the Republicans coming out ahead on the trade. If you look at similar maps from 1974, you'll find the Democrats winning many rural districts — and, looking closely, Republicans winning some wealthy "silk stocking" districts (as they were often called) in the larger cities.
Why the switch? Many reasons, but two areas seem especially important to me, social issues and the environment.
(For the record: I'd like to see both parties try a little harder to compete in places where they may not feel entirely welcome, Democrats in rural areas, Republicans in central cities.)
- 2:41 PM, 6 November 2014 [link]
Keith Koffler Watched President Obama's Press Conference So We Didn't Have To: Even better, Koffler has translated it for us.
“Obviously, Republicans had a good night, and they deserve credit for running good campaigns.”
- 12:46 PM, 6 November 2014 [link]
An Egyptian Bus Driver Got Some Good News And Some Bad News, at the same time.
A bus driver who tried to avoid a being caught out in a drugs test by using his wife's urine got busted when officials told him, 'Congratulations, you're pregnant.'To avoid similar tricks, the transport authority will be using blood tests, too, in the future.
- 12:24 PM, 6 November 2014 [link]
Mistaken Election Predictions: Politico has a list of them. (Not all of them are actually mistakes. A person who says that something might happen is not necessarily wrong if it doesn't happen.)
I don't blame the Democratic party leaders for what they said before the election, just as I don't blame coaches of heavy underdogs who insist that their team has a chance in the upcoming game.
But I do blame those who are paid to make predictions and one of them, the new numbers guy at the New York Times, Nate Cohn, may have let his hopes for Democratic victories bias his predictions.
For example, here is what he wrote about the governor's races.
On balance, Democrats seem set to pick up two or three states, mainly because the Republicans enter the elections with twice as many Republican-held seats. But it is easy to imagine the Republicans holding their advantage — there are 29 Republican governors and 21 Democratic ones — or the Democrats picking up a half-dozen seats.Republicans held all those states (and Maine), and picked up Massachusetts, Maryland(!), Illinois, and Arkansas. The Republican loss in Pennsylvania cut their net gains to three.
At best, Cohn thought Republicans might break even, but in fact they made gains, net.
(To be fair to Cohn, he may have been fooled by his newspaper's own polls, or may have missed a late shift to the Republicans.
Here's the Real Clear Politics table, for those who want to see some numbers.
Alaska is a weird special case. The Republican incumbent, Sean Parnell, is losing to a former Republican running as an independent, Bill Walker. Former governor Sarah Palin has endorsed Walker, so you could see this general election as something like a Republican primary, in Alaska's often fractious Republican Party. (It would be an open primary, since the Democrats have joined Walker.) This Wikipedia article will give you some of the basics on the race.)
- 7:37 AM, 6 November 2014 [link]
This Election Result Will Make Andrew Stiles Happy: Republican Chris Gibson defeated Democrat Sean Eldridge in New York's 19th House district by 30 points, just as Stiles had requested.
(It occurs to me that Eldridge may have lost so badly in part because he was unable to fake a sincere interest in the district, convincingly. For example, he should have — speaking for the moment as a political tactician — promised to stay in the district, win or lose.)
- 6:57 PM, 5 November 2014 [link]
In October, I Called For Seattle To Allow People There to view "The Simpsons".
Very much to my surprise, the authorities listened, and Seattle voters did the right thing.
Seattleites overwhelmingly rejected a request to spend $2 million per year for yet another layer of transportation planning, this time to explore a monorail line.It's another victory for freedom of speech.
- 1:56 PM, 5 November 2014 [link]
The Biggest GOP Senate Majority Since? The Virginia race may be subject to a recount, the Louisiana race will go to a run-off in December, and Mark Begich has still not conceded in Alaska. If the Republicans win all three — the first is unlikely, the second and third likely — they will have a net gain of
To beat that 55, you would have to go all the way back to . . . . 1928.
And the 56 seats that the Republicans won that year were in a Senate with just 96 members. (Alaska and Hawaii first elected senators in 1958.)
- 1:27 PM, 5 November 2014 [link]
The Biggest GOP House Majority Since? Our "mainstream" journalists are looking for historical comparisons for the enlarged Republican majority, and mostly, judging by the stories I've heard, getting them wrong.
This CNN story is typical.
In the House, CNN projected the GOP will have at least 246 seats, its largest majority since World War II. Speaker John Boehner, celebrating a widened majority, said he is "humbled by the responsibility the American people have placed with us."(Emphasis added.)
Actually, it is likely to be their largest majority since 1928. Republicans had 233 House seats going into this election. A 13 seat gain would give them 246 seats, the same number they won in 1946. More would give them more seats than they won in any election since 1928.
As I write, Real Clear Politics is saying that the Republicans have gained 13 seats — and there are a number of seats that are still undecided.
These are not obscure historical facts. You can find them, for instance, in an inexpensive almanac. But somehow doing that simple check is beyond many of our "mainstream" reporters. And their editors.
(Here's another, even worse, example from, of all places, Roll Call: "on track to win their largest majority since 1949".)
- 12:59 PM, 5 November 2014 [link]
Worth Reading: (If you are at all interested in political tactics.) This Washington Post article on how the Republicans won the Senate.
Minutes after landing at Reagan National Airport one day early this year, many GOP Senate hopefuls found themselves besieged at baggage claim by people with cameras yelling questions at them about abortion and rape.That last supports my contention that President Obama spends most of his time on politics, not management, or even policy.
But not very effectively. According to the article, congressional Democrats believe that "Obama’s political operation functioned purely for the president’s benefit and not for his party’s". From what I can tell, they are right about that, which shows how short-sighted he is, politically, since his success depends on theirs.
- 6:13 AM, 5 November 2014 [link]
As Expected, The Washington Gun Control Initiative I-594 is passing easily.
Which leads, naturally, to this question: If it is that popular with the voters, why couldn't a similar measure pass the Washington legislature, especially when it was controlled by the Democrats, as it was, fairly recently?
My tentative answer: Most likely because legislators understand that opponents of such measures are far more intense than supporters, more likely to vote against a legislator on that issue alone.
So the supporters of this measure were probably right, tactically, to run an initiative, rather than try to lobby the legislature one more time.
And we shouldn't be terribly surprised if a future legislature, working with a Republican governor, repeals the initiative, in whole or part.
- 9:08 PM, 4 November 2014 [link]
Arch Moore Is Smiling Tonight: After decades of battling Jay Rockefeller in West Virginia, unsuccessfully, he is seeing his daughter, Shelley Moore Capito, take over the Senate seat that Rockefeller is giving up.
(In general, I have no more objection to children following their parents into politics than I do to children following their parents into any other business, though I can certainly name many children (and grandchildren) that I wish hadn't. I would say that Capito's career, so far, shows that she has inherited at least some of her father's political skills, without, as far as I know, any of his willingness to cut ethical corners.
As you probably guessed, Jay Rockefeller was able to outspend Moore by a very large margin. In one of their contests, a Moore supporter urged him to make Rockefeller spend it all, presumably for what it would do for the West Virginia economy.)
- 6:54 PM, 4 November 2014 [link]
My Half-Empty Ballot: This is an off year here in Washington, so there are no statewide offices on the ballot. As it happens, neither of our senators is up for re-election, so the top office on my ballot was the congressional race. (And I voted for Pedro Celis without any hesitation.)
After that, it was discouraging to see how many candidates had no opponents. There were four races for the state supreme court; two of them had just one candidate. There were three races for the appeals court; all three had just one candidate. There were seven races for local courts; four of them had just one candidate. (I cast a write-in vote for Rob McKenna, our former attorney general, in one of the supreme court races.)
There were three legislative races; in one of them the Democrat had only a Libertarian opponent (who I voted for).
Finally, the Republcan county prosecutor, Dan Satterberg, had no opponent. I have no idea why this very Democratic county, King, did not produce at least a nominal opponent for Satterberg.
- 3:03 PM, 4 November 2014 [link]
The NYT's Big Senate Campaign Graph: Which you can see here.
For the record: Although I disagree with some of their numbers — for example, I would make Scott Brown a small favorite in New Hampshire — I love the way they present all that data from their forecasts.
(For those who have to present charts from time to time: This is an especially good example of "small multiples", a term popularized by Edward Tufte. Search images on his name, and that phrase, for examples.)
- 1:27 PM, 4 November 2014 [link]
Exciting Canadian Sex Scandal: That sounds like the punch line to a joke, doesn't it?
But they have an exciting sex scandal, one important enough so that I am interrupting my flow of posts on the election to describe it for you. (And, it's a sex scandal that involves our National Public Radio, tangentially.)
Jian Ghomeshi was a very popular radio host for the very respectable Canadian Broadcasting Corporation (which is, of course, supported by Canadian taxpayers).
But no longer, and it wasn't because of his performance on the air.
TORONTO (AP) — Police have begun an investigation of a prominent former Canadian Broadcast Corp. radio host on sexual assault charges after three women filed complaints against him.Ghomeshi has claimed that this sexual practices, though perhaps kinky, were consensual, but enough women have come forward with complaints so that we can have some doubts about his defense.
Now here is what I consider the most exciting part of that sex scandal — though some may disagree. According to a respected Canadian journalist, Ray Heard, some CBC officials have known about Ghomeshi's attacks on women for "12 years".
No one, no adult anyway, should be surprised by the possibility that a bureaucracy may have been covering up a scandal. That's what bureaucracies often do. (Some would say, routinely.) Especially when they are protecting someone in the bureaucracy they value.
But this scandal, and the far larger (as far as we know now) Jimmy Savile sex scandal at the BBC make me wonder whether sex scandals are especially common in news bureaucracies, and whether those bureaucracies are especially likely to cover them up.
(Here's Ghomeshi's Wikipedia biography, if you are wondering about his background, or even his rather un-Canadian name.
How is NPR involved? Many NPR stations carried his show.)
- 8:11 AM, 4 November 2014 [link]
Piers Morgan Turns Against President Obama: In 2008, the British journalist celebrated Obama's election
When Barack Obama was first elected President in 2008, I celebrated by having an exquisitely romantic dinner on a tiny, temporary sand dune in the middle of the Indian Ocean.Yesterday, the conservative Daily Mail published Morgan's article giving ten reasons why Obama deserves a "shellacking" in today's election. (I agree with all of them except number 6, since I have learned, through long experience, that boring elected officials are often better for us than exciting officials.)
Here are two of those ten:
8. HE CAN’T NEGOTIATE. I wouldn’t trust Obama to renegotiate my monthly gardener’s fees. He just doesn’t know how to make a deal. Politics, at its best, is about strong, powerful men and women getting in each others’ faces, arguing the toss about issues they care about, and agreeing points of compromise. Clinton was a master at it. Obama, by general consent, is useless. Donald Trump may not be everyone’s cup of tea but I can tell you this – he’d negotiate way better deals for America than Obama. It’s in his blood.Morgan does not, at least in that article, discuss why he was so easily fooled, but I suppose we should be mildly pleased that he has come this far, however late. And displeased that so many of his fellow journalists have not faced the obvious, as he has.
(It isn't hard to understand why Morgan, and so many other journalists, were fooled in 2008. Obama told them what they wanted to hear, they loved a symbolism of a black president (however inauthentic he might be), and as people who earn their livings with words, they naturally pay less attention to deeds — or lack of them — than they should.)
- 6:51 AM, 4 November 2014 [link]
The French, Too, Have Figured Out That President Obama Is In Trouble: The lead story in Le Figaro, a moderate conservative newspaper, is headlined: "Barack Obama face à une débâcle électorale annoncée". Which you can get the general sense of, even if you don't know any French.
Similarly, Le Monde, a leftist newspaper, has this headline on a chart illustrating the fight for the Senate: "Le Sénat, un casse-tête pour Barack Obama".
(A "casse-tête" could be translated as a headache, a problem, or a puzzle, depending on the context.)
- 5:15 PM, 3 November 2014 [link]
"5 State Senates And 9 State Houses" That's how many that political scientist Carl Klarner is predicting that Republicans will gain in tomorrow's election.
According to his model, Democrats have a 46 percent chance of keeping control of the state senate in New York, and a 25 percent chance in Washington.
In Washington state, most political observers would be surprised by his prediction that Republicans will take control of the house, as well as the senate, because the Democrats had a bigger majority in the house (55-43) than in the senate (26-23) after the 2012 election. (Those numbers are from the 2014 Almanac of American Politics, and have changed slightly since the book was published.)
Caveat: The brief description he gives of his model sounds plausible, but I haven't looked at it (and don't have time to before the election), nor do I know what the model's track record is, though I assume he tested it against past elections. (You can find some of his data here.)
One thing he has going for him in building this kind of model is the sheer number of legislative elections, which makes it easier to find historical patterns.
One general point: Although most attention is paid to the glamorous races at the tops of tickets, I think these down-ballot races give us better measures of the underlying strengths of the two parties. And, of course, the state legislatures often serve as "farm teams" for the national parties.Cross posted at Sound Politics.
- 3:03 PM, 3 November 2014 [link]
What Do The Bettors Say About Senate Control? As I write, the closing price at the University of Iowa site gives the Republicans about a 90 percent chance of controlling the Senate (and a 100 percent chance of controlling the House).
That seems about right to me, though I am still thinking about the problem.
Here's their graph, for those who want to see the trend. A year ago, that chance was about 20 percent, though I would guess that the trading was thin, back then, given the way the graph bounces around.
- 10:50 AM, 3 November 2014 [link]
Australian Native Fred Turned 100, Got A Letter From The Queen, and a nice birthday party.
A sulphur-crested cockatoo has apparently made it to his 100th year, surpassing most birds of his kind by 60 years.He probably didn't care about the letter, but he appeared to enjoy the party.
If he and the queen were ever to meet, they would probably get along just fine.
But Fred doesn't warm to everyone - he has a preference.Though she may want to leave Prince Philip behind.
(Naturally, I wondered whether that age was possible for a cockatoo. Judging by this Wikipedia article, I would say that it may be, since there are documented cases in which they have lived — in captivity — to 70 or more, and there are stories of a few living even longer.)
- 9:18 AM, 3 November 2014 [link]
The Guardian Is Expecting Leftist Losses In The United States Tomorrow, and has already prepared an excuse.
The prospect of Democratic losses stems from more than just the unfavourable fundamentals of an off-year election and an unpopular president (eligible voters currently disapprove of Obama’s performance by a 53-42 margin). Democrats are also suffering in 2014 from the retirement of key senators and the bad timing of having multiple re-election battles playing out in states where Republicans performed especially well in 2012.So there you are; if the Democrats lose close elections tomorrow, it will because those mean Republicans aren't letting people, "particularly from Democratic-leaning backgrounds", vote.
Probably, the Guardian believes that argument, as so many on the left here in the United States do. Probably, the Guardian does not care about the possibility that vote fraud may tip some close races to the Democrats.
(Since it is the Guardian, you expect mistakes, and I noticed a big one. They say that George W. Bush lost control of the Senate "midway" through his presidency, which would be 2004, rather than 2006.
Having said that, I'll add that the survey article also includes an interactive map on voting rules, and some links that you might find useful.)
- 8:37 AM, 3 November 2014 [link]
Why Is This Ebola Outbreak So Much Larger Than Previous Outbreaks? (Since I have been thinking about disease the past few days.)
First, some numbers:
The disease typically occurs in outbreaks in tropical regions of Sub-Saharan Africa. From 1976 (when it was first identified) through 2013, the World Health Organization reported 1,716 confirmed cases. The largest outbreak to date is the ongoing 2014 West Africa Ebola virus outbreak, which is affecting Guinea, Sierra Leone, Liberia and Nigeria. As of 29 October 2014, 13,567 suspected cases and 4,922 deaths had been reported.(Which are probably correct within a factor of 2, maybe even correct within 25 to 50 percent. The three countries most affected, Guinea, Sierra Leone, and Liberia, have a population together of about 20 million, so the disease has hit more than 1 in 2,000 in those three countries.)
Past outbreaks hit at most a few hundred; this one has already infected more than 10,000. Why?
As far as I can tell, no one knows the answer to that question, so what follows is speculation.
One possibility, of course, is that the virus has mutated so that it spreads from person to person more easily. (It's an RNA virus, so it has a high natural mutation rate.)
That's possible, but I think it less likely than another explanation: globalization and the resulting rising prosperity in parts of Africa. How would that help spread the disease? By breaking down the old village-based social structures, so that relatives are more spread out, and transportation to and from the villages is easier.
Epidemiologists appear to believe that the disease was spread, at the beginning, by the funeral services for the first victims. (In that part of Africa, most funeral services involve much contact with the deceased.) In the old village-based societies, those attending the services would have been almost entirely from the village where the dead person had lived. Those who lived farther away might not know of the death, and might not have had any practical way to get to the village. Now, however, they can learn of the death by phone, and often get to the services by bus, or even car.
This isn't, at all, a novel explanation. Almost everyone knows that wars often spread diseases; so, oddly enough, can peace and prosperity, by making it safer and easier for people to travel. Right now, airline passengers are carrying new strains of flu all around the world, spreading them far more quickly than they could have spread when the fastest way to travel was a sailing ship, and when the average person made at most one long trip in his or her life time. (The Black Death was spread by both war and peaceful trade. The parts of Europe that were spared were mostly the parts where there was little trade.)
Although this is not at all novel, it is not something that advocates of globalization are likely to stress, or even mention, in their arguments.
But the rest of us should recognize that it is one of the inevitable costs of globalization.
(There is a third possibility that should be mentioned: bad luck. The chains of transmission are probabilistic. It is possible that, through chance, the first few people with the virus spread it to many more than you would expect, ordinarily. Or perhaps this is just the first time the outbreak began in a large market town.)
- 7:58 AM, 3 November 2014 [link]
A Few Days Ago, I picked up a bug* — perhaps from reading too many Ebola stories — which is why you haven't seen any posts here for a few days, and why I am even further behind on my email than usual.
Bad timing, since I was just getting ready to do a series of posts on the election — and I do plan to do some today and tomorrow, just not as many as I had hoped. And I am hoping to get to the email backlog soon, perhaps as early as this Friday.
(*Nothing very serious, but I figure if I have a fever I shouldn't go out and share it with other people.)
- 6:18 AM, 3 November 2014 [link]