May 2015, Part 1
Jim Miller on Politics
Why Did The Modelers Miss On Their British Election Predictions? Because the modelers were relying on the polls, and the polls were wrong, in one very important way.
Ben Lauderdale, one of the team that made predictions for Nate Silver's site, FiveThirtyEight, explains:
The most obvious problem for all forecasters was that the polling average had Labour and the Conservatives even on the night before the election. This was not just the average of the polls, it was the consensus. Nearly every pollster’s final poll placed the two parties within 1 percentage point of each other. Based on the polling average being level, we predicted Conservatives to win by 1.6 percentage points on the basis of the historical tendency of polls to overstate changes from the last election. This kind of adjustment is helpful for understanding how the 2010 result deviated from the national polls on election day, as well as the infamous 1992 U.K. polling disaster, when the polls had the two parties even before the election and the Tories won by 7.5 percentage points. The Conservative margin over Labour will be smaller than that when the 2015 totals are finalized, but not a lot smaller (currently it is 6.4 with all but one constituency declared). So our adjustment was in the right direction, but it was not nearly large enough.(Emphasis added.)
So, why were the polls wrong in their predictions of the vote shares for Labour and the Conservatives? I have seen some speculation, but no definitive arguments on that question.
But I am troubled — and this won't surprise those of you who have studied statistics — by that "consensus". It seemed to me before the election that the pollsters' results were too similar. That can happen if some pollsters, worrying that they might be wrong, add a little too much "secret sauce" to their results to make them taste like everyone else's.
If you want to think about this problem, you can find the final poll results from eleven pollsters, here. I'd suggest starting, not with Labour and the Conservatives, but with the polling results for UKIP: 12, 14, 12, 11, 12, 11, 11, 12, 16, 14, 15.7. The average of those eleven polls is 12.8 percent; the current UKIP vote share is 12.6 percent,
So whatever affected the poll estimates for the Conservative vote didn't affect the estimates for the UKIP vote.
Any explanation for the errors in the polls should explain that fact, too.
(If you'd like to see a picture showing how the polls missed (and didn't), the Daily Mail has one, in this article.
I don't know whether there are any pure economic models of British elections, and, if so, whether any of them out-performed the models that used polls.)
- 1:27 PM, 8 May 2015
More: If you follow that Daily Mail link, you'll find, in a sidebar, an admission from one of the pollsters, Survation, that they got a result showing a solid Conservative lead (37-31), but didn't publish it, because it "seemed so out of line" with all the other results.
- 2:14 PM, 8 May 2015 [link]
The NYT Is Going To Regret That Headline: May even regret it already. In my copy of the New York Times, which I picked up at lunch, the newspaper headlined the British election results: "Exit poll Sees Gains for Britain's Conservatives, but No Clear Winner".
A little wishful thinking there? Quite possibly. I don't think many at the Times are celebrating the Conservative victory.
Scroll down just a bit to see my own headline for that same exit poll.
(The headline and article may both change, as the newspaper absorbs the result.)
- 12:28 PM, 8 May 2015 [link]
Michelle Lee He Yee Gets A Fact Check Right: A week ago, I criticized her for botching a fact correction, so it is only fair to note that she got this one right.
Granted, Minority Leader Nancy Pelosi's claim that House Republicans had cut the Veteran Affairs budget was an easy fact check to do, but that doesn't mean it isn't worth doing.
To call the funding in the House bill a “cut” is not accurate. The president’s budget proposal is just that: a proposal. The House bill is, indeed, lower than what the president and VA proposed, and what many veterans groups hoped to see. But the House appropriations bill increased VA’s total budget and its discretionary spending over fiscal 2015 levels. This is more money than the agency ever has had, and the total VA funding has grown by nearly 73 percent since 2009. It is also difficult to see how rescissions of bonuses or pay raises would lead to fewer veterans receiving care.Did Pelosi know that what she said at that press conference was false?
I'm not sure, but I can say that she often seems careless about mere facts, that, at best, she often doesn't take the time to check what she says, before she says it.
(I assume that all of you know that bureaucracies routinely ask for more money than they think they will need.)
- 8:33 AM, 8 May 2015 [link]
Who Won The British Election? Maggie Simpson.
Oh, all right. The Conservatives won in England and Wales, the Scottish National Party won in Scotland, and the usual minor parties divided up Northern Ireland in much the usual ways.
Much to the surprise of practically everyone — including me — the Conservatives won an absolute majority. Currently, the BBC is projecting that David Cameron's Conservatives will win 331 seats.
That's just large enough to allow for a few defections, and a few possible losses in by-elections. (In Britain, as in the United States, there is a tendency for the ruling party to lose more special elections than you would expect from party strengths in the individual districts.)
Who lost? Labour everywhere, but especially in Scotland, and the Conservative's coalition partners, the Liberal Democrats. So British voters simultaneously rewarded the Conservatives for their actions in government, and punished the Liberal Democrats, for those same actions.
And the pollsters lost; all of them were wrong.
They are still counting the votes, but the count is complete enough so that you can make some tentative conclusions from the data.
The best single presentation of the numbers that I have found is this page in the Guardian. One surprise: Labour gained 1.4 percent of the popular vote over 2010, while the Conservatives gained only 0.4 percent. As predicted, UKIP won almost 13 percent of the popular vote, and 1 seat; as predicted, the Scottish National Party won less than 5 percent of the vote, and 56 seats. The Greens won almost four-fifths as many votes as the Scottish National Party, 1,146,703 versus 1,454,436, but just a single seat.
(Here's the Wikipedia article on the 2010 election for comparisons, and here's their article, which will be changing rapidly, on this year's election.)
- 6:03 AM, 8 May 2015 [link]
Here Are The Results From the 2010 British Election, if you want to look at the changes that this election is bringing. (Assuming the exit poll is roughly right.)
- 3:03 PM, 7 May 2015 [link]
The Exit Poll Looks Very Good For The Conservatives In Britain: How good? The BBC is projecting that they will win 316 seats, short of a 323 majority, but allowing for several formal, or tacit, coalitions that would allow David Cameron to stay on as prime minister.
To say the least, I am surprised, but will remind you that, in the United States, exit polls have sometimes been quite wrong.
(A reminder on the basic arithmetic: There are 650 seats in the House of Commons. One of them is held by the Speaker, who agrees to be nonpartisan (and is usually not opposed for re-election), so a majority would appear to be 325 seats. However, Sinn Fein usually wins 5 seats in Northern Ireland and those 5 members do not participate in parliament, which makes a working majority 323.)
- 2:26 PM, 7 May 2015 [link]
You Don't Have To Be British To Vote In British Elections: As I just learned, though I have been following these elections for years. Here's what the BBC says are the requirements: "You must be registered to vote, be at least 18 years old on polling day, be British or be a Commonwealth or Republic of Ireland citizen living in the UK".
So, for instance, a citizen of India, who happens to be living in Britain. can vote in British elections.
There are probably a few qualifications besides those three. I would like to think, for instance, that the certifiably insane can't vote there. But I could be wrong.
(Here's a reminder on which 53 nations are in the British Commonwealth. I like to think I could have guessed most of them, but I may be flattering myself.)
- 2:01 PM, 7 May 2015 [link]
British Pollsters Are Mostly As Confused As I Am about who will win today's election, who will be the next prime minister.
Ben Page, chief executive of Ipsos Mori: ‘Toss a coin’There are four of the eight who who think they know who will win, three saying Labour's Miliband and one saying the Conservative's Cameron.
- 1:12 PM, 7 May 2015 [link]
The Guardian Walks Right Up To The Fence Separating Serious Journalism From Parody: And then jumps over it. This headline is almost perfect: "How can our future Mars colonies be free of sexism and racism?"
There's more along the usual lines in the piece. You may be pleased to learn that the author, Martin Robbins, does have some hope for those future colonies.
This would be funnier were it not for this fact: The BBC, probably the most influential news organization in the world, recruits many of its people from, and through, the Guardian.
By way of, among others, David Thompson. His post is followed by a number of instructive and amusing comments. (Don't miss the xkcd cartoon.)
- 9:54 AM, 7 May 2015 [link]
Predicted Votes And Seats In Tomorrow's British Election: The Wikipedia article on the election says, with traditional British understatement: "The first-past-the-post system used in UK general elections means that the number of seats won is not closely related to vote share."
Predicted Votes And Seats In 7 May UK Election, by Party
Sources: For the popular vote, I am using the ICM numbers from the Wikipedia article; for the seat predictions, I using this model from FiveThirtyEight. Both may change slightly by tomorrow.
You can see, I think, why the amount of tactical voting will be important tomorrow. The UKIP popular votes, being spread across England are likely to be almost entirely wasted, whereas the votes for the Scottish National Party, being concentrated in Scotland, will be extremely effective.
Presumably, some voters who prefer UKIP will vote for another party in order to prevent Labour from winning the election, or to keep the SNP from sharing power — but I haven't seen any good estimates of just how many will do that.
(Numbers review: If past election patterns hold in Northern Ireland, 323 seats will be just enough to win control of parliament. The SNP has promised to support Labour — at a price. The Conservatives can probably count on 8 votes from the Democratic Unionist Party in Northern Ireland. The other minor parties are likely to support Labour.
I have no strong reasons for using either source in that table. ICM is a well established pollster in Britain, and the FiveThirtyEight has had some success in their own predictions.)
- 5:18 PM, 6 May 2015 [link]
What Happens When Leaders Don't Compromise And Voters Don't Vote Tactically? Then you get results like yesterday's Alberta election, where a socialist party, the New Democrats, won a big victory in Canada's most conservative province. The Conservatives had won every Alberta election since 1971, which will give you an idea just how big that victory was. Or, you can look at the number of seats the New Democrats won: 53 of 87.
Why did the New Democrats win that big victory? Because the conservatives split, with a conservative party, Wildrose, breaking away from the established Conservatives. Yesterday, the Conservatives won 27.78 percent of the popular vote and 10 seats; Wildrose won 24.23 percent of the popular vote and 23 seats. (The New Democrats won 40.57 percent of the popular vote.)
As I have said before in similar situations, you can't just add up the votes and conclude that, if there had been a single conservative party, it would have won. But I think that, considering past elections, if the leaders of the two parties had worked together — if they had, in short, behaved as adults, they would have won.
(One short-term possibility would have been a "non-compete" agreement. When parties make such agreements, each party agrees not to compete with the other in some seats.)
Or, the parties would have won if their voters had voted tactically, voted for Wildrose where that party had the best chance of defeating the NDP, or for the Conservatives, otherwise.
But instead they voted with their hearts, rather than with their minds.
For a rough American equivalent, imagine the Texas Republicans splitting, and a far-left Democratic Party taking control of the state.
(I wish I could give you a quick explanation for the split, a quick description of the issues and leaders in this split — but I can't. Oddly enough, the newspapers I read regularly don't often cover Alberta politics.
But I can explain the name of the party; Alberta's official flower is the wild rose.)
- 2:16 PM, 6 May 2015 [link]
How Much Tactical Voting Will There Be In The British Election? The Daily Mail is hoping there will be a lot, and has even provided seat-by-seat directions.
Labour’s collapse in Scotland means Ed Miliband will almost certainly miss out on the 323 seats needed for a majority.Even if you are only mildly interested in the British election, you'll probably find it interesting to glance over their recommendations, if only to understand the thinking behind them.
I have no idea how much tactical voting there will be, but I agree with the Mail that — if there is enough — it will keep the Conservatives in power.
(Here's the Wikipedia article on tactical voting, which almost certainly includes more than most of you will want to know.)
- 10:07 AM, 6 May 2015 [link]
Why Is Outgoing Nigerian President Goodluck Jonathan Attacking Boko Haram Now? With considerable success.
Outgoing Nigerian President Goodluck Jonathan isn't acting like a lame duck when it comes to going after Boko Haram, the murderous Islamist army that has terrorized the northern part of Africa’s most populous country.Here's my cynical speculation: Before the election, Jonathan was trying not to offend Muslim voters by a vigorous campaign against the Islamic terrorists. After the election, he had no reason to pander to them, and is now doing what he should have done all along.
Whether that speculation is correct or not, I think we can all agree that armies are more effective against terrorists than hash tags, that sending large bodies of armed men to kill terrorists works better than displaying signs saying "Bring Back Our Girls".
The delay in opening a serious military campaign against Boko Haram meant that many of those girls suffered in ways I don't even like to think about.
(There are a few interesting comments after that Fox article about what the United States has done (not much) to help Nigeria, and what we could have done. As secretary of state, Hillary Clinton blocked military aid to Nigeria, a policy partly reversed by John Kerry. She believed, correctly no doubt, that the Nigerian army was corrupt, and sometimes treated the population badly. But you can't always pick your allies in wars.
Here's the Wikipedia biography of Goodluck Jonathan, with more than the usual caveats.)
- 9:10 AM, 6 May 2015 [link]
Two More Firsts In Space: The first Italian woman astronaut — who celebrated by making the first espresso in space.
Astronaut Samantha Cristoforetti boldly brewed where no astronaut has ever done so before -- on board the International Space Station. As if she couldn't get any cooler, Cristoforetti wore her Star Trek uniform as she sipped her cup of espresso from a cup designed for zero gravity use. The Italian astronaut shared the sweet snap on Sunday with her Twitter followers.And she probably shared the espresso with the rest of the crew.
(This ABC article doesn't mention it, but the New York Times article, where I first saw this story, says that she is the first Italian woman in orbit. The Times article also includes a a partial description of the espresso machine made for this task. It weighs 44 pounds (or, more likely, approximately 20 kilograms), which is a lot for cooking equipment sent into orbit.)
- 12:39 PM, 5 May 2015 [link]
Finland Is Getting Nervous About Russia: And showing it by military actions and preparations.
At about 3 A.M. local time last Tuesday, ships of the Finnish Navy dropped depth charges into the waters near Helsinki in the vicinity of what has been reported to be a possible Russian submarine. The Finns initially refused to confirm that this “possible underwater object” was in fact a submarine, let alone a Russian submarine, and a spokesman was later quick to emphasize that the depth charges were “not intended to damage the target,” but to “let the target know that it has been noticed.”Of course not.
Finland's history gives the Finns good reason to be wary of Russia.
- 8:04 AM, 5 May 2015 [link]
Matt Drudge Is Playing Juxtaposition Games Again: As I write, the Drudge site has the following five items at the top of its right column: a picture of Obama, a story on the predominance of government jobs in Baltimore, a story on Attorney General Loretta Lynch's plan to visit Baltimore, a story on a CBS/NYT poll showing worsening race relations, and a story on Obama's plan to continue his "project" after his presidency.
It isn't hard to understand the argument Drudge is making; government polices have made race relations worse under the Obama administration, and Obama will continue his efforts to make them worse, after he leaves the presidency.
(Do I think that implicit Drudge argument is correct? Yes. In 2008, when I saw that Obama had Al Sharpton as a close ally, I thought it likely that Obama's policies would, on the whole, tend to divide the races. It was possible, I thought, that Obama would use Sharpton and then abandon him, but not likely. And Obama's past experiences as a "community organizer" suggested that he would seek to divide us, because that's what community organizers do.)
- 6:55 AM, 5 May 2015 [link]
UCLA Professor Volokh Supports Free Speech: At one time, the idea that a UCLA professor would support free speech — and would say so, publicly — would not have been news, but now it is, not big news, but still news.
You can see that support in his gentle reproof of our two Muslim congressmen, Keith Ellison and André Carson, who called for the Dutch legislator, Geert Wilders, to be banned from this country, for his views on Muslims.
Here's Volokh's conclusion:
Whether “Christian culture is superior to other cultures,” which groups should be allowed to immigrate into a country, and even whether Islam should be viewed as an ideology rather than a religion (an unsound distinction, in my view) are matters that the First Amendment allows us all to debate. The Congressmen quite clearly don’t want to allow Rep. Wilders to debate such matters here in the U.S. But their “In the U.S.” paragraph suggests that they view even such debates by Americans as constitutionally unprotected.You can decide for yourself whether Volokh should have mentioned the congressmen's religious beliefs, or their frequent association with radical Islamists.
Whatever you decide, I hope you will join me in admiring his stand, which would not be approved by all of his colleagues at UCLA. (I suspect a fair number of them agree, in part, with Volokh, but wish he would not say such things publicly, not about Muslims, anyway.)
- 5:09 PM, 4 May 2015 [link]
Females Are More Likely To Ask People For Directions: That's not news, you may be thinking, but the females I am talking about are female dogs.
Researchers set more than 400 pedigree beagles an impossible puzzle. They were put in front of three boxes, each of which had a clear lid and contained a biscuit.The researchers seem to think this shows the superiority of females, just as some people see the greater willingness of women to ask for directions as showing superiority. Actually, I think both strategies have advantages and disadvantages. Asking for help (or directions) may solve an immediate problem more quickly; trying to solve it by ourselves may let us learn something that will help us in the future.
- 7:42 AM, 4 May 2015 [link]
Does Ed Miliband Think He's Moses? The Labour leader made one of the oddest campaign promises I've seen:
Just days before one of the closest General Elections in years Ed Miliband will no doubt be delighted to discover he is currently the number one talked about topic on Twitter - but his joy will be short-lived when he realises he is not necessarily trending for the right reasons.And, naturally, the Daily Mail has a few pictures of them to show you.
One of the pictures may need an explanation for American readers. The face on the naked woman is that of the head of the Scottish National Party, Nicola Sturgeon. Few expect her party to play an entirely constructive role in the election, and possible negotiations, afterward.
- 7:10 AM, 4 May 2015 [link]
The Betting Odds On The Next British Prime Minister Have Now Shifted so that some bookies are offering the same odds on Conservative David Cameron and Labour Ed Miliband.
(Currently, five firms are giving these odds for both men: 10/11. I am guessing that means that if you bet 11 pounds on either and he wins, then you win 21 pounds. After all, there has to be a little profit for the bookies.)
As I said before, I don't plan to make any predictions, because I know so little about individual constituencies, and because the race is so complex. To make a good prediction, you would have to do a seat-by-seat analysis of the "marginals", as the British often call them. And make a guess at how many voters will decide, in the last few days, to cast "tactical" votes, to vote for their second or third choice parties.
- 6:37 AM, 4 May 2015 [link]
The Maduro Government Is Making It Difficult to export cocoa beans.
The quality of Venezuela’s cocoa is legendary. Cocoa plants all over the world have been cross-bred to make them resistant to plagues, but ours have been preserved, so much so that our cocoa is often times heralded as the world’s best.Why is the Venezuelan government doing this? Reuters couldn't find out, couldn't even get an answer from Maduro and company.
So I will turn to pure speculation: Perhaps one or more of the producers did or said something that annoyed the regime.
- 7:36 PM, 3 May 2015 [link]
Socialists Good, Anarchists Bad: Seattle had its annual May Day riot, as everyone expected. It was smaller and less destructive than many past Seattle riots.
Perhaps it was that smaller scale that made me notice something I should have noticed years ago: For our local journalists, socialists are good, but anarchists are bad.
Here's an example of the first: One of our local news readers showed us a picture of Fidel Castro meeting Venezuela's Maduro for a celebration of May Day. She was obviously pleased by the meeting and the recent Obama-Castro rapprochement. (Is she misinformed about Castro and Maduro, or does she actually like what the two men have done to their countries? I don't know, but I suspect it is mostly the first.)
Similarly, the anti-police demonstration and the amnesty-for-illegals demonstration on May Day here (which were mostly peaceful) received the usual respectful coverage.
But our local journalists don't like the anarchists, and when they began to demonstrate around 6 PM on May Day, the coverage became negative. There were none of the these-people-have-legitimate-grievances excuses that you hear, routinely, in other kinds of leftist demonstrations. Instead, there were mostly these-people-are behaving-badly, for-no-good-reason stories, and luckily-the-police-seem-to-be-prepared-for-them stories.
I am not entirely sure why the local journalists dislike anarchists so much; perhaps it is simply that the journalists want to use the power of the state, and the anarchists want eliminate that power.
(The coverage was extensive, with stations dropping their ordinary Friday night programs to show us pictures, from the ground and the air, of what was going on in Seattle, and how the police were trying to keep the anarchists away from downtown Seattle, and Interstate 5.
If you are feeling serious, you can take a look at the Wikipedia article on anarchism — or you can just appreciate this joke:
Why do anarchists use only lower case letters?
Because they are against capitalism.
The "proper tea" joke is pretty good, too, but perhaps little too English for a general audience.)
- 7:05 PM, 3 May 2015 [link]
Worth Reading: In spite of the misleading headline, and the muddled presentation of the data. (Which, to be fair, has problems.)
Here, however, is the bottom line of this front page New York Times article.
A number of criminologists believe police homicides are near their nadir. In New York City, for example, 91 people were fatally shot by police officers in 1971 — and a record-low eight in 2013, the last year for which figures are available. In Los Angeles, officers used “categorical” force — gunfire, chokings and other violence that could lead to death — in 84 of nearly 149,000 arrests in 2012, down 17 percent in seven years.(Michael Wines and Sarah Cohen appear to be using absolute, rather than per capita, numbers, in their comparisons. But, because of the rise in the population of the United States, just over 200 million in 1970, and about 320 million now, one would expect the total shootings by police to rise, by about 50 percent.)
So, almost certainly, killings by police have been declining for decades, and the decline has been dramatic since crime peaked decades ago.
Some of that decrease, just like the general decrease in murder, has been caused by improvements in emergency care. Many people who would have died years ago now survive, though they may be crippled.
Since this issue is prominent because so many of our political leaders and "mainstream" journalists are obsessed with race, I'll add this from further on in the article.
“Blacks are three times as likely to be killed by cops as are whites, on a per-capita basis,” said Dr. Moskos of John Jay. But part of that is because of crime in predominantly black neighborhoods.In general, I'll repeat what I have said before: If a trend is good, you should probably continue doing what you are doing. These declines show that our police have been doing something right in recent decades, though it may not be clear what they have been doing right.
(The article would be much better, if it were illustrated with a simple graph or two, along, of course, with some caveats about the quality of the data.)
- 2:13 PM, 1 May 2015 [link]
How To Do Fact Checking, And How Not To: Glenn Kessler of the Washington Post gives us an example of the first.
It may be the case that Baltimore schools should get more federal funding — or that too much was spent in Afghanistan, with not much impact. But [Jon] Stewart suggested that an extraordinary amount was spent on schools in Afghanistan while Baltimore did not even get a “taste.”(Full disclosure: I've never watched Stewart, so I can't say whether Kessler is correct in his generally positive assessment of the man.)
Michelle Lee He Yee of the Washington Post gives us an example of the second. Senator Ted Cruz noted that the IRS tax code is longer than the Bible, and Yee spent an entire column addressing that common comparison, finally concluding:
This is a nonsense fact, something that is technically correct but ultimately meaningless. Thus it is not worthy of a Geppetto Checkmark but neither does it qualify for a Pinocchio.Yee concludes that what Cruz said was true — but that he shouldn't have said it, which is not what a "fact checker" ought to be doing.
Nor is it correct to say that the comparison is "meaningless". We use word lengths (or, a close substitute, page counts) all the time to compare reading and comprehension times. It's an imperfect measure, but not a meaningless one.
(My own view: The comparison of the Bible to the IRS code has been used so many times, that, were I Cruz, I would have looked for a different comparison, just to keep my audience awake.)
- 7:37 AM, 1 May 2015 [link]
The WSJ Fills In Another Piece Of The Warren Weinstein Ransom Puzzle: When I read their first story on the attempt to ransom the al Qaeda captive, I assumed that the family had made the attempt, because of the amount of money paid, $250,000. It sounded like what a family like his might be able to raise, not the amount available to a government or a wealthy benefactor.
But what puzzled me was how the family had found a middleman, something I doubt most of us could do. (I'm not even sure how I would go about finding a trustworthy middleman in Pakistan, even if I had months to do so.)
Yesterday, the lead story in the Journal explained how the family checked out that middleman.
The Federal Bureau of Investigation helped facilitate a 2012 ransom payment to al Qaeda from the family of kidnapped aid worker Warren Weinstein, senior U.S. officials said, in an unsuccessful bid to secure the release of the American, who was killed in January in a U.S. drone strike.It isn't hard to think of how the FBI might have enough contacts in Pakistan so that they could vet this middleman, and it is understandable that officials in the FBI, wanting to help the family, might violate US policy.
What is still unknown is whether the middleman was, in fact, trustworthy, but was conned by al Qaeda, or whether he was simply a scammer, whether, in short, he was a victim or a con man.
(If you read the whole article, you will probably conclude, as I did, that the FBI does not agree with official US policy, and has undermined it more than once. This won't surprise anyone familiar with bureaucracies, which often go their own ways.)
- 5:42 AM, 1 May 2015 [link]