Moonbat in flight
Ticker November & December 2012

31 December 2012 - New Years' Resolutions - updated 8 January 2013

Today is the day we compose the last draft of those resolutions that come into effect tomorrow! No more on-line solitaire. No more swiping candy from the receptionist's bowl. Fifty sit-ups a day.

The trick, as psychologists repeatedly tell us, is practice. But practice is hard, and it is much easier to get a self-help book from the local bookstore and read, for the five thousandth time, that practice makes perfect. Indeed, the self-help book industry brings to mind the old line Roman doctors had about their patients' skin disorders: they never die, and they never get better - it's perfect!.

This attitude is not consistent with the, um, rationalist slant of The City, which bills itself as "the nationís premier urban-policy magazine," although it is also a creature of the right wing Manhattan Institute for Policy Research - a think tank dedicated to the proposition that people follow their own rational self-interest and if they don't then it's their own fault.

The City ran a column on The Paperback Quest for Joy: Americaís unique love affair with self-help books by Laura Vanderkam, who delved into the history of self-help books - which are a lot older than Vanderkam seems to realize (a substantial fraction of classical Chinese and Buddhist philosophy consists of self-help) - but she flounders around the issue of their effectiveness with the following classical economics boilerplate: "self-help readers rarely do everything that a book advises; after examining their own lives, they use their judgment to decide whatís worth trying." This is probably rubbish - how many readers deliberately decided that the RCAF 11-minutes-a-day program for men or the 15-minutes-a-day program for women were not worth trying or continuing? A sideways glance at popular humor suggests that people found them worth trying, and tried them for a few days, and then subsequently ... they never got around to it.

This is a sign of irrational human behavior, something that classical libertarian economists tend to regard as a mythological beast whose existence is posited only by left-wing charlatans. But we do see it every year, this time of year, don't we? And the vast self-help literature is not that much help. Meanwhile, I've come to the rational conclusion to quit on-line solitaire, cold turkey. If the classical libertarian economists are right, it should be easy as pie. Here goes ...

*              *              *

29 December 2012 - Happiness and Helping People

Is it a good idea - pragmatically speaking, not idealistically - to go out of one's way to help people?

This is a longstanding bone of contention between those committed to random acts of kindness, and those convinced that nice guys finish last. But this debate isn't confined to people on barstools: social scientists can actually collect data.

One of the recent studies looked at preadolescents (yes, this is what a scholarly social science article looks like). 415 schoolchildren, ages 9 to 11, spent four weeks committing three random acts of kindness daily, keeping a log of what they did, and also what they felt. Meantime, they and their peers were surveyed about who was popular and who wasn't. At the end of the four weeks, the kind students had gained an average 1.5 friends, and reported being happier. And their popularity increased.

The authors had been worried about bullying, and in their conclusion they recommended that students be taught that helping others (as opposed to picking on others) makes one more popular. But this study, like others in the literature, suggests that the advice to help others may have a pragmatic justification.

*              *              *

22 December 2012 - First Glance at the Florida Senate - revised 21 January 2013

The final election results for the November 6 General Election by precinct were supposedly reported to the Florida Division of Elections by December 20, and the IT staff there believe that they can get the results posted by January 4. While we wait, we can continue to get an outline of what's going on by looking at the Florida legislative elections, following the lines we set in the 22 November 2012 posting. Today, let's look at the Florida Senate.

Florida has forty senate seats, and of these, ten (# 5, 17, 18, 19, 23, 28, 36, 37, 38, and 40) were uncontested in the last election.

  • Eight of these ten uncontested seats are now occupied by Republicans, and the remaining two by Democrats. (Cynics may observe that the two Democrats are minorities.) This means that for those districts, the opposing party was unable to find anyone willing to run.
  • For the remaining thirty seats, three (# 2, 22, 27) were between a party candidate and a write-in candidate (# 2 and # 22 were Republicans, # 27 was Democratic). As is usual, the write-in candidates got uniformly microscopic support. More to the point, these were also districts in which the Democratic Party was unable to find anyone willing to run.
  • That leaves two, for districts # 1 and 11, between a Republican and a candidate of No Party Affiliation. In District # 1, Richard Harrison, NPA, won 25.59 % of the vote against incoming Senate President Don Gaetz, who got 74.41 %. In District # 11, John Iler, NPA, won 28.44 % of the vote against 71.56 % for Republican Alan Hays. These results suggest that in such districts, solid moderates with party support may have a chance.
So there were no major party opposition candidates for fifteen of the forty senate districts; in fact, there were no Democrats in twelve of the forty races. There are two explanations that these analyses were supposed to address:
  • Despite the Fair Districts amendments, the Florida Legislature managed to gerrymander the district lines so as to remain in power and keep the Democrats behind the eight ball.
  • In spite of the opportunities provided by public unhappiness with a discredited governor and an extreme legislative leadership, the Florida Democratic Party failed to make sufficient effort in this election and therefore remains behind the eight ball.
We should keep these two explanations in mind during this and coming analyses. These two explanations could explain the above as follows.
  • The gerrymandering was so severe that the above thirteen districts were so Republican (and the other two so Democratic) that credible and willing opposing candidates could not be found. But this would suggest that the very substantial number of Democrats in Florida are in the remaining twenty-five districts.
  • The Democratic Party did not encourage or recruit candidates for the thirteen districts in question. It is not clear at this point what this explanation would say about these districts.
Of course, the reality could be a mixture of the above explanations. Anyway, we have reached the twenty-five of the districts in which a Republican faced a Democrat.
  • In these twenty-five races, 2,602,316 voted for the Democratic candidate and 2,440,297 for the Republican. There were no No Party Affiliation candidates in these races, and only nine write-in votes, so only the Democratic-versus-Republican votes matter here. Notice that this means that in these twenty-five races, 51.6 % of the votes were for a Democrat and 48.4 % were cast for a Republican.
  • The results for these twenty-five races were: eleven Democrats won and fourteen Republicans won, i.e. 44 % of the Democrats won and 56 % of the Republicans. Not nice, but for only twenty-five districts sprawled across a diverse state, we will need additional analysis to guess how fishy the results were.
The bottom line is: Democrats skipped twelve races (of uncertain prospects) and now hold only fourteen seats (out of forty) in the Florida Senate.

*              *              *

30 November 2012 - Secret Societies

So what are secret societies really like? We are familiar with predatory secret societies like La Cosa Nostra and the thuggee, but fans of Dan Brown's political fantasy novels may get a somewhat overblown impression of, say, freemasons (a secret society that included Benjamin Franklin and George Washington among its members, and was described in Leo Tolstoy's War and Peace - and in one of Monty Python's skits).

A glance at a more typical society may be provided by a recently decrypted text used at an initiation ceremony of an "Oculist" society in the Eighteenth century. European governments were then far more paranoid about new ideas than they are now, and many of these societies were essentially forums where "respectable" society could meet without getting into trouble. But they were also places where new notions about science and politics were discussed, and they played a role in the political revolutions of the Eighteenth and Nineteenth century.

Subsequently, all that ritual & paraphernalia attracted imitators from college fraternities & sororities & honor societies to Hollywood. And the KKK, complete with ersatz masonic symbols and pseudo-Druidic robes.

*              *              *

25 November 2012 - The Marriage Racket

Marriage has been a contract ever since two Cro-Magnon chiefs arranged a wedding between their children in order to cement a hunting agreement. Contracts are matters of law, and laws are a matter of government, so of course the government is involved. But during the Renaissance, the poets foisted this notion of marriage for love on us (remember Romeo and Juliet?), and since we're talking about living with someone, which is about as personal as it gets, we're talking about rights. And rights are also the affair of government.

Our government has gotten itself into a pickle. The standard interpretation of the Fourteenth Amendment is that the Federal government will protect the rights of all citizens, equally - that equal figuring prominently in Bush v. Gore, which is a precedent unless the U. S. Supreme Court is utterly without honor. But over the years, the governments have banned bigamy, polygamy, polyandry, interracial marriage, and, oh yes, gay marriage. Most of these bans were imposed by the states, but the Fourteenth Amendment explicitly trumps states rights (it was passed after, and therefore modifies, the Tenth Amendment (on states' rights)), and therefore unless the U. S. Supreme Court wants to make a spectacle of itself, all these bans will have to go.

What a mess. Law enforcement and prosecutors in Arizona, Nevada, New Mexico, Texas, and of course, Utah, are waiting to see if the current court is going to throw the rubbish out, or cling to this Nineteenth century garbage heap. (And in Salt Lake City, the elders are wringing their hands, wondering what on Earth they're going to say if the unspeakable happens.) And the unspeakable had better happen, for if the Defense of Marriage Act is constitutional, so are the Anti-miscegenation laws.

Perhaps the time has come to revisit the notion that marriage is some special kind of contract, and instead let people (and their lawyers and paralegals) draw up standard marriage contracts for those couples, triples, etc., who feel the need for pre-nuptial agreements - but otherwise get the government out of the marriage business. The government has a legitimate duty to protect children and hear contractual lawsuits (after all, the real threat to marriage - and the only marriage issue addressed explicitly by Jesus Christ - is divorce), but other than that, leave it to the churches, synagogues, temples, commercial establishments, etc.

*              *              *

22 November 2012 - Was We Robbed? - revised 19 January 2013

Now that the election is over and the Republicans kept the House, the Democrats kept the Senate, the Democratic presidential candidate narrowly won the White House (winning the electoral college by a larger margin) and the Republicans held the statehouses (by larger margin), the big question would seem to be: did either side actually win? But the pundits and politicians and even many of the people have decided that the Democrats won. In fact, that loud whining sound is the Right, squealing denials and even secession - the latter winning the inevitable snarky comment that since it's the red states that balance their budgets on the federal dole, they're not in any position to go it alone (Greece, anyone?).

But as all politics is local, the alleged Democratic victors are looking at the statehouses and wondering whether they really won. Not only did CNN observe that House Democratic gains may not equal victory, but the fact that the House districts are drawn by state legislatures led Washington Post columnist Harold Meyerson to Credit gerrymandering for GOP House control. More to the point, his colleague Aaron Blake cites Post staffer Dan Keating's conclusion that of all the votes cast for House candidates, 48.8 % went to Democrats while 48.5 % went to Republicans, and Blake concludes that if Democrats got more votes but Republicans won more races, it's because Republican legislatures gerrymandered the districts.

Let's take a closer look at Florida - because that's where I live. Florida is not really part of the Deep South - although it was the Confederacy's source of meat during the Civil War (the original "cracker" was a Florida cowboy, driving his cattle to the Georgia market with a whip) - and it was solidly southern Democratic up into the 1990s. It was a Democratic legislature in 1990 that refused to reform redistricting practices. During the 1990s, the Republican Southern Strategy finally took hold, with Jeb Bush winning the gubernatorial race in 1998 while the legislature turned Republican; the statehouse has been Republican for over a decade. Right now, the governor and his elected cabinet are all Republicans, and the only Democrat to hold a statewide elected position is (recently re-elected) Senator Bill Nelson.

After being gerrymandered by Democrats in 1990 and by Republicans in 2000, a grassroots movement (with Democratic and out-of-state support) pushed through two somewhat vague and not particular tough Fair Districts amendments that required that the Legislature be fair in redistricting. The amendments did have an effect - including litigation - and the Tampa Tribune reports that Fair Districts helped Democrats make gains in Florida election - although Democrats, Republicans cheer gains in Legislature.

But in the election, Obama carried Florida (very narrowly) while Nelson won re-election (not narrowly at all), which raises the question, why did only eight Democrats win House seats while nineteen Republicans did? Perhaps gerrymandering, perhaps (as the Tribune insinuates) something else, perhaps a bit of both.

First, the basic facts from voter registrations and the basic election returns.

  • As of November 6, 39.8 % of all voters were registered as Republicans, 35.6 % as Democrats, 21.7 % with no party affiliation, and the rest in minor parties. Total registration was 12,390,280.
  • On November 6, among those who voted, 50.0 % voted for President Obama while 49.1 % voted for Governor Romney, and 8,474,179 votes were cast in that election.
  • Also on November 6, 8,189,946 votes were cast in the race between Senator Nelson, who won with 55.2 % of the vote, versus Congressman McGillicuddy ("Mack"), who lost with 42.23 %.
  • And on November 6, 7,513,534 votes were cast in the Congressional elections. 50.9 % were cast for Republican candidates, while 45.2 % were cast for Democratic candidates.
So if the (Republican) legislature had drawn the lines fairly, one would expect perhaps thirteen of the 27 candidates elected to Congress to be Republican, twelve of them to be Democrats, and maybe one independent. Of course, we all know that this won't happen: votes for independents are scattered about, so we actually shouldn't expect any independents to be elected. What we would actually expect is about half Republican and half Democratic.

Actually, we got seventeen Republicans and ten Democrats. But since two candidates faced no opposition, and hence walked into office without an election, we elected sixteen Republicans and nine Democrats. Does 64 % for the Republicans and 36 % for the Democrats mean that the districts were gerrymandered? How do we answer a question like that?

The simple-minded statistical test would be to propose as a null (or default) hypothesis that the districts had been so fairly drawn that the distribution of Democratic and Republican votes in the 27 districts was effectively random. This is the standard statistical assumption, but in this case, does it does this make sense? Remembering that we are assuming this extremely fair districting, let's go through the computation step by step: if there are 27 districts casting perhaps 278,000 votes each, we can use the old Normal approximation to the Binomial distribution to estimate how many Republican congressmen would get majorities in their districts just with 50.9 % of the vote (never mind winning with pluralities): (Hold onto your hats, because people - like supreme court justices - who are unfamiliar with how statistics works may find this very odd ...)

  • There are 278,000 votes, 50.9 % for Republicans, 49.1 % for non-Republicans, giving a variance of 70,000, and hence a standard deviation of 264.
    (The variance and standard deviation measure how far away the observation is from the expected value that we would expect to get; when statisticians talk of statistical fluctuations, they usually mean that they aren't surprised when the observation is one, two, or even three standard deviations away from the expected value - but they are surprised when the observation is further away than that.)
    (When a supreme court justice talks of statistical fluctuations, it often is just an attempt to wave away an inconvenient statistic.)
  • The expected value of Republican votes for any district is 0.509 times 278000 or 141,500 votes, while a majority is anything over 139,000 votes.
  • 141,500 - 139,000 = 2,500 votes, which is 9.5 standard deviations. For a Republican to lose in any district would mean that the vote would be off by 9.5 standard deviations. What does this mean?
  • The Normal approximation to the Binomial breaks down when computing the probability of being as many as 9.5 standard deviations away from the expected number (at least for a population of over 278,00 voters): the probability of being as much as 9.5 standard deviations away from the expected number (in the Normal distribution) is a thousand billion billions to one.
  • My Maple program took 26 seconds to compute one of the 139,000 terms it would have to compute to find the precise probability of a Republican losing. So I will just do the usual thing ...
  • Usually, one uses the Normal Approximation to three standard deviations (which is reasonable for 278,000 votes), and the probability of being off by three standard deviations - never mind nine standard deviations - is one in 741. Three standard deviations from the expected 141,500 votes would give us 140,700 votes - a majority - and there better is better than a 96 % chance of Republicans getting a such a majority in all 27 districts.
Huh??? In real life, a party with a narrow advantage does not walk away with all the electoral victories, so something odd is happening. And in this case, we know there is a lot of oddness: no Democrats ran in districts 4 or 25, and no Republicans in districts 20, 21, and 24: nearly a fifth of the districts were effectively uncontested. Then sitting back and thinking about it: voters are not uniformly distributed so that such super-fair districts can be drawn; after all, the Old Money in south Tampa is very different from the Diverse Community of south St. Petersburg, so the voters in those two areas will vote for different people. The statistics are not that simple.

This is the first of a series of postings on the fall 2012 returns for Florida's congressional and legislative elections. In future postings, we will look into the question of how congressional, state senate, and state house districts were drawn and how that affected the size of the party delegations elected. En route, we will show something of how statistics - and gerrymandering - work.

*              *              *

14 November 2012 - And You Thought Order of the Phoenix was Fiction

We often assume that there is a fundamental difference between Americans and the European and Asian ideologues who broke their academies and made them into lapdogs. Academia has long relied on the ancient immunities granted (grudgingly and not reliably) to prophets, sages, jesters and minstrels -- immunities mentioned repeated in the Bible (especially for prophets). But as Joseph Stalin - a noted breaker of academies - would have asked, how many troops do the academics have?

The evisceration of Shorter University by its parent Georgia Baptist Convention is a reminder of Stalin's point. According to the report in Inside Higher Ed, a new administration, imposed by the Convention, has cleansed its undergraduate campus of 35 of its 94 full-time faculty, using tools familiar to academics who have gone through this sort of thing: and ideological statement everyone is to adhere to, and a list of proscribed behaviors, ranging from an actual crime to a thought crime -- and none of this (as if it matters) is scriptural.

Fortunately, the Convention has no powers other than the ability to wreck a university. But it is a reminder of what such organizations do when they get even greater powers. It is also a reminder to academics in similarly situated institutions of Mencius' warning that when the signs turn ominous, the wise men leave.

*              *              *

12 November 2012 - Reality Check

Ever since 1964, each Right wing electoral debacle would be followed by a discussion about when the Republican Party will free itself from that millstone around its neck and return to being the party of Abe Lincoln and Teddy Roosevelt. This time is no different,even though the election results - Obama scored a narrow popular win over Romney and held the senate, while his coattails swept a handfull of state politicians into office - could hardly be called a Right wing debacle. The reality is that even a toxic crank like Allen West can win over 49 % of the vote, and indeed, in state and local races (where the GOP historically does well), the GOP controls 27 legislatures (to the Democrats' 19, and only twelve states with divided legislatures); in many of these states, the legislative leadership (like Florida's) is troglodyte Right.

Put bluntly, the Southern Strategists who tried to use the machinery of state government to suppress the vote are still in office. And of course, they are unrepentent and have every intention of doing it again.

Still, hope blooms eternal in the progressive breast. James Carville has long been claiming that a rainbow of diverse youth will change the game, and writes that the GOP was routed by reality, while the Tampa Tribune reports that Young, tech-savvy residents changing political landscape and thus making the GOP's predicament worse. Meanwhile David Horsey gloats - er, writes - that Right wingers careen into craziness to explain Obama's victory.

Yet the most realistic view may well be Rick Santorum's, who apparently is following Charlie Brown's advice by vowing, Just Wait 'til Next Year. History suggests that Santorum is right.

*              *              *

8 November 2012 - Calling the Election

Statistics arose as a tool for demographers and insurance companies. It's great success story was modern medicine: during the Nineteenth century, doctors discovered that by tracking diagnoses, treatments, and outcomes, they got a far better picture of what worked and what didn't than any pile of armchair theorizing and anecdotes. As statistics became hot, it naturally got into horse racing and politics.

Political polling began roughly the same way demography itself did: just as a Seventeenth century demographer would pore over ancient parish records, ticking off the births and deaths, so Nineteenth century pollsters would stand on street corners, asking passersby if they were voting for Mr. Adams or General Jackson. This was a great stunt for the media, and starting in 1916, the Literary Digest conducted polls to project the winners of presidential elections : the Digest would send out umpteen jillion postcards and tally up the responses. This method predicted the victories of Wilson in 1916, Harding in 1920, Coolidge in 1924, Hoover in 1928, and Roosevelt in 1932; and with 3.2 million returned postcards in hand, the Digest predicted Alf Landon's victory in 1936.

So 1936 was the year that George Gallup introduced America to the complexities of statistical analysis. The Digest's problem was that a poll requires a random sample of prospective voters: if one relies on, say, automobile license registries, phone directories, or subscription lists to literary magazines, that excludes many voters and is thus biased. And since 1936, sampling, bias, and other jargon of statistics has crept into horse racing and political handicapping.

Political pundits are very conservative in the sense that they prefer to use the tools that they know, and they don't know much statistics. Pundits have gotten used to polls, so they can speak authoritatively on television about trends appearing in sequences of polls, but like Nineteenth century doctors, they are full of armchair theorizing and anecdotes (not to mention, ahem, a habit of letting their personal preferences affect their prognostications). Meanwhile, over the last few years, a number of statistical types have discovered that in the computer age, there is a lot of data out there, and a lot of available number-crunching capability. (Indeed, this is the year that Mathematics Awareness Month features Big Data.) The result is statistical "meta-analysis", which involves mathematical techniques for combining sequences of studies to draw global conclusions.

The competition between meta-analysis and the pundits is getting hot. The pundits are scornful: Peggy Noonan wrote that "nobody knows anything" about the 2012 election, David Brooks derided meta-analysis as "sillyland", and Jennifer Rubin wrote that "averaging polls is junk". David Horsey more sensibly wrote that such predictions "... reveal the fears, anxieties and hopes that many Americans feel on election day". This last point led to some right wing criticism of an observer who predicted a narrow Obama victory.

As Tom Bartlett in the Chronicle of Higher Education (subscription required) reported, the pundits claimed that the 2012 presidential race was too close to call, while the meta-analysts were consistently claiming that Obama was ahead by a nose. John Sides predicted at least 303 electoral votes for Obama, Nate Silver said that Romney had an 8 % chance of winning, Drew Linzer gave Obama 332 votes, Sam Wang vowed to eat a bug if Romney won Ohio, and Simon Jackson claimed his model won 51 out of 51. So Tom Bartlett (subscription still required) reported that "The Poll Quants won the election".

As Bartlett observed, this was not the first victory for meta-analysis. But Bartlett is over-optmistic when he suggests that the pundits will wake up and smell the results (although the Christian Science Monitor's DC Decoder raised the issue). As one of the commentators on his article observed, punditry is really a form of entertainment. Besides, FAIR has been complaining for years about the consistently bad predictions of many leading pundits, and they are still leading because they are very good at what they do - which is (1) entertaining us and (2) keeping us up on what the Beltway is thinking.

And the Beltway is still phobic about statistics.

Footnote: Florida also gave pollsters a Truman v. Dewey moment. The Tampa Bay Times reports that several pollsters put Romney well ahead of Obama in Florida, and quoted Mason-Dixon pollster Brad Coker saying that Mason-Dixon had repeated Dr. Gallup's 1948 error in not keeping up with the voters. Meanwhile, Florida's performance was again an embarrassment, and we're still counting ballots...

*              *              *

3 November 2012 - Racism and Obama, revised Nov. 4

As we approach Election Day, CNN's John Blake reviews, for about the five zillionth time, the possible racism of President's opponents. He compares modern attacks on Obama - his birth certificate, his religion, his alleged socialism - to attacks on black politicians during the Reconstruction. Many historians find noticeable parallels, and so the argument to Obama's more hysterical critics seems to be, if the shoe fits ...

On the other hand, Obama's aloof and occasionally arrogant attitude played a major role in landing him in his predicament. Even the New Yorker, one of his most reliable supporters, has complained about his failure to schmooze , which is part of his more general inability to deal with people of differing views (a failing addressed in the New Yorker's endorsement). His treatment not only of the Left but also of Keynesian liberals and base Democrats has bordered on the contemptuous and dismissive, while he seems unable to get any support across the aisle. The result is an almost toxic distrust: the Democratic base is already quietly preparing for the possibility of Obama throwing benefits under the bus after a victory.

In addition, toxic partisanship is as American as apple pie. Clinton faced it, Kennedy faced it, Roosevelt faced it, even Jefferson faced it. And at the bottom of this toxicity was a mixture of interest politics and reaction.

On the other hand, a lot of the hostility towards Obama comes not only from outer space, but from an identifiable galaxy known as racial stereotypes. Consider those portrayals of Obama as stupid, as a monkey, and so on. Some critics will point to, say, the U. K. Guardian's Steve Bell's portrayal of George Bush as an ape; not nice, but there is one important difference between that and, say, the N. Y. Post's Sean Delonas' portrayal of Barack Obama as an ape. Bell's portrayal was ferocious and based on his perception of Bush's own record and conduct; Delonas's portrayal was ferocious and based on his view of Obama's ethnicity.

  • The Right's defensiveness about Bush -- from his lackluster college years, his mysterious missing year in service, his chronic incuriosity and occasional sadism, his dependence on aides (from Cheney to Rice) to handle policy complexities beyond his own abilities, not to mention his tendency to drift into a fugue state during a crisis -- all suggested an awareness that Bush wasn't up to the job. Bell may have been out of line, but he was not off target.
  • Delonas, on the other hand, missed it by a mile. Obama has always been the smartest guy in the room, and he knows it, and he has no idea why he is even listening to your stupid ideas. Google image Obama vulcan and you'll many examples of this theme. Vulcans are smart, superior, priggish, intolerant, compulsively honest, high-minded, and they look down on us. Obama is all of these things, which is why cartoons like Delonas' are so bizarre. They seem motivated by something other than his conduct, character, or record. Skin color, perhaps.
Romney, of course, is a Ferengi.

*              *              *

Back ...