According to a google search I am the webs fourth leading source on "national guard tcrs" which is odd since I had to look up tcrs (at google) to find out what I meant.
Sunday, November 28, 2004
Social Security Book Cooking
Blogger ate a long post I wrote on this recently. Republicans are still talking about not describing the increase in the deficit due to the Bush administration's partial privatisation of social security as an increase in the deficit. Interestingly, they seem to be willing to claim that the partial privatisation will move the social security administration towards solvency.
This is crazy. Furthermore there is no way to use financial tricks to hide the complete craziness of the argument. When a reform is proposed, its proponents should, and I trust will, be forced to discuss three numbers which can't be fudged. The date when the Social Security Administration will begin to spend down it's trust fund (seems to be 2005 according to current proposals) the date after which, without further reform, SSA outlays will exceed revenues for the forseeable future and the date when the trust fund will hit zero. The plan to "save" social security will bring all of these dates closer. Therefore the claim that it addresses the shortfall is laughably false.
Of course the plan is not at all a plan to save social security as is suggested by the argument in its defence.
The above paragraph, written by RICHARD W. STEVENSON by in The New York Times seems to me to mainly demonstrate the extreme strain that a reporter faces when required by the canons of the profession to write a balanced article on a debate when one party to the debate (the Republican one) insists on saying things that make no sense whatsoever.
Here the defender of the proposal is anonymous, which is odd given that it's the President's proposal. The defence is laughably weak, since it is based on the claim that the proposal will be a "catalyst" for something completely different, which is not described at all.
The honest argument would be that, yes the private accounts would make the shortfall worse, but once they are in place it will be politically possible to cut (or eliminate) social security, which would be a good thing. Obviously that is not politicially possible right now, so the anonymous source chose to say nothing about what exact "broader changes" he or she has in mind.
I really think that 55 senators are not enough to win the debate in the country (or even maybe the Senate) when one is so totally utterly unwilling to address reality. Of course, I've always been wrong before.l
Blogger ate a long post I wrote on this recently. Republicans are still talking about not describing the increase in the deficit due to the Bush administration's partial privatisation of social security as an increase in the deficit. Interestingly, they seem to be willing to claim that the partial privatisation will move the social security administration towards solvency.
This is crazy. Furthermore there is no way to use financial tricks to hide the complete craziness of the argument. When a reform is proposed, its proponents should, and I trust will, be forced to discuss three numbers which can't be fudged. The date when the Social Security Administration will begin to spend down it's trust fund (seems to be 2005 according to current proposals) the date after which, without further reform, SSA outlays will exceed revenues for the forseeable future and the date when the trust fund will hit zero. The plan to "save" social security will bring all of these dates closer. Therefore the claim that it addresses the shortfall is laughably false.
Of course the plan is not at all a plan to save social security as is suggested by the argument in its defence.
A reasonable amount of borrowing now, the proponents say, would avert a much bigger financial obligation decades later. They say personal accounts would yield higher returns for individuals than the current system and could be a catalyst to broader changes that would bring the benefits promised by Social Security into line with what the system, which is also about to come under intense financial strain from the aging of the baby boom generation and the increase in life expectancies, can afford to pay.
The above paragraph, written by RICHARD W. STEVENSON by in The New York Times seems to me to mainly demonstrate the extreme strain that a reporter faces when required by the canons of the profession to write a balanced article on a debate when one party to the debate (the Republican one) insists on saying things that make no sense whatsoever.
Here the defender of the proposal is anonymous, which is odd given that it's the President's proposal. The defence is laughably weak, since it is based on the claim that the proposal will be a "catalyst" for something completely different, which is not described at all.
The honest argument would be that, yes the private accounts would make the shortfall worse, but once they are in place it will be politically possible to cut (or eliminate) social security, which would be a good thing. Obviously that is not politicially possible right now, so the anonymous source chose to say nothing about what exact "broader changes" he or she has in mind.
I really think that 55 senators are not enough to win the debate in the country (or even maybe the Senate) when one is so totally utterly unwilling to address reality. Of course, I've always been wrong before.l
Friday, November 26, 2004
Rejecting the Alternative
Elementary statistics makes the front page of the New York Times
Michael Kinsley take notice.
"Witth his condition deteriorating from Parkinson's disease last year, Steve Kaufman gave up making improvements to his home in Algonquin, Ill. "I couldn't even hold a nail stable," he recalled.
The drug, however, is no longer available to Mr. Kaufman or other Parkinson's patients in clinical trials. In June, its developer, Amgen, announced that the drug, which is called glial cell line-derived neurotrophic factor, or GDNF, had not proved better than a placebo.
"To control for the placebo effect, Amgen began a 34-patient trial last year in which all the patients had the pumps and catheters implanted but for the first six months half got GDNF and the other half saline solution. The results, announced in June, showed that while the group getting GDNF did somewhat better than the control group on a scale measuring the severity of symptoms, the difference in the results was not statistically significant."
Let us review the two claims in bold, both due to AMGEN the manufacturer of GDNF. In the first it is asserted that the treated group did not do better than the group given the placebo. In the second it is asserted that they did do better but that the difference was not statistically significant.
The statistically insignificant positive effect is considered proof that the true effect is zero. This is an exceedingly common but blatent error due to incomprehension of the meaning of the phrase "statisticially insignificant". Another approach would be to perform a larger trial in order to get more precise estimates and find out if the effect of the treatment is positive as weakly suggested by the initial trial.
Note the number in bold -- 34. Amgen appears to have decided that if a trial with a sample size of 34 doesn't show a significant effect, there is no reason to look further.
Now let's consider the cost of type I and type II error in this case. Type II error (failure to reject the nul that GDNF does not work even though it does) means people with Parkinson's suffer horrible disability and early death. Type one error means that a potentially large number of un-necessary surgery involving drilling holes in peoples skull will be performed until it becomes clear that the treatment doesn't work. Both are costly, but in this case, type II error is even worse, so rejecting the alternative is worse than a mistake, it is a crime.
Of course a larger study would be expensive "Amgen had spent hundreds of millions of dollars over 10 years on the drug's development." This is an absurdly tiny sum compared to the reasonable chance of a useful treatment for Parkinson's. Recall the voters of California just voted for 3 billion on a long shot.
And the point estimate was positive! It boggles the mind.
Elementary statistics makes the front page of the New York Times
Michael Kinsley take notice.
"Witth his condition deteriorating from Parkinson's disease last year, Steve Kaufman gave up making improvements to his home in Algonquin, Ill. "I couldn't even hold a nail stable," he recalled.
Earlier this year, after taking an experimental drug in a clinical trial, Mr. Kaufman built new kitchen cabinets and an outdoor deck. He was so steady he could walk across a narrow piece of lumber like an Olympic gymnast on the balance beam."
But
|
|
|
 |
"To control for the placebo effect, Amgen began a 34-patient trial last year in which all the patients had the pumps and catheters implanted but for the first six months half got GDNF and the other half saline solution. The results, announced in June, showed that while the group getting GDNF did somewhat better than the control group on a scale measuring the severity of symptoms, the difference in the results was not statistically significant."
Let us review the two claims in bold, both due to AMGEN the manufacturer of GDNF. In the first it is asserted that the treated group did not do better than the group given the placebo. In the second it is asserted that they did do better but that the difference was not statistically significant.
The statistically insignificant positive effect is considered proof that the true effect is zero. This is an exceedingly common but blatent error due to incomprehension of the meaning of the phrase "statisticially insignificant". Another approach would be to perform a larger trial in order to get more precise estimates and find out if the effect of the treatment is positive as weakly suggested by the initial trial.
Note the number in bold -- 34. Amgen appears to have decided that if a trial with a sample size of 34 doesn't show a significant effect, there is no reason to look further.
Now let's consider the cost of type I and type II error in this case. Type II error (failure to reject the nul that GDNF does not work even though it does) means people with Parkinson's suffer horrible disability and early death. Type one error means that a potentially large number of un-necessary surgery involving drilling holes in peoples skull will be performed until it becomes clear that the treatment doesn't work. Both are costly, but in this case, type II error is even worse, so rejecting the alternative is worse than a mistake, it is a crime.
Of course a larger study would be expensive "Amgen had spent hundreds of millions of dollars over 10 years on the drug's development." This is an absurdly tiny sum compared to the reasonable chance of a useful treatment for Parkinson's. Recall the voters of California just voted for 3 billion on a long shot.
And the point estimate was positive! It boggles the mind.
Wednesday, November 24, 2004
Private sins and public service
I have long been totally irritated with the gossipy interest in personal scandals unrelated to fitness to serve in public office (long means before Clinton). Now I wonder. I still feel a bit guilty about paying attention to Bush's going AWOL, since it was long ago, but now I think Bush and Cheney's record of fraudulent accounting matter. It seems they are now planning to cook the federal books by counting the social security surplus as offsetting part of the general revenue budget deficit while keeping the costs of their social security partial privatisation off the books.
"As they lay the groundwork for what will probably be a controversial fight over Social Security, Republican lawmakers and the Bush administration are examining a number of accounting strategies that would allow the expensive transition to a partially privatized Social Security system without -- at least on paper -- expanding the country's record annual budget deficits. The strategies include, for example, moving the costs of Social Security reform "off-budget" so they are not counted against the government's yearly shortfall. "
To paraphrase Dirkson "a trillion here a trilion there and soon you are talking real money."
Now I can't imagine a general revolt over accounting standards, so I guess the economy is screwed. I thought they couldn't shock me anymore but this made my jaw drop.
Via Atrios of course.
I have long been totally irritated with the gossipy interest in personal scandals unrelated to fitness to serve in public office (long means before Clinton). Now I wonder. I still feel a bit guilty about paying attention to Bush's going AWOL, since it was long ago, but now I think Bush and Cheney's record of fraudulent accounting matter. It seems they are now planning to cook the federal books by counting the social security surplus as offsetting part of the general revenue budget deficit while keeping the costs of their social security partial privatisation off the books.
"
To paraphrase Dirkson "a trillion here a trilion there and soon you are talking real money."
Now I can't imagine a general revolt over accounting standards, so I guess the economy is screwed. I thought they couldn't shock me anymore but this made my jaw drop.
Via Atrios of course.
Thursday, November 18, 2004
Faint Praise From Senator Leahy
"the president, with the majority he has in the Senate, could have sent up Attila the Hun and got him confirmed," Mr. Leahy said. "But Judge Gonzales is no Attila the Hun; he's far from that, and he's a more uniting figure."
and furthermore Judge Gonzales is a member of a much more important demographic group, is a US citizen and is alive. On balance, I agree that he is a better choice that Attila the Hun. The grim thing is that Sen Leany's joke is barely exaggerated given the Leninist level of discipline of most Republicans.
Oh and if they think they can take us Huns for granted, they better watch out (OK so tehcnically I'm a Magyar but Americans call the Magyar Koztarsasad Hungary).
"the president, with the majority he has in the Senate, could have sent up Attila the Hun and got him confirmed," Mr. Leahy said. "But Judge Gonzales is no Attila the Hun; he's far from that, and he's a more uniting figure."
and furthermore Judge Gonzales is a member of a much more important demographic group, is a US citizen and is alive. On balance, I agree that he is a better choice that Attila the Hun. The grim thing is that Sen Leany's joke is barely exaggerated given the Leninist level of discipline of most Republicans.
Oh and if they think they can take us Huns for granted, they better watch out (OK so tehcnically I'm a Magyar but Americans call the Magyar Koztarsasad Hungary).
I hope Alberto Gonzales doesn't suffer organ failure during his confirmation hearings
but I also very much hope that he is closely interrogated under oath about his views on what constitutes torture, how quaint the Geneva conventions are and why judge Robinson just declared the military tribunals unconstitutional. I was not pleased to see Ashcroft leave, because I suspected that Bush would try this.
The Senate has a chance to finally declare itself irrelevant by confirming the appointment of Gonzales whose contempt for the legislature is as profound as his concept for human rights and the very concept of law.
but I also very much hope that he is closely interrogated under oath about his views on what constitutes torture, how quaint the Geneva conventions are and why judge Robinson just declared the military tribunals unconstitutional. I was not pleased to see Ashcroft leave, because I suspected that Bush would try this.
The Senate has a chance to finally declare itself irrelevant by confirming the appointment of Gonzales whose contempt for the legislature is as profound as his concept for human rights and the very concept of law.
Wednesday, November 10, 2004
It's all in the timing - John Carson
Kristof writes "Paging China! Help us! Urge the U.S. government to respect freedom of the press!...three different U.S. federal judges, each appointed by President Ronald Reagan...U.S. District Judge Ernest Torres found Mr. Taricani in contempt for refusing to identify the person he got the videotape from, and the judge fined him $1,000 a day... Then there's Patrick Fitzgerald, the overzealous special prosecutor who is the Inspector Javert of our age. Mr. Fitzgerald hasn't made any progress in punishing the White House officials believed to have leaked the identity of the C.I.A. officer Valerie Plame to Robert Novak. But Mr. Fitzgerald seems determined to imprison two reporters who committed no crime, Judith Miller of The New York Times [Finally we get to the point] and Matthew Cooper of Time, because they won't blab about confidential sources.
Jesusof Maryof and Josephof, I understand that blood is thicker than logic, but couldn't you have just said that you consider the sanctity of your employer more important than the law?
Freedom of the press does not imply that the sacred promises made by journalists to criminals
must be allowed to help criminals violate the law. Our constitution, in parchment and ink, does not stand between judges and journalists' sources. "Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press;..." does not imply that any law inconvenient to journalists is unconstitutional. People who witnessed a crime can be compelled to testify against the criminals other than themselves. "nor shall be compelled in any criminal case to be a witness against himself" does not mean "nor shall be compelled in any criminal case to be a witness against his or her confidential source"
The thing that really bothers me about this idiotic op ed is the pretense that it is about Mr. Taricani in Providence not Judith Miller in New York. The three seperate examples just show that there is nothing unusual about Fitzgerald or the Plame case. Promises of confidentiality of journalists to their sources have no particular constitutional status. The claim that refusal to defer to such promises is equivalent to censorship reminds us that the political power of journalists has caused enough arrogance to overcome their ability to read the plain english of the 1st and 5th amendments..
Kristof writes "Paging China! Help us! Urge the U.S. government to respect freedom of the press!...three different U.S. federal judges, each appointed by President Ronald Reagan...U.S. District Judge Ernest Torres found Mr. Taricani in contempt for refusing to identify the person he got the videotape from, and the judge fined him $1,000 a day... Then there's Patrick Fitzgerald, the overzealous special prosecutor who is the Inspector Javert of our age. Mr. Fitzgerald hasn't made any progress in punishing the White House officials believed to have leaked the identity of the C.I.A. officer Valerie Plame to Robert Novak. But Mr. Fitzgerald seems determined to imprison two reporters who committed no crime, Judith Miller of The New York Times [Finally we get to the point] and Matthew Cooper of Time, because they won't blab about confidential sources.
Jesusof Maryof and Josephof, I understand that blood is thicker than logic, but couldn't you have just said that you consider the sanctity of your employer more important than the law?
Freedom of the press does not imply that the sacred promises made by journalists to criminals
must be allowed to help criminals violate the law. Our constitution, in parchment and ink, does not stand between judges and journalists' sources. "Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press;..." does not imply that any law inconvenient to journalists is unconstitutional. People who witnessed a crime can be compelled to testify against the criminals other than themselves. "nor shall be compelled in any criminal case to be a witness against himself" does not mean "nor shall be compelled in any criminal case to be a witness against his or her confidential source"
The thing that really bothers me about this idiotic op ed is the pretense that it is about Mr. Taricani in Providence not Judith Miller in New York. The three seperate examples just show that there is nothing unusual about Fitzgerald or the Plame case. Promises of confidentiality of journalists to their sources have no particular constitutional status. The claim that refusal to defer to such promises is equivalent to censorship reminds us that the political power of journalists has caused enough arrogance to overcome their ability to read the plain english of the 1st and 5th amendments..
Tuesday, November 09, 2004
Truth Justices and the American Way
Everywhere including the Washington Post
Everywhere including the Washington Post
Friday, November 05, 2004
Class War
The Democrats decades long effort to be Republican light has not succeded yet. I have long thought that an extreme proposal to increase the progressivity of the tax code would be both good policy and good politics. The Republicans would reply by accusing the Democrats of class war. So ? Seems a pretty feeble argument. It would make rich journalists hostile. Like they are so sympathetic now. It would make it hard to raise money, but not impossible with the small donations on the internet and the Kerry campaign showed the limits of money without slogans.
I am sure that this would be popular. Whenever it is polled it is popular (this is very rarely).
I am also sure it would be good policy. Roughly a good case would be that this is what Clinton did with his increase of income taxes on the top 2 million households and increase in the EITC.
I think there is clearly an active margin of people deciding whether to participate in the labor market including married women and young people. Problems with high taxes on rich people seem to me to be difficult to detect in practice by people who don't cheat (no names but there are two in my mind).
I'd say eliminate income tax on all earnings under $ 50,000 for a family of 4. This isn't even much money.
Now a better policy would be payroll tax reform, but eliminating the income tax for many households would be better politics.
This seems obvious to me. I have never heard an even semi convincing argument against it.
The Democrats decades long effort to be Republican light has not succeded yet. I have long thought that an extreme proposal to increase the progressivity of the tax code would be both good policy and good politics. The Republicans would reply by accusing the Democrats of class war. So ? Seems a pretty feeble argument. It would make rich journalists hostile. Like they are so sympathetic now. It would make it hard to raise money, but not impossible with the small donations on the internet and the Kerry campaign showed the limits of money without slogans.
I am sure that this would be popular. Whenever it is polled it is popular (this is very rarely).
I am also sure it would be good policy. Roughly a good case would be that this is what Clinton did with his increase of income taxes on the top 2 million households and increase in the EITC.
I think there is clearly an active margin of people deciding whether to participate in the labor market including married women and young people. Problems with high taxes on rich people seem to me to be difficult to detect in practice by people who don't cheat (no names but there are two in my mind).
I'd say eliminate income tax on all earnings under $ 50,000 for a family of 4. This isn't even much money.
Now a better policy would be payroll tax reform, but eliminating the income tax for many households would be better politics.
This seems obvious to me. I have never heard an even semi convincing argument against it.
Wednesday, November 03, 2004
There are an estimated 250,000 provisional ballots in Ohio. These are ballots cast by people whose right to vote was challenged. Bush currently has a lead of 127,992 with 97 % of precincts reporting. The remaining 3% are mainly in pro-Bush counties. There are also tens of thousands of uncounted absentee ballots (some of which haven't arrived). They will also probably break for Bush. My guess is that most of the provisional ballots are marked for Kerry by African Americans and pushed into provisional by Republicans. If this is true, this election too will be decided by the courts.
I got this from Kevin Drum
I got this from Kevin Drum
Random Vote totals at C-Span 2
For New York with 0 % of precincts reporting C-Span 2 has 17 votes for Bush and 17 Votes for Kerry. Yellow Pig. I think it is a joke and that the guy who pulled it knows I can't get a screen shot.
For New York with 0 % of precincts reporting C-Span 2 has 17 votes for Bush and 17 Votes for Kerry. Yellow Pig. I think it is a joke and that the guy who pulled it knows I can't get a screen shot.
Congratulations Senator Castor
Bit allarmed about the Senate exit polls which seem to be from Luntz. However, I am calling Florida for Castor based on official vote counts
Huh they show Marinez ahead. Yeah but look at this
Castor minus Marinez - (kerry - Bush) = 4.7%. Seems that Republican precincts have reported first. I am quite sure that Bush will not win Florida by 4.7 % (I stand by prediction that Kerry will win Florida). Thus I call it for Senator Castor. Am I the first? May I call you Betty ?
Update: Looks like I wiffed this one too. Martinez appears to have won. Sorry Ms Castor
Bit allarmed about the Senate exit polls which seem to be from Luntz. However, I am calling Florida for Castor based on official vote counts
|
| %
Reporting | Mel Martinez
(REP) | Betty Castor
(DEM) | Dennis Bradley
(VET) | Peter Blass
(WRI) | Rachele Fruit
(WRI) | Mark Stufft
(WRI) | Nancy Travis
(WRI) |
| Total |
| 822,960 | 756,235 | 40,646 | 0 | 0 | 0 | 0 |
| Percent | 17.5% | 50.8% | 46.7% | 2.5% | 0.0% | 0.0% | 0.0% | 0.0% |
Huh they show Marinez ahead. Yeah but look at this
|
| %
Reporting | Bush Cheney
(REP) | Kerry Edwards
(DEM) | Peroutka Baldwin
(CPF) | Badnarik Campagna
(LIB) | Cobb LaMarche
(GRE) | Harris Trowe
(SWP) | Brown Herbert
(SPF) | Nader Camejo
(REF) |
| Total |
| 1,158,153 | 971,176 | 1,525 | 3,906 | 1,021 | 584 | 712 | 9,996 |
| Percent | 23.0% | 53.9% | 45.2% | 0.1% | 0.2% | 0.0% | 0.0% | 0.0% | 0.5% |
Castor minus Marinez - (kerry - Bush) = 4.7%. Seems that Republican precincts have reported first. I am quite sure that Bush will not win Florida by 4.7 % (I stand by prediction that Kerry will win Florida). Thus I call it for Senator Castor. Am I the first? May I call you Betty ?
Update: Looks like I wiffed this one too. Martinez appears to have won. Sorry Ms Castor
CNN says the election will be decided by Gays Lesbians and Bisexuals
They don't dare call it but their exit polls say President Kerry and that, if only straights could vote, Bushwould have been re-elected. Historic.
| ARE YOU GAY, LESBIAN OR BISEXUAL? |
|  | KERRY |  | NADER | |||
TOTAL | 2004 | 2000 | 2004 | 2004 | ||||
 | ||||||||
Yes (4%) | 23% | -2 | 76% | 1% | ||||
| ||||||||
No (96%) | 50% | +0 | 49% | 1% | ||||
They don't dare call it but their exit polls say President Kerry and that, if only straights could vote, Bushwould have been re-elected. Historic.
Exit polls from Wonkette who is hanging in there by switching to text only
Hot, Fresh Polling #
5:40 p.m.
CO Bush 50 Kerry 48
FL Kerry 51 Bush 49
IA Kerry 50 Bush 49
MI Kerry 51 Bush 47
MN Kerry 54 Bush 44
NV tied
NH Kerry 53 Bush 45
NJ Kerry 54 Bush 44
NM Kerry 50 Bush 48
OH Kerry 51 Bush 49
OR still too early to get accurate reading
PA Kerry 53 Bush 46
WI Kerry 51 Bush 48
Disclaims a birdie: "*** There appear be problems with exits in the following states that could be tipping numbers toward kerry: MN, NH, VT, PA, VA, CT, DE. described only as 'serious' issues we're looking at. so i would not put too much faith in those results."
Tuesday, November 02, 2004
Exit Polls Via kos
Tue Nov 2nd, 2004 at 18:03:32 GMT
Jerome over at MyDD gets the big scoop on the 2 p.m. VNS numbers:
Tue Nov 2nd, 2004 at 18:03:32 GMT
Jerome over at MyDD gets the big scoop on the 2 p.m. VNS numbers:
AZ CO LA PA OH FL MI NM MN WI IA NH
Kerry 45 48 42 60 52 51 51 50 58 52 49 57
Bush 55 51 57 40 48 48 47 48 40 43 49 41
These depend only on early voting, and so are presumably biased.
In 2000 the early exit polls showed Bush winning big. Now we see
hunger for Bush's scalp.
Monday, November 01, 2004
Faulty Polls or Faulty Analysis
One thing many poll watchers seem able to agree on this year is that there is something wrong with the polls. It appears that polls by different agencies give very different results, much too different to be explained by the stated sampling errors. In particular it appears the different agencies likely voter filters give systematically different results. I have been an enthusiastic proponent of this view since 2000.
Now I think that we may have seen patterns in random data and taken out our frustration over a deadlocked race on the pollsters.
I strongly suspect that some people are treating the margin of error of the polls as a margin of error of the difference Bush - Kerry when it is not. In fact the standard error of the difference Bush minus Kerry is almost twice the standard error of support for Bush or Kerry (see explanation below). I further suspect that some people assume that the variance due to sampling in the difference between results in two polls is about the same as the variance of each of the polls when in fact it is the sum of the variance of the two polls. These two mistakes would lead those people to underestimate by a factor of about 2.8 the variance in differences in Bush - Kerry in two valid (or equally biased) polls taken at the same time. This would explain why such people think there is something strange going on. It is less clear to me why people who understand simple statistics had the same impression.
After Luigi Giamboni and I attempted unsuccesfully to improve on pollsters raw numbers reweighting with internals, I finally checked the variance accross polls compared to the stated sampling variance, or taking the square root, I compared the standard devation of polls to the mean square average sampling standard error. The mean squared average is close to the ordinary average standard error which I will present here.
The variance across polls should be the sum of the sampling error, squared differences in bias across pollsters and the squared changes in actual opinion. Nonetheless it is only slightly greater than the stated average sampling variance.
I will present a lot of standard devations across polls and average sampling error for Bush-Kerry, that is percent who say they would vote for Bush minus percent who say they would vote for Kerry.
First consider all 98 polls of likely voters who were prompted with Nader as well as Bush and Kerrytaken in 2004 and reported at www.pollingreport.com including an estimate of sampling variance. The standard deviation across polls of
Bush-Kerry is 3.97 % . The average standard error due to sampling alone in the 98 polls whihc report this is 3.11 %. The difference of 0.86 % is partly due to all of the variance in true opinions over the year. Now a kindof table
Likely Voter 3 way 98 polls standard deviation 3.97 average sampling standard error 3.11%.
Likely Voter 2 way 97 polls standard deviation 3.64 average sampling standard error 2.81 %.
Registered voter 3 way
105 polls standard deviation 3.90 average sampling standard error 2.90 %
Registered voter 2 way
77 polls standard deviation 5.23 average sampling standard error 3.17 %
The variation across polls is only slightly higher than one would expect from sampling error alone even though some were taken in August and others in September.
Now look at polls whose sample period ended in October
Likely Voter 3 way 41 polls standard deviation 3.22 average sampling standard error 3.13 %.
Likely Voter 2 way 17 polls standard deviation 2.04 average sampling standard error 3.10 %.
Registered voter 3 way
29 polls standard deviation 2.85 average sampling standard error 2.95 %
Registered voter 2 way
13 polls standard deviation 2.56 average sampling standard error 3.19 %
There is somewhat less than zero evidence of anomalous variation across polls. That is the variation across polls is slightly lower than one would guess given the stated sampling error.
Some explanation of the calculation
Bush-Kerry is 100% times the average over respondendts of a variable which is 1 if they support Bush -1 if they support Kerry and 0 if they are undecided or support someone else. The variance of the variable is the probability someone supports Bush or Kerry. A sample estimate of this probability is the fraction of people in the poll who say they support Bush or Kerry (Bush+Kerry)/100%. The variable is averaged across respondents and multiplied by 100% to give Bush - Kerry which therefore has variance 10,000*((Bush+Kerry)/100)/N where N is the number of people polled. The sampling standard error in Bush -Kerry is therefore
10 times the square root of (Bush+Kerry) divided by the square root of N %.
I am very very embarassed to admit that Brad DeLong had to explain this to me in 2000, but I trust no one will read this far.
One thing many poll watchers seem able to agree on this year is that there is something wrong with the polls. It appears that polls by different agencies give very different results, much too different to be explained by the stated sampling errors. In particular it appears the different agencies likely voter filters give systematically different results. I have been an enthusiastic proponent of this view since 2000.
Now I think that we may have seen patterns in random data and taken out our frustration over a deadlocked race on the pollsters.
I strongly suspect that some people are treating the margin of error of the polls as a margin of error of the difference Bush - Kerry when it is not. In fact the standard error of the difference Bush minus Kerry is almost twice the standard error of support for Bush or Kerry (see explanation below). I further suspect that some people assume that the variance due to sampling in the difference between results in two polls is about the same as the variance of each of the polls when in fact it is the sum of the variance of the two polls. These two mistakes would lead those people to underestimate by a factor of about 2.8 the variance in differences in Bush - Kerry in two valid (or equally biased) polls taken at the same time. This would explain why such people think there is something strange going on. It is less clear to me why people who understand simple statistics had the same impression.
After Luigi Giamboni and I attempted unsuccesfully to improve on pollsters raw numbers reweighting with internals, I finally checked the variance accross polls compared to the stated sampling variance, or taking the square root, I compared the standard devation of polls to the mean square average sampling standard error. The mean squared average is close to the ordinary average standard error which I will present here.
The variance across polls should be the sum of the sampling error, squared differences in bias across pollsters and the squared changes in actual opinion. Nonetheless it is only slightly greater than the stated average sampling variance.
I will present a lot of standard devations across polls and average sampling error for Bush-Kerry, that is percent who say they would vote for Bush minus percent who say they would vote for Kerry.
First consider all 98 polls of likely voters who were prompted with Nader as well as Bush and Kerrytaken in 2004 and reported at www.pollingreport.com including an estimate of sampling variance. The standard deviation across polls of
Bush-Kerry is 3.97 % . The average standard error due to sampling alone in the 98 polls whihc report this is 3.11 %. The difference of 0.86 % is partly due to all of the variance in true opinions over the year. Now a kindof table
Likely Voter 3 way 98 polls standard deviation 3.97 average sampling standard error 3.11%.
Likely Voter 2 way 97 polls standard deviation 3.64 average sampling standard error 2.81 %.
Registered voter 3 way
105 polls standard deviation 3.90 average sampling standard error 2.90 %
Registered voter 2 way
77 polls standard deviation 5.23 average sampling standard error 3.17 %
The variation across polls is only slightly higher than one would expect from sampling error alone even though some were taken in August and others in September.
Now look at polls whose sample period ended in October
Likely Voter 3 way 41 polls standard deviation 3.22 average sampling standard error 3.13 %.
Likely Voter 2 way 17 polls standard deviation 2.04 average sampling standard error 3.10 %.
Registered voter 3 way
29 polls standard deviation 2.85 average sampling standard error 2.95 %
Registered voter 2 way
13 polls standard deviation 2.56 average sampling standard error 3.19 %
There is somewhat less than zero evidence of anomalous variation across polls. That is the variation across polls is slightly lower than one would guess given the stated sampling error.
Some explanation of the calculation
Bush-Kerry is 100% times the average over respondendts of a variable which is 1 if they support Bush -1 if they support Kerry and 0 if they are undecided or support someone else. The variance of the variable is the probability someone supports Bush or Kerry. A sample estimate of this probability is the fraction of people in the poll who say they support Bush or Kerry (Bush+Kerry)/100%. The variable is averaged across respondents and multiplied by 100% to give Bush - Kerry which therefore has variance 10,000*((Bush+Kerry)/100)/N where N is the number of people polled. The sampling standard error in Bush -Kerry is therefore
10 times the square root of (Bush+Kerry) divided by the square root of N %.
I am very very embarassed to admit that Brad DeLong had to explain this to me in 2000, but I trust no one will read this far.