The Rest of the Story

Those of us who have been following the news about global warming are worried about the upcoming hurricane season.  But now another storm is brewing in Florida, and it has nothing to do with global warming.  

The storm is being generated by Black Box Voting, which describes itself as "a nonpartisan, nonprofit, 501c(3) organization. We are the official consumer protection group for elections, funded by citizen donations."  They have released a report of their inspection of voting machine logs from Palm Beach, Volusia, and Broward County, used in the November 2, 2004 elections.  Their report shows that the logs contain over 100,000 error messages.  In addition, the logs show many votes were recorded in October 2004, even though the specific machines were not used for early elections.  

After investing over $7,000 and waiting nine months for the records, Black Box Voting discovered that the voting machine logs contained approximately 100,000 errors. According to voting machine assignment logs, Palm Beach County used 4,313 machines in the Nov. 2004 election. During election day, 1,475 voting system calibrations were performed while the polls were open, providing documentation to substantiate reports from citizens indicating the wrong candidate was selected when they tried to vote.

Another disturbing find was several dozen voting machines with votes for the Nov. 2, 2004 election cast on dates like Oct. 16, 15, 19, 13, 25, 28 2004 and one tape dated in 2010. These machines did not contain any votes date-stamped on Nov. 2, 2004. [...]

The Palm Beach County Supervisor of Elections, Arthur Anderson, said that his staff had looked into the problem and that the votes were normal, it's just that the dates somehow changed.  [...]

"The votes were normal, it's just that the dates somehow changed."  If that was intended to be reassuring, I would say Anderson failed rather miserably.  If, as he says, his staff looked into it, then there should be a written report.  Anderson should say to the journalist, "We prepared a report, and I would be happy to send you a copy," or something like that.

Google News lists about 100 news articles that reference the Black Box Voting article.  Most of these are reprints of an Associated Press article dated 2/23/2006, by Brian Skoloff.  The AP article was picked up mostly by small organizations, although some of the big media outlets, such as The Washington Post, also printed it.  A couple of conservative sites printed it as well, including Town Hall and The Conservative Voice.  Another news article on the subject appeared in the South Florida Sun-Sentinel, dated 2/24/2006, by Kevin Connolly.  

Both news articles are disappointing, although they live up to the usual journalistic standards.  

The AP writer, in a superficial effort to be fair and balanced, contacted the spokesperson for the manufacturer of the voting machines, and asked for a reply:

Sequoia spokeswoman Michelle Shafer disputed the findings, saying the company's machines worked properly. Sequoia's machines are used in five Florida counties and in 21 states.

"There was a fine election in November 2004," Shafer said.

She said many of the errors in the computer logs could have resulted from voters improperly inserting their user cards into the machines. The remaining errors would not affect the vote results because each unit has a backup system, she said.

The problem with that, is that there is no analysis of the adequacy of Sequoia's response.  The Black Box Voting article indeed does cite voting card errors, but that is only one type of error that they document.  There are many more kinds of error reported, and most of them have nothing to do with the voting cards.  Furthermore, a reasonably inquisitive reporter could wonder how it is that the machines handle these card errors, and what the procedures are for responding to the errors.  One might wonder why the logs do not show what was done to correct the error.  Furthermore, the respons that "The remaining errors would not affect the vote results because each unit has a backup system" is inadequate.  How does that address the problem of incorrect date stamps?  

Many such errors that Black Box Voting reported simply are not addressed in the AP article.  For eample:

Polls closed and results report messages would be expected to appear on every voting machine at the end of the voting cycle, but these revealed problems with poll worker training and procedures at the administrative/training level. Some logs reported one report printed, some two, three, four or five, and several not only had no results tape printed but showed no closing of the polls. (Closing the polls tells the voting machine not to accept any more votes).

Card encryption bad and Card read fail errors also appeared, with the encryption error message the more frequent of the two.

Some logs show that no results were printed, indicating that there is no paper record of the votes that the machine recorded.  Some show that the machine was not properly shut down at the end of the polling session.  That would not neceassirly be a problem, if all the votes had reliable date stamps; and vote recorded after the polls were closed could be discarded.  But as we have seen, the date stamps are not reliable.  I would like to think that a vigilant reporter would notice this, and ask about it.

Likewise, the South Florida Sun-Sentinel article is only a superficial report.  They quote a Volusia County official:

Volusia officials said the charges are groundless and questioned Harris' credibility.

"If you wish hard enough for a problem, your mind can imagine it," said former Volusia County Election Supervisor Deanie Lowe, who ran the 2004 election.

"I don't know of any election or of any voting system she has ever programmed, so she does not understand the situation. . . . She doesn't know what she's talking about."

I think it is fine to give the County officials a chance to respond, but the reporter could do better than that, by pressing for specifics.  Lowe's response is a sweeping rejection of all claims, based upon a vague dismissal of the credibility of the Balck Box Voting director, Bev Harris.  Vague ad hominem attacks tell us nothing.  I want to know what the county officials have to say about the individual error messages.  I want to know what happens when error messages occur while votes are being cast.  Are poll workers aware of the errors?  Is there a log kept by poll workers, showing when the errors occurred, what was done in response, and which poll workers responded?  Is there a policy and procedures manual that specifies how these situations are handled?  If so, how can citizens get access to that manual?  What does it say a poll worker should do, if, for example, a voting machine reports a card encryption error?  

Bloggers have been commenting on this, typically with a high degree of skepticism.  For example, an IT guy, Truthspew, writes:

Can a president elected through fraudulent means be tossed out?

Because Black Box Voting is really going gangbusters identifying serious election hanky panky.

For example, while I find the grammar and syntactical skills of the journalists at the Associated Press to be very disappointing, this article ineloquently states what the problems encountered happened to be.

Even during the 2000 Judicial Fiat we knew something was wrong. Same happened in 2004 and Dubya's plunging numbers only reinforce the notion that something is seriously rotten in Denmark.

Elisabeth, writing on Infomanic, picked up on the story, and got a useless comment:

Um, Elisabeth. We need to talk. The DUmmie FUnnies has LONG chronicled the foibles of the veracity challenged Bev Harris. Even the Democrat Underground has BANNED her from their leftwing site. Check out our December 2004 archivies of the DUmmie FUnnies and then do a search on the MANY references to Bev Harris and that should get you up to speed on the whole Black Box Voting thing.

The only thing Bev Harris and Black Box Voting are really good for is as a reliable source for comedy material and for that we thank her.---P.J.

Again, attacking Bev Harris may be interesting, but it is a different story.  We need to see an analysis of the specific claims made by Black Box Voting, and a credible explanation of each claim.  Harris provided the actual logs -- the raw data -- that were used by Black Box voting.  If someone wants to complain about her reporting, that person needs to look at the logs and explain what all those error messages mean, and show why they should be disregarded.    

Independent Report picks up on the story, and points out:

Does this sway the election results any? No, probably not. However, we find it curious that the voting machine manufacturers fought, kicking and screaming, against providing their source code or the machine's capabilities. "Trade secrets" they claimed, while also claiming the machines worked wonderfully well and could not be sabotaged. Well, we know they can be hacked pretty easily, and these irregularities point to some fishy business going on somewhere.

Bev Harris does not claim to know if election results would have been any different had these errors not occurred.  From the AP artiicle:

However, Harris said it was impossible to determine what information was altered or if votes were shifted among candidates.

That, in itself, is newsworthy.  The fact is, the results of the election cannot be verified, because the machines do a lousy job of telling us what actually happened on voting day.  As one comment states here:

I started reading the actual log files... as far as I can tell, it doesn't look like the numbers were manipulated to benefit a specific candidate. It does look like the company is totally inept, and isn't qualified to count jellybeans.

Some people read the Black Box voting article and concluded that there were problems, but the problems probably did not change the election results.  Others looked at the same article and concluded that the problems do indicate election fraud, such as this post at What a Mockery: Concrete Evidence that 2004 Vote was Rigged in Florida.  Some echo the report, and imply that the report indicates fraud, but do not state that explicitly; Ranting and Venting: Florida Voting Machine Logs Reveal Anomalies, by Mindwolf,  is an example of this approach.  

It seems like all this work the GOP is doing is just to make it easier to stage an election so they can seize power and make it look like it was legal.

Others merely report on the findings, without adding anything; presumably, they are content to let readers draw their own conclusions.  For example, Rob Galgano, at The Great Leap Forward, does exactly that.  In a way, that approach is preferable to the approach taken by news writers who provide a half-hearted attempt at being fair and balanced, but no analysis of their findings.

What I notice about this situation, is that the Black Box Voting article, and the news stories about the findings by the Black Box Voting (BBV) organization, both illustrate similar concepts.  What BBV found, is that one of the fundamental guarantors of our democracy -- accurate vote counting -- is no longer credible.  Similarly, another fundamental guarantor of our democracy -- good journalism -- suffers from a credibility problem.  Both of these problems are serious, but having the two together is especially bad.  Together, the lack of credible voting, and the lack of credible journalism, threaten to create a new storm system in our political ecosystem.

Aspazia left a comment with some questions, pertaining to my last post.  Since the response is way to long to fit into a comment box, I've responded here.  Note that the response might not make a lot of sense unless you first read the original post, and the comment, here.

Yes, the term hysteroid dysphoria is attributed to Donald Klein, from the late 60's. At this point (year 2006), it is more of historical and sociological interest, than medical interest. As for the question of whether MAOIs or Prozac would be better for treating rejection sensitivity, the answer is the predictable one: it depends on the patient, and to some extent, on the doctor.

I am confident that the magnitude of the therapeutic effect is greater with MAOIs, and that a higher percentage of patients will respond. But at least 30-35% will stop taking an MAOI because of adverse effects, compared to 10-15% with Prozac.  

From the psychiatrist's perspective, the judgment about which is "better" will depend upon a number of factors.  MDs who see a lot of young, relatively healthy patients, with clean diagnoses, and without a long history of multiple medication trials, will naturally see Prozac as the better choice for their typical patient.  MDs who see patients with multiple diagnoses (e.g. depression and panic disorder and PTSD), patients who have had multiple trials on modern antidepressants, will tend to think of MAOIs as better.  That's because there is no point in starting yet another trial of yet another SSRI in such a patient.  (If that was going to work, it would have worked already.)  

Some younger psychiatrists have little or no experience prescribing MAOIs, and would need to do their homework before doing so.  Some will do their homework, while others will not.

Psychiatrists vary considerably in their practice patterns, thus there are systematic differences in their patient populations.   It is important to keep this in mind when listening to a psychiatrist talk about his or her own experiences.  This is because it is impossible to know what conclusions can be drawn from those experiences, unless you know something about the population of patients among whom those experiences occurred.  

As for whether patients can pressure doctors into prescribing antidepressants, I would like to think that it would not happen very often.  It is pretty easy to say to someone, "Look, I know you think this would be best, but I really have to prescribe according to my own judgment..."

A bigger concern is that it is fairly easy for someone to come in and give all the correct answers, to manipulate the doctor into prescribing an antidepressant.  There is no good defense against that.  As a physician, one has to assume that the patient is acting in good faith, and the suspicion level is going to be low unless the patient seems to be fishing for a controlled substance.  

Make SSRI's available without a prescription?  First you would have to get a company that would be willing to sell them without a prescription, and I tend to doubt that anyone would do that in the USA; the liability issues would be horrendous.  But I realize that is a different issue.  Assuming that a company would do it, would it be advantageous for society?  The issues I see are these:

1. Selling antidepressants over the counter (OTC) might do something to destigmatize depression, and mental illness in general, and there could be benefit there.  

2.  It could trivialize mental illness, which would be bad.

3.  A lot of people who don't need them would end up taking them.  Most of those people would suffer no harm, except for the wasted money, and would stop taking them after a while.

4. Some people with bipolar disorder would became manic or hypomanic, and that would be a problem; in some cases, it would be a big problem.

5. Some people would try to abuse them.  Most of those people would get no appreciable effect, and would stop, no harm done.  Some would combine them with other substances, and that could be very dangerous.  

6. Some people report that antidepressants intensify the effect of alcohol.  That is not directly dangerous, assuming that the person does not intentionally overdose, but it can be hazardous because the person may not realize how much impairment there is.  (Of course, that can happen with alcohol alone, but it is more likely to happen with some kind of CNS drug involved.)

7. Some people who need treatment, but who for some reason will not see a doctor to get treatment, might end up getting beneficial treatment.  While not optimal, that would be more good than bad, except for #8...

8. Some people who are at risk for suicide would try to treat themselves in isolation.  Isolation is very bad for persons who are suicidal.  Self-treatment almost certainly would be a bad idea, and could have disastrous consequences.

Policy wonks may have some way of estimating the magnitude of those good and bad factors, and coming up with some kind of pronouncement about whether the good would outweigh the bad.  Personally, I would be highly skeptical of any such analysis.  So, to answer the question about whether SSRIs should be available OTC, I would not attempt a numerical analysis.  Rather, I would rely on the following kind of judgment:  

• In general, it is a bad idea to make a radical change in how a medication is used, unless the likely consequences can be anticipated and quantified.
• Selling antidepressants OTC would be a radical change in how they are used.
• The likely consequences cannot be anticipated, nor quantified.
• Therefore, it would be a bad idea.

I realize that that is a simplistic analysis, but even so, it pretty much sums up what I think about the idea.

Yesterday, ambling around the Internet, I read an article, and a blog post, that seemed to call for some kind of response.  The article is on The Nation's website: Brave Neuro World: The Ethics of the New Brain Science, by Kathryn Schulz; the blog post is Prozac Feminism?, by Aspazia.  What was bugging me, is that my intuition was telling me that there is some kind of important connection between the themes of the two pieces, but I could not quite put my finger on it.

Tonight, I am going to try to clarify for myself what the connection is.  This is not a post that was thought out before writing it.  Rather, I let my unconscious mind wrestle with it, and now I am going to sit down and start typing and see what comes out.  Ah, the joy of blogging!  

I will refer to the Schulz' article as BNW, and Aspazia's as PF, just as a kind of shorthand.

In BWN, Schulz poses a number of questions that are created by advances in neuroscience, and in our ability to control what the brain does.  Essentially, she points out that the prospect of enhancement technologies will pose ethical quandaries.  If enhancement is possible, to what extent will humans be free to make use of those enhancements?  Will it be possible, or proper, for anyone to coerce another to make use of such enhancements?  What are we to do about the inevitable unequal access to enhancement technologies?  

Schulz can no more answer these questions that anyone else.  Admittedly, I was sort of hoping that she would venture some answers, no matter how speculative or premature.  

Aspazia takes a different approach.  Not surprisingly, she ends up with different questions.  Rather than viewing Prozac as an enhancement technology, she views it as a means of promoting conformity to social expectations.  In that way, Prozac becomes a metaphor for the social pressures that promote conformity and compliance in women.Viewed in that way, Prozac is only an "enhancement" technology to the extent that conformity and subservience are improvements over the original.

I will leave aside the question of whether Prozac actually does that, or can do that.

What occurs to me first, is sort of an odd parallel.  I realize that understanding this requires some specialized knowledge, and that I am too impatient to expand on all of the prerequisites in sufficient detail, so I may loose some readers by not explaining things sufficiently.

In the 1950's, three particular new molecular entities were developed for the treatment of depression: phenelzine, isocarboxazid, and tranylcypramine.  These were marketed successfully as Nardil, Marplan, and Parnate, respectively.  All three are monoamine oxidase inhibitors (MAOIs).  Although there are highly effective, they were supplanted rather quickly by the tricyclic antidepressants (TCAs).  The reason is that all of them required that the patient follow a special diet, in order to avoid an interaction with a chemical found in certain foods.  (See the Wikipedia article for details.)  MAOIs also tended to cause weight gain, sexual dysfunction, and various other adverse effects.  

Although the MAOIs were supplanted by the TCAs, and the TCAs were supplanted by Prozac, they continue to excel in one respect.  MAOIs have particular efficacy for the treatment of atypical depression.  Atypical depression is characterized by a cluster of four symptoms: overeating, oversleeping, leaden fatigue, and rejection sensitivity.  Ordinary depression, in contrast, tends to involve loss of appetite, insomnia, intentional social isolation, and fatigue that does not have a peculiar somatic sensation associated with it.  MAOIs are particularly effective at reducing the symptom of rejection sensitivity, and thus are the preferred agents for treatment of Social Phobia.

As an historical aside, there is a disused term lurking in the dustbin of psychiatry: hysteroid dysphoria.  The term hysteroid was used in reference to an ancient (but not ancient enough) belief that certain emotional symptoms were caused by a wandering uterus.  (Really.  People actually believed that.)  It was thought that certain emotional symptoms tended to cluster together, along with rejection sensitivity.  That cluster was given the name hysteroid dysphoria.  In the 1980s, someone got around to analyzing the statistics, and it was found that the proposed symptom cluster had no validity.  The term has been abandoned since then.  In retrospect, it seems that it was merely a term of disparagement that was applied to women who had emotions that were inconvenient to men.  (As a part of the process of discarding pejorative terminology, the symptom of rejection sensitivity now is often referred to as mood reactivity.  The two are not exactly synonymous, but close enough for some purposes.)

Anyway, in the 1950's, the treatment of mood reactivity was hard: the patient had to memorize a list of foods to avoid, learn new recipes, get the family to accept a new menu, and scrupulously avoid dietary indiscretions.  Likewise, in the 1950's, housework was hard.  

Prozac changed the treatment of mood reactivity, sort of like the way the microwave oven changed housework.  Just push a button, and you're done.  Nothing hard about it.   That is the odd parallel that I referred to earlier in this insufferable stream of consciousness.

Both BNW and PF pose questions about bioethics.  Specifically, they ask how society should deal with the fact that we are developing new ways to alter brain function.  Both essays point out that there is a potential for coercion.  

BNW concludes with this:

If we fail to have that discussion, we risk winding up with a social policy for neuroscience based on tactical decisions, not ethical ones; benefiting the few, not the many; and obscuring the complex relationship between personal decisions about our minds and public decisions about our culture. That is a social policy we need like a hole in the head.

PF concludes with this:

What sort of culture do we become when we can gender engineer ourselves right into the sort of personality types that kick ass in business, that make us less sentimental about sex, and less overly sensitive to the needs of others?

Are we ready for this medically enhanced post-modern Feminism?

I think these issues are important.  What is important to note, though, is that society does not put people on medication.  Doctors put people on medication.  Society may exert unarticulated pressure on companies to develop products, on doctors to prescribe those products, and on people to go to their doctors to get those products, but ultimately it is the responsibility of the physician to be aware of those pressures, counteract them, and to make the prescribing decisions based on clinical grounds, with the patient's interest and value system in mind.  

I have mixed feelings about this.  On the one hand, many medical decisions are complex.  With that in mind, it seems that the need for a physician to sort out sociological and gender issues and consider them in the prescribing decision is no more complex than the decision about who should get bypass surgery.  On the other hand, the medical profession has a spotty record when it comes to that kind of thing.  It has been reported that women with heart disease have not gotten treated as aggressively, or as appropriately, as men.  There are other examples of bias in medical decision-making, such as racial or class biases.  I don't think the problem of potential biases is intractable, but I do think it requires some attention.

It is not something that worries me, really, because I do think that the medical profession is responsive when such biases are documented.  If we are paying attention, we should be able to avoid such biases, so hopefully they will not occur in the first place.  The authors are correct to point out the potential for such biases.