When elections - and clinical trials - go wrong

As I write this, media outlets have just declared that Phil Murphy has won New Jersey’s race for governor.

Murphy has less than a 1 percentage point lead, and a significant number of votes remain uncounted. Nonetheless, the Associated Press was eager to declare the election settled - no doubt in part because editors there knew that the longer it went without a declared winner, the more conspiracy theories would flourish.

Other media outlets were happy to follow. Whether they would have done so equally fast if Jack Ciattarelli, the Republican, had won, is an fine question - though irrelevant to this discussion.

The broader issue is that it should not be the media’s job to call a final winner in a very close race before election officials have counted all - or substantially all - of the votes. The AP is stepping in because the authorities have failed to do their jobs in a timely manner.

And these failures are happening more and more often.

For generations, elections have generally been called and decided within hours after the polls closed, except in those rare cases when they genuinely were too close to call, where a mere handful of votes separated candidates.

No more. Vote-counting now routinely drags on for days, sometimes weeks.

This is in sharp contrast to the trends for almost every other form of data. Tracking information has now become so easy and cheap that - for example - anyone can go on line and watch every commercial flight in the United States in real time for free, or receive a notification of potential credit card fraud and instantly have the option to close the card,

This failure of speedy vote-counting may seem of interest mainly to journalists and political scientists. After all, the final counts do eventually arrive.

But it is no small matter. It is profoundly dangerous.

It erodes confidence both in the vote itself and in the state’s basic ability to carry out its duties. If New Jersey cannot even determine who has won its highest office with clarity, precision, and speed, why would we trust it to do anything else? It is probably no coincidence that claims of stealing elections have surged since 2000, the first time since 1960 when the voting process itself could credibly have been said to affect a Presidential election.

Both the left and the right are now willing - if not outright eager - to make wild claims that elections will be or are being stolen. Yes, the left too.

So why did New Jersey fail?

Not because turnout was overwhelming, though it was higher than normal for an off-year election. Not because of weather problems or other unforeseeable disasters.

Our voting process has two core problems, one long-standing and one new.

The old is that elections are run at the state, county, and city level, and some of these local authorities are - how do I put this nicely? - more committed to transparency and speed than others. Not surprisingly, the old machine states in the Northeast and Midwest, where election boards have provided a quiet but steady source of patronage for generations, tend to be the worst.

Still, the machine creaked along reasonably well - until the federal move to allow and even encourage voting-by-mail.

Personally, I don’t think people should be allowed to opt out of voting in-person without a documented medical reason. This is NOT an argument against early voting, or weekend voting. I think voting should be as convenient as possible. But making people go to a centralized polling place where observers from both major parties and independents can watch the process is likely to discourage fraud.

I AM NOT SAYING WIDESPREAD FRAUD HAPPENED IN THE 2020 PRESIDENTIAL ELECTION - just that the risk seems higher (at least to me).

But even without fraud, mail-in voting has changed our elections for the worse, especially in the machine states. They are struggling to adapt to the mailed ballots - shocking, I know, that sleepy government bureaucracies are having difficulty with a major structural change. And some states have adopted rules that seem almost designed to make the counting as slow as possible.

So here we are. The process of elections is getting measurably worse - at a real cost to public trust - and no one seems to care.

What does this have to do with clinical trials - and specifically Pfizer’s clinical trial for Comirnaty, aka BNT162b2, aka the mRNA Covid vaccine I am not letting my kids get anywhere near?

Yesterday the BMJ - the only major medical journal to look at all skeptically at the Covid vaccines - published a piece about problems in the Comirnaty pivotal clinical trial last year.

The piece centered on a whistleblower who worked briefly (as in for two weeks) at a company running several trial sites for Pfizer in Texas. The problems she describes do not come close to rising to the level of fraud that would compromise the trial.

But they are ugly and a sign of a deeper problem - in this case, Pfizer’s desperate rush to complete the trial. Just like the vote-counting crisis, they damage public confidence - as they should.

They would, anyway, if anyone in the media would report on them. But unlike election officials, Pfizer doesn’t have to worry about that risk. It can be just about certain no one will follow up on the BMJ’s reporting.

And that failure is even more damaging to public health than the trial errors.