Wednesday, June 29, 2016

The cost of presidential elections

I've been reading Big Money by Kenneth P. Vogel. The book talks about the rise of Super PACs and big money in politics.

I'll start by saying big money has always been around in politics, and it doesn't always help. I was in Minnesota in 1982 when Mark Dayton ran for US Senate, spending $7 million of his own money (a record for any senate campaign at the time) and losing.

What struck me is thinking about the cost per vote of an election. A total of $7 billion was spent by he candidates, parties, and other grups in the 2012 presidential election. So what is the cost per vote?

About 127 million votes were cast. This means a over $50.00 was spent per vote cast, quite a sum. However, since most votes are pre-determined before the election (a large number of voters automatically vote for their party of choice and aren't going to cross party lines) so the real cost per vote must consider the number of votes which might be swayed. This also includes the number of people are convinced to go vote who wouldn't otherwise vote, along with the number who are convinced not to vote who might otherwise vote (this being money spent by the other side to "inhibit" a vote).

My guess is that, barring a candidate with extreme charisma (e.g. Ronald Reagan getting working class democratic votes) at most 10% of voters are subject to persuasion. So this implies that roughly 10-12 million voters are the target of all the political advertising, recorded phone calls, etc.

Given this smaller number, for a major party candidate, the actual cost per vote of a presidential election is more on the order of $500.

Thursday, June 2, 2016

The Theranos promise of finger prick blood tests -- fact or fiction?

The New York Times reports: Elizabeth Holmes, Founder of Theranos, Falls From Highest Perch Off Forbes List.

I've been seeing articles on Ms Holmes for the last year or so. She claims to have technology which allows blood tests using a finger prick of blood rather than multiple vials. She dropped out of college to start a company to develop her new technology. The company is one of the "unicorns" often talked about -- private startup companies valued at over a billion dollars.

For the last 6 months the company has come under criticism and has had a lot of bad press. There are claims that the company's tests are inaccurate and that they are using conventional blood testing machinery for many of the tests. On the business side the company board of directors apparently doesn't include medical experts.

So is Theranos a scam? A new technology which didn't pan out? The company hasn't provided peer reviewed data on its technology, and the reports so far indicate that at the very least they are having trouble getting the product to market, not unexpected for a new technology.

Yet there is another aspect to this story. There are several very large players (existing labs, existing producers of blood testing equipment) who stand to lose a lot if Theranos' technology pans out. All of medicine is also highly regulated. Regulatory agencies are not known for their flexibility or acceptance of anything new. So the question arises -- how big is the real problem at Theranos and how much of the problem is the inability of regulatory agencies to deal with change?

Time will tell. My feeling right now is that Elizabeth Holmes really did have a new technology but that it's hit some snags (or doesn't work well enough). Investors have put $700 million into the company and they'll probably have a hard time raising more money, especially with the bad press. Hopefully we won't lose a promising new technology to bureaucratic incompetence or lack of funding to deal with unexpected problems.

Wednesday, June 1, 2016

Cellphone Radiation Linked to Cancer in Major Rat Study - maybe

Cellphone Radiation Linked to Cancer in Major Rat Study - IEEE Spectrum

A new, major study is showing a slight link between cell phone radiation and cancer. Yet the results of the study appear to show even more so the problem with studies. As with many cancer studies, rats were used, and exposed to much higher doses of radiation than provided by a cell phone, wireless, or other RF sources. As one commentor of this article said, a 50 lb rock hitting your head is very bad for your health, 50 lbs worth of marshmallows hitting your head over a period of time aren't a problem.

Complicating the study is the fact that the rats exposed to radiation, while showing a slight increase in cancers, apparently lived longer overall than the control rats. Should we worry about cell phone radiation? I'm not convinced.

However, maybe the tinfoil hat brigade has been right all along.