Robot Cars Will Change Everything

Last week famed entrepreneur Elon Musk announced that new models of the Tesla S electric vehicle would be capable of driving themselves. While the announcement might be largely marketing, it is the first portent of a new age. Autonomous cars are coming, whether we are ready or not, and it will be revolutionary.

1. Lives

Self-driving vehicles are robots and thus do not get tired, drunk, or impatient, and they certainly do not break the law. This means that the carnage on the national road system will disappear, saving 32000 lives per year. In stark actuarial terms, eliminating road crashes will save the economy a staggering $800 billion per year. More importantly, thousands of families will be spared the crushing sorrow of the needless death of a loved one.

2. Climate

Despite the claims of a few of our more dinosaurish politicians, it is increasingly clear that burning carbon is causing climate change. If temperatures rise too far too fast the resulting stresses on ecosystems are likely to cause large-scale upheavals. If this is not enough reason to act, we are rapidly running out of conventional oil (although there is unconventional oil aplenty) so it wouldn’t hurt to think ahead.

A lot of atmospheric carbon comes from our cars. About 30% of the national energy budget is expended on transportation and this pumps about 4 trillion pounds of carbon per year into the air. On a personal level, every mile you drive contributes 3/4 of a pound of carbon dioxide to the total. To make matters worse, the process of moving a bag of Cheetos from the grocery store to your house is colossally inefficient: about 3/4 of the energy content of gas is wasted as heat, primarily in exhaust and heating the engine.

Here’s the good news: besides being quiet and having phenomenal acceleration, electric vehicles are far more efficient than carbon burners. If we can generate the electricity cleanly we will have gone a long way towards solving our carbon addiction problem. But there is an issue: current battery technology does not permit trips longer than about 60 miles, which might be okay for commuting, but not much else. Possible solutions to this problem are to develop more efficient batteries or batteries that can recharge very quickly. Even if these are feasible options, both will take time and substantial research investment to develop.

Fortunately there is a third option that is technically and economically feasible right now: electrify the national highway system. The idea is that you drive your electric vehicle to the nearest on-ramp using your car’s batteries and park. The car’s computer then takes over and merges you onto the highway where it collects energy either through an overhead wire or by induction. The computer drives you safely and efficiently to your exit where you take over to complete the trip. Autonomous highway driving is much simpler than city driving and vehicles already exist that can do this. While you relax with a cappuccino, your batteries are charged and your account is debited for the electricity.

Supplying the power to the road system will not be easy, but is within reach. Taking into account the much better efficiency of electric motors, about 100 gigawatts of power are required for the nation’s long distance driving. This can be generated with current technology by lining the national highway system with a 60 foot wide strip of photovoltaic panels. The total cost, again with current technology, will be about one trillion dollars. This sounds like a lot, but in practice would amount to $50 billion per year for twenty years. This is comparable to the national highways budget and is less than one tenth of what is spent on the military (and don’t forget the savings due to eliminating accidents). At a more local level, electrifying the road system will cost about $7 million per mile. This should be compared to construction costs of $4 – $10 million per mile of four-lane highway.

Autonomous cars help here too. Once they enter the highway they can link together in virtual electric trains, making for much more efficient travel. And because they are aware of other vehicles the density of traffic can be much higher and only two-lane highways need be constructed.

3. Future Impact

Robotic cars will do more than save lives and the environment, they will change our culture. If a car can drive itself, why bother to own one? If you want to go to the store, just tap a button on your smartphone and one will appear to pick you up in a few minutes. It would be like a super taxi service — an uber Uber. So, no more driveways or garages.

It will be much easier to live in highrises so we will be able to live in denser cities, thereby creating larger and more readily accessible green areas for all to enjoy. Our cities will be quieter and greener.

Neighbourhood roads can be much narrower since people need not park on them and autonomous cars can share them with skill.

Leaving football games will amount to an orderly robotic dispersal guided by efficient traffic flow algorithms.

Smaller vehicles will deliver groceries and other goods to your door.

Traffic laws will be obsolete: why stop at every intersection when connected self-aware autonomous vehicles can whiz through without slowing? Speed limits will be irrelevant because cars will be able to safely travel at speeds only constrained by comfort and technical feasibility.

Auto insurance companies will suddenly find themselves without clients and will need to seek other business.

People will live longer because car exhaust is a major contributor to small particulate matter pollution. This is no minor thing; in Pittsburgh small particulate matter pollution ranges from 10 to 20 micrograms per cubic meter which lowers our life expectancy by seven to fourteen months.

A few enthusiasts will insist on driving their own cars for sport, but most people will regard them as extravagant eccentrics.

A healthier and less environmentally damaging future is in store if we don’t drop the ball. There is work to be done: laws will have to modified to account for the new robotic reality, privacy concerns will need to be addressed, and a new national infrastructure will have to be built to take advantage of it all. We built the interstate system and the internet, landed a man on the moon, and eliminated small pox. We can do it again.

Praying for Science

Can prayer help the ill?

Invoking God’s good will to help the sick has been a mainstay of Christianity for 2000 years. God’s role as physician-in-the-sky is manifest at Lourdes, Fatima, and countless other sites that attract hordes seeking supernatural help for natural afflictions. But with the increasing scientific basis of medicine God’s role has diminished to a fall back position — the plan B god.

Our question, can prayer help the ill?, lies at the intersection of science and religion, of the ancient and the modern. At first sight, it appears to be scientific since it is testable and therefore falsifiable. But I will argue that things are not so simple. Furthermore, the theology is even more suspect than the science. It turns out it is not easy living at an intersection.

The earliest attempt to answer the question was made by Sir Francis Galton in 1872. Galton, a cousin of Darwin, was a Victorian gentleman with an epic range of interests and talent. One of a small cadre of polymaths that Europe and America seemed to throw up at will in the 19th century, Galton had a hand in founding or expanding psychology, geography, meteorology, genetics, statistics, forensics, and anthropology. Fascinated with his cousin’s work on evolution, Galton threw himself into studies of human nature, and introduced the use of statistics and surveys in his research. Twin studies were first used by Galton in his research on inheritance, as were the phrases, regression to the mean and nature versus nurture. Galton also coined the word eugenics and was an early proponent of the idea of improving the lot of mankind through careful breeding. Of course the idea and the field fell out of favor after it was appropriated by the Nazi intelligencia. These days the Galton Chair of Eugenics at University College London is called the Galton Chair of Genetics.

Galton’s idea to test the efficacy of prayer was simple: royalty in the UK are regularly prayed for since they head the Anglican Church. He therefore undertook to compare the average lifetime of members of the royal family to other persons of wealth. His results were published in the Fortnightly Review as Statistical Enquiries into the Efficacy of Prayer.

Galton prepared a table of average lifespans based on data he collected himself which showed no evidence for royal longevity, and hence, no evidence for the efficacy of prayer in promoting good health: “The sovereigns are literally the shortest lived of all who have the advantage of affluence. The prayer has therefore no efficacy, unless the very questionable hypothesis be raised, that the conditions of royal life may naturally be yet more fatal, and that their influence is partly, though incompletely, neutralised by the effects of public prayers.”

Although not definitive, one might think that Galton’s study would have dampened any enthusiasm for further scientific tests. But it is estimated that five million dollars is spent annually on the subject, that about 1/2 of American medical schools offer courses on spirituality and health, and that about 1000 articles are published per year on the subject.

Let’s take a look.

Religious Factors and Hippocampal Atrophy in Late Life

Dr Amy Owens of Duke University and colleagues decided to use MRI scans to check the correlation of brain structure with religiosity. Surprisingly they found one!: “Significantly greater hippocampal atrophy was observed from baseline to final assessment among born-again Protestants, Catholics, and those with no religious affiliation, compared with Protestants not identifying as born-again.” Of course the correlation was immediately turned into causation by the media: “New research finds membership in a minority religion seems to hasten a loss of volume of the hippocampal region of the brain.”

I am not sure that this bizarre result deserves any more thought, other than to remind ourselves just how flakey MRI research can be (dead salmon knows!).

A Randomized, Controlled Trial of the Effects of Remote, Intercessory Prayer on Outcomes in Patients Admitted to the Coronary Care Unit

Even though the paper starts with the line

From time immemorial, prayer for the sick has been a common response to the illness of a loved one,

this is a more serious effort, published in JAMA’s Archives of Internal Medicine. Dr W.S. Harris and pals examined 990 patients who had been admitted to the Coronary Care Unit at the Mid America Heart Institute in Kansas City. The study was double blind and has a decent sample size (good!) and did indeed find a significant difference (p=0.04) between outcomes of the group being prayed for over the control group.

Remote, intercessory prayer was associated with lower CCU course scores. This result suggests that prayer may be an effective adjunct to standard medical care.

It’s looking good for prayer! But the authors note

There were no statistically significant differences between groups for any individual component of the MAHI-CCU score. Mean lengths of stay in the CCU and in the hospital (after initiation of prayer) were not different, and median hospital stay was 4.0 days for both groups. There was no significant difference between groups using Byrd’s hospital course score.

In other words, if you look at other ways to score the patients, no difference in outcomes is seen. This is, in fact, a serious issue in study design. The metric you use can swamp the observable. The same thing happens in particle physics, where scientists can choose to exclude parts of their data (called a cut) with a click of the mouse. The temptation to experiment with cuts is irresistible and inevitably leads to an anomaly. If you do something 100 times, on average you will find a one percent occurrence. To prove the point, a 2006 study with 1802 CCU patients at six hospitals found no effect.

Theology

I am no theologian, and I haven’t consulted any in writing this post, but there seems to be something seriously amiss on the religious side of this story. Surely it is sacrilegious presume that we can explore the mind of God like he is a weighted spring in a middle school science experiment. If prayer studies made sense one could determine which religion God prefers

or how much the piety of the prayer can influence God’s actions

or how receptive God is to prayer as a function of distance

Study of the therapeutic effects of proximal intercessory prayer (STEPP) on auditory and visual impairments in rural Mozambique

I had looked at these examples as amusing illustrations of the foolishness of prayer studies. But bless her heart, along comes Dr Candy Brown of Indiana University who blithely jumps right in. Yes indeed, Dr Brown and colleagues have found that proximity of intercessor and patient is correlated with health outcomes!

This wretched research came with a commentary in the Southern Medical Journal written by Dr. John Peteet of Harvard. Dr Peteet was open-minded, writing

But the conviction that we live in a closed system governed only by naturalistic processes is an expression of faith in a world view rather than a conclusion logically demanded by the scientific method.

I have difficulty parsing this thought. Is he saying that prayer lies outside of “naturalistic processes”? If so, then how can its efficacy be tested in experiments? What does he mean by a “closed system”? I suppose he is thinking of some sort of spiritual metauniverse with ‘natural’ and ‘supernatural’ portions. Can one then call the union of the two closed? And is he trying to say that science is not a logical demand of science? I suppose that this is a true, but utterly useless, statement about metalogic. Frankly, I am not remotely surprised that this level of muddleheadedness is associated with prayer studies.

In the end, prayer studies are not scientific either. Omniscient and omnipresent deities need not respond in predictable ways to intercessory invocations. I am sure that the religious would say “God does not choose to reveal himself” to any outcome whatsoever from a prayer study. So what is really being studied? It certainly isn’t science and it certainly isn’t religion.

Can the Humanities find meaning in a meaningless universe?

Two Solitudes — one solitary man

I have to admit that the Chronicle of Higher Education is not my go-to choice for late night reading. I do enjoy the blog posts on language, but the tendency to publish articles on impressing the promotion committee or writing good proposals leaves me cold. And the taint of postmodern new-ageism in some of the feature articles is depressing. In spite of this experience, I was not prepared for the level of dismay that Visions of the Impossible invoked.

In Visions Jeffrey Kripal, a professor of religious studies at Rice University, (i) argues that dream precognition is real, (ii) says precognition is not studied because science is caught in dreary materialism, (iii) asserts an imaginary demise of the humanities. His bold solution to all these problems is to recast the humanities as a “study of consciousness coded in culture”. Kripal, you see, imagines that consciousness is beyond science and, just maybe, the humanities can ride to the rescue.

Dream Precognition

The essay starts with a stirring account of Mark Twain dreaming the death of his brother, down to minute details, weeks before it all comes to pass!

Most scholars have no idea what to do with such poignant, powerful stories, other than to dismiss them with lazy words like `anecdote’ or `coincidence’.

After realizing that he really meant it, I read on, all the while experiencing a growing sense of involuntarily taking part in a slow motion train wreck.

Let’s take a moment to deconstruct his assertion. Apparently referring to things that are coincidental and anecdotal as coincidental and anecdotal makes one “lazy”. Presumably the intent of the insult is to point out that all things, no matter how ludicrous or at odds with the opus of scientific knowledge, should be investigated with equal vigor. Postmodernism like this may sit well with creationists or vulgarizers of Jacques Derrida, but scientists tend to be busy and cannot afford to waste their time on nonsense.

But Kripal is prepared for scientific arrogance of this type. He notes, for example, that there are thousands of examples of precognition over many centuries. Its got to be true! Never mind that there are thousands of stories of elves, ghouls, goblins, and poltergeists over millennia — those are just anecdotes apparently. Kripal even has an explanation for why precognition and its cousins have not been discerned in the laboratory:

…the answer to why robust events like those of Twain … do not appear in the lab is simple: There is no trauma, love, or loss there. … The professional debunker’s insistence, then, that the phenomena play by his rules and appear for all to see in a safe and sterile laboratory is little more than a mark of his own ignorance of the nature of the phenomena in question.

That’s the problem with science! Its so clinical, with all those white overcoats, and fume hoods, and women with their hair in buns, and latex gloves, and logic.

What’s worse,

… we will not invest those resources in the study of anomalous states of cognition and consciousness, and so we continue to work with the most banal models of mind — materialist and mechanistic ones.

I have banal news for Dr. Kripal: science, by definition, deals with the totality of phenomena in the universe. If dream precognition were a real phenomenon people would study it. Although most scientists spend their days slogging away at some small corner of the modern scientific paradigm, all of them dream of discovering something truly new. What better than precognition? The person who conclusively demonstrated it would be bigger than Einstein. And yet, almost no one works on paranormal research. The reason is painfully clear: there is no way to study something that does not exist.

The irony is that if precognition were found and studied, it would enter the realm of materialism, and hence would fail to satisfy Dr. Kripal’s needs for `something else’. He would doubtless be forced to invent some new fantasy into which to pour his hopes. The god of the gaps is a feeble fellow indeed.

The Mystery of Consciousness

The good doctor is obsessed with paranormal happenings because he sees in them a portal to something even more mysterious: human consciousness. For Kripal, consciousness is a thing of itself, something distinct from life, yet bound to it. With a magical certitude that
consciousness is more than “cognitive modules, replicating DNA, quantum-sensitive microtubles in the synapses of the brain, or whatever”, Kripal opines

Many want to claim that … consciousness is not its own thing, is reducible to warm, wet tissue and brainhood. But no one has come close to showing how that might work. Probably because it doesn’t.

But isn’t it just as likely that this is a failure of imagination on the part of the author than of science? As a test, I offer a scientific explanation of consciousness:

Consciousness is an emergent property of complex biological systems.

I find this perfectly acceptable, in the same way that saying that memory is an emergent property of complex biological systems (presumably something Kripal would not argue with since explicit models of memory have existed since the days of Hopfield and his neural net). But I think Kripal would not accept this — he seeks something more, something deeper.

Let’s back up for a minute. It is easy to imagine that a virus is a collection of biomolecules that interacts with its environment in a predictable manner. Presumably a bacterium is not much different. The neurological structure of multicelled hydra is completely mapped and decent virtual hydra have been created on computers. Perhaps Dr. Kripal can believe that organisms as large as ants can be “explained” by biochemistry. Where does one draw the line between a biochemical description of life and consciousness and something magical? Surely not at homo sapiens, since it is clear that apes and other large mammals have emotions, can communicate, and feel empathy. Even my cat must be self-aware to some degree since it knows when it can safely attack small things and when it must run from large ones. Perhaps the line is between ants and cats? Surely a more rational thought would be that there is no line — there is no magical definition of life.

Those obsessed with teleological impulses long for meaning. The thought of a life as a mere collection of large molecules, as amphiphilic bags of self-replicating matter, as specks in a cold and hostile universe, leaves them afraid and yearning for something more. Homo timere seek a special place in the universe, and being alive, ascribe that specialness to life. Why do they never ask about the meaning of a piece of coal? Why does one clump of carbony matter deserve meaning while another does not?

The universe is filled with wonder and magic aplenty. If Kripal and his friends cannot manage to find meaning in the glorious higgledy-piggledy of it all, I am afraid that is their problem.

“Krauthammer” is not German for “scientist”

Normally I don’t read Charles Krauthammer. It would be like taking time out of my day to listen to the rantings of the guy who fills the bus with the smell of urine and his theories on black helicopters and communists massing at the border. But his latest missive mentioned science (The notion of “settled science” is in itself anti-scientific) so, against my better judgment, I bit.

What set Krauthammer off was this comment in the State of the Union Address: “The debate is settled. Climate change is a fact.” Apparently this tweaked a deeply buried, muddled, and rather moldy sense of scientific outrage in our intrepid journalist, prompting him to thunder

“There is nothing more anti-scientific than the very idea that science is settled.”

The rest of the article was spent ranting about mammograms, the nonagenarian Freeman Dyson, and the climate `pause’. I’ll get back to these later. But first, what about his claim? Is it scientific to talk about settled science?

 I am currently teaching a course called “Science and Society” and one of the things I have stressed to my students is that science is not entirely rational, and in fact it cannot be (you cannot deduce your way to new knowledge). Science has cultural and historical influences; scientists must deal with biases and self-deception; and everything is contingent. In short science is a human endeavor. In principle this means science can never be settled. But no scientist ever uses the phrase “in principle” — science is far too practical to dwell in the philosophical world of might-have-been or could-become. Instead, science talks about in practice.

The phrase “in practice” is implemented in a specific way in science, known as setting a scale. Let me illustrate this with an example. Say that your wife hears scratching in the attic and sends you to look for a mouse. You return and announce that there is no mouse. But your wife is a philosopher (or a Krauthead) and asks how you know that you have seen no mouse. In principle you have no answer because you cannot guarantee that the things you did see were not very odd looking mice. Fortunately you are a scientist and have a response for her: the things you did see were not mice within a certain tolerance — none of them had ears, were furry, or scurried into corners. This tolerance is the scale that I am referring to (excuse me, “to which I refer”). When physicists use the notion of scale they often refer to a given length or energy, say one meter or 1000 joules.

What does this have to do with settled science? When scientists say they understand something they always mean within a certain range of scales. For example, light is one of mankind’s most perfectly understood phenomena: our theories for light work on scales ranging from 10^-28 meters to 10²¹ meters — an incredible 49 orders of magnitude (I am being conservative here, the lower end can be pushed to the Grand Unification scale of about 10^-32 meters, and the upper end can probably be pushed to the size of the universe, about 10²⁶ meters). Another example is the conservation of energy, which is verified over similar length scales. The science of light and of conservation of energy are completely settled over these 49 orders of magnitude. Only a madman would waste time looking for violations of this settled science; instead scientists look to probe theories beyond the scales at which they are known to function. This is why the LHC particle accelerator; is so big and expensive. (An alternative is to look for emergent properties at known scales, but this is taking us too far afield).

Climate science is interesting because it happens over a large range of length scales. Dealing with all of these scales is what provides the intellectual challenge to the field, and spurs further research. The largest scales are set by jet streams and hurricanes and thus are a good fraction of the size of the Earth, about 10⁷ meters. The smallest scales are set by chemical processes which occur at around 10^-10 meters. One such chemical process is the depletion of ozone caused by the dissociation of CFC molecules by UV radiation and the subsequent combination of chlorine with ozone. Another process of concern is the absorption of thermal radiation by greenhouse gases.

The claim that climate change is occurring due to greenhouse gases rests on the properties of these gases, their interactions with light, and the conservation of energy. Recall that the science of these things is settled in a range of scales from 10^-28 to 10²¹ meters, which is much larger than the scale of climate science. Our understanding of these phenomena is not going to change, or if they do, they will change at scales that are completely irrelevant to climate science. Thus the statement that CO2 causes global warning is entirely settled.

The questions that fill the time of professional climate scientists do not involve these fundamental laws of nature, rather they involve details that answer questions like “how rapidly will the earth heat?”, “how much will it heat?”, and “what effects will this have?” These are clearly important questions and the answers will likely change somewhat as the science improves, but the basic principles and effects are set in stone.

 

With that settled, I want to return to a valid point that Krauthammer made, namely too many people are claiming “climate change” as the cause for their favorite disasters. It is far too early to correlate climate change with weather (be careful to distinguish short term and volatile weather from long term climate patterns!). The cold snap this winter is a random (actually deterministically chaotic, but that is another story) event, just like the California drought, and the latest megastorms.

And finally, Freeman Dyson is a (very) old man and has lost it. It happens. One of last century’s greatest scientists was Linus Pauling, who regretably spent the last years of his life ranting about vitamin C. As for the “pause” — it is a fabrication of the deniers.

 

Cellphones and Cancer: manufacturing fear

Three billion people use cell phones on a regular basis. That makes eight billion pounds of brain matter being subjected to cell phone radiation every day. Is this cause for concern? For a lot of people the answer is yes. Specifically, the idea that cell phones cause brain cancer has been floating through cyberspace and the national media for several years now.

The Science

The genesis of the cellphone cancer scare is an intricate story involving bumbling scientists, complacent reporting, populist politicians, and an establishment-wary public. But let’s settle something first: cell phones cannot cause cancer. Here’s why.

Light Basics

First we need to clarify that the “radiation” we are talking about is electromagnetic radiation. Electromagnetic radiation can be thought of as a wave with two properties, wavelength and wave amplitude. Wavelength is a measure of the distance between peaks of a wave, while the amplitude measures the height of a wave. If you are a boater you can appreciate that short wavelengths make for choppy waters while large wave amplitudes can be pretty darn scary. “Radio waves”, “microwaves”, “UVA”, and “light” are all types of electromagnetic radiation that differ only in their wavelength. All of this together is called the spectrum of light.

You are probably more familiar with characterizing radiation by frequency, which is how many times it oscillates per second (the unit of this is called a Hertz, abbreviated Hz). Frequency is related to wavelength by the formula c = wavelength x frequency, where c is the speed of light, about 300 thousand kilometers per second. (In case you are wondering, this is the same c that appears in Einstein’s formula E = m c².) We can use the equation to freely translate between frequency and wavelength — they are just different ways of characterizing one of the properties of radiation.

It is also possible to think of electromagnetic radiation in terms of particles, called photons. This is the famous “wave-particle duality” quantum mechanical nature of light. Photons always move at the speed of light but can have different energy. The connection between the wave and the particle picture is provided by Planck’s Law, which states that the photon energy is inversely proportional to the radiation wavelength. In other words, the shorter the wavelength the higher the energy. You can refer to the figure above or the table to see the translation between wavelength, frequency, and energy.

name	frequency	wavelength	energy (eV)	energy (j)
x-ray	1 billion GHz	0.3 nm	4,000	6 10-16
UVB	1 million GHz	300 nm	4	6 10-19
UVA	750,000 GHz	400 nm	3	5 10-19
visible	430,000 – 750,000 GHz	400 – 720 nm	1.7 – 3	(2.7 – 5) 10-19
microwaves	10 Ghz	3 cm	4 10-5	6 10-24
cellphone radio	1 Ghz	30 cm	4 10-6	6 10-25
FM radio	100 Mhz	3 m	4 10-7	6 10-26

Check it!
You can measure the wavelength of microwaves by covering a plate with marshmallows and microwaving them. Use low power and turn off the turntable. You should see spots of melted marshmallow about 7 cm apart (this depends on the frequency your microwave works at)! I tried this and got a gooey mess, let me know if you can get it to work!

A word about the units in the tables. A meter (m) is about 3 feet, a centimeter (cm) is about the height of your little finger. A nanometer (nm) is 10^-9 meters or 10^-7 cm. One MHz is one megahertz, or one million Hertz, or one million cycles per second. 1 GHz is one billion cycles per second. The units of energy used here are electron volts (abbreviated as eV) and joules (abbreviated as j). One electron volt is about 1.6 10^-19 joules. So there are a lot of electron volts in one joule. It takes 4.2 joules of energy to heat one gram of water one degree Celsius (which is called a calorie). The energy of a marble dropped from a height of 3 feet is about 1/10 of a joule. Clearly we are talking about some pretty tiny energies here.

One other thing about light: it interacts with matter by bouncing off of charged particles, mostly electrons in the outer orbits of atoms and molecules. Light is absorbed by a brick because the brick is dense with electrons that love to interact with visible light. Glass is transparent because the electrons that help make it up do not play well with visible light.

Light and Cancer

Can electromagnetic radiation cause cancer? Yes! This is the reason your dental assistant leaves the room when she x-rays your loose tooth. Lead does a good job of absorbing x-rays, which is why that apron that covers your delicate bits is heavy. Next on the list is UVB, which is also dangerous because it causes sunburn and skin cancer. UVA is dangerous, but in an indirect way (it creates free radicals that damage DNA). What about visible light? You’ve probably seen truck drivers with tan left arms — tan because he drives with his arm out the window and hence exposes it to UVA and UVB radiation. The rest of him is not tan because glass absorbs UVA and UVB radiation very well. Of course it does not absorb visible light (otherwise we couldn’t see through it!). We conclude that visible light does not cause sunburn or skin cancer. This is a good thing because most of us spend 16 hours a day bathed in (mostly artificial) visible light.

Check it!
Go outside on a sunny day; cover part of your arm with a piece of glass, another part with cling film, and leave a third uncovered. Which parts get sunburnt? (By the way, I have no idea what the light absorption properties of cling film are. Let me know what happens!)

Cells Gone Wild!

Why are x-rays, UVA, and UVB carcinogenic and why is visible light not? To understand this we need to recall some things about cancer. Cancer happens when cells start reproducing in an uncontrolled fashion. Since cell reproduction is controlled by DNA we might conclude that x-rays and other high energy radiation damage DNA molecules in some way. In fact this is exactly what happens. Light radiation with wavelengths less than about 300 nm have sufficient energy that they can scramble DNA base pairs when they smash into their electrons. Visible light simply does not have enough energy to cause this kind of damage. DNA lives in a quantum world and if a photon does not have enough energy to cause a base pair to scramble, no amount of photons (or radiation intensity) can ever cause it to happen. This is why you can expose your skin to visible light for hour after hour, day after day.

Cellphones Can Not Cause Cancer

What about cellphones? They typically broadcast between 800 Mhz and 2.1 GHz, which corresponds to a wavelength of 30 cm (let’s use 1 GHz for now). The energy carried by cell phone radio waves is not sufficient to damage DNA! In fact it is about one million times less energetic than visible light! Similarly, radio waves, TV transmissions, and microwaves are all too feeble to cause DNA damage, and hence cancer. Cellphones do not cause cancer.

Cell Phone Radiation and Heating the Brain

What cell phone radiation does do is jostle the molecules in brain tissue. Jostled molecules are better known as heated molecules. Thus the physics of radiation tells us that the biology of radiation damage is the biology of heat damage.

Of course too much heat is not a good thing. I would not be happy if my cell phone were slow-cooking my brain. But what levels of heating are we talking about? Heat energy deposited in tissue is measured as “SAR”, which stands for “specific absorption rate”. This tells us how many joules of energy are deposited in tissue per second per kilogram of tissue. The units for this are watts per kilogram, written as W/kg. In the US cell phones are not permitted to have SAR ratings of over 1.6 W/kg as measured in any 1/1000 of a kg in the brain (this is about 1 cubic centimeter of tissue). Now the brain naturally produces its own heat just by running all that high-powered thinking ability. The heat output from standing there thinking is about 15 W/kg, about ten times what the cell phone contributes. If you go outside into the sunlight and you are not wearing a hat, your head will pick up about 20 W/kg of extra heat energy. The upshot is that the cell phone is making a fairly small contribution to the heat budget of your brain. If that causes cancer we have a lot more serious things to worry about than cell phones.

If you would like to see what your phone’s SAR rating is, visit CNET, who have rated cell phone power levels, or, more ominously, “measur[ed] the quantity of radiofrequency energy that is absorbed by the body”.

One of the more famous images from this saga shows (dangerously red!)
absorption of cell phone radiation in the brain. Of course alarmists never bother to explain the scale (or that the image is actually a simulation, not experimental data). It turns out blue represent about 0.5 W/kg, whereas the scary orange-red is 1 W/kg. Again, these figures are much less heat energy (per second per kilogram) than is received in natural and common situations.

The brain has evolved over hundreds of thousands of years in a thermal environment, namely it is warm. The survival of the species then depends on the stability of biological processes under the jostling due to thermal energies that we are likely to experience in a lifetime. Run hard from a saber tooth tiger and your brain will warm up. Argue with your wife over who gets to eat the giant sloth bone, and your brain will warm up. Spend the night sleeping outside of the cave, and your brain will cool down. The brain can handle it, and it doesn’t get cancer.

Glioma Rates

And just in case you are not entirely convinced, consider this study, published in the Journal of the National Cancer Institute. The authors examined the rates for a type of brain cancer (glioma) in the entire adult population of Denmark, Finland, Norway, and Sweden (these countries maintain thorough national cancer registries). The results are summarized in the figure below.

The top figure (A) is for men, the bottom (C) is for women. The points represent glioma incidence by year. Circles are for people aged 20-39, squares are 40-59, and triangles are 60-79. We learn some interesting things: the older you get the more likely it is that you get brain cancer; women get less brain cancer than men; and the cancer rates have been roughly constant since 1974. To be fair this is a logarithmic plot, so slight slopes can mean a lot. For example, young men saw their cancer rates rise from about 3/100,000 in 1974 to 4.5/100,000 in 1988. One can speculate about the cause for this, especially since all the other rates are essentially unchanged. One thing for sure, it is not due to cell phone use, which only took off in the mid-1990’s. In fact, if we take the data too seriously, it looks like cell phone use has lowered the rate of cancer (at least among young men).

This All Seems Familiar…

Do you remember the power line fiasco of the 1980s and 90s? The idea was that the magnetic fields from power lines (and computer monitors and the wiring in your house) caused cancer. The National Academy of Sciences reviewed the published literature and announced that power line radiation led to an increase in the rate of leukemia by a factor of two. But it was all bad science. The studies upon which the distinguished Academicians based their decision were riddled with bias.

experimental bias

Imagine that you want to find out if power lines cause leukemia. You find little Jimmy is sick and go to visit his house. You ask, “Do you live near a power line?”. “No”, “Are you sure?”, “I think I lived near one in my previous house”. You put a little check next to Jimmy’s name. You have just introduced observer bias into your experiment. As a researcher you are a little attached to the project you have been working on for two years. And you know that Jimmy has leukemia, surely he lived near a power line at some time.

On top of this, Jimmy has introduced recall bias into the experiment. He has leukemia, there must be a reason he got it. The guy in the lab coat in front of him seems to be implying that it has something to do with power lines and he heard something like that on the news a while ago, and, yeah, he does kinda remember a power line near his old house!

You see the problem. A good experiment avoids these issues by running a “blind” experiment. This means, for example, that the researcher should not know if the person he is interviewing has leukemia or not. Recall bias can be avoided by actually measuring the strength of magnetic fields in houses of leukemia sufferers and a control group of people who have no cancer.

In the case of power lines, such blinded experiments were eventually performed. No surprise here: the evidence for higher leukemia rates vanished. Power lines don’t cause cancer.

Its easy to dismiss the whole saga as a nuisance for a minority of people. But it wasn’t — there are estimates that the power line fiasco cost American taxpayers twenty billion dollars. That’s $300 of our tax money per family.

There are Studies and there are “Studies”

After the humiliating and expensive experience with power lines, things must be better now, right? Well, maybe not. The European Union has spent $24 million conducting an enormous study of the correlation of cell phone use with the incidence of brain cancer, called the Interphone Study. Remember, everyone uses cell phones, so to show an effect you have to show that people who use cell phones more tend to get brain cancer more. Brain cancer is a rare disease, so these numbers have to be collected very carefully. Are they? Not even close.

Imagine you are a cancer researcher in Sweden, you go to the hospital to interview poor Mrs. Sundstrom who is dying from brain cancer. You ask her, “How much did you use your cell phone?” “Oh, not much”. “Are you sure?”, “Well I did use it quite a bit when I was trying to corner the pickled herring market.”. Its our old friend observer bias. And Mrs. Sundstrom sees a man in a white coat in front of her asking about cell phones, and she’s heard something on the news about cell phones, and there’s got to be a reason she got brain cancer. Yep: recall bias.

It is difficult to know if the Interphone Study group has implemented blind methodology to avoid observer bias. However, there are published reports stating that the study suffers from recall bias. The researchers are asking cancer victims about their cell phone use, not measuring it (say, via provider records). As a result final publication was delayed for three years.

In an attempt to paper over the cracks, Interphone is tried to compare people’s recall of cell phone usage with actual cell phone usage. Canthey actually account for the different recall between healthy and cancerous people? Can they do it with any accuracy? It seems unlikely for a number that is already difficult to tease out of the data. The upshot is that the Interphone Study is rife with experimental bias. They may as well have flushed $24 million bucks down the john.

The Cell Phone Panic

In March of 2008, the Independent, a British newspaper reported that a study by Dr. Vini Khurana shows that mobile phones could kill more people than smoking. The story was quickly repeated on this side of the Pond by Fox News. A few months later Tara Parker-Pope of the New York Times stirred the pot with two stories, Experts Revive Debate Over Cellphones and Cancer and How Much Radiation Does Your Phone Emit?. In an alarming introduction, the first article says,

Along with Senator Edward M. Kennedy’s recent diagnosis of a glioma, a type of tumor that critics have long associated with cellphone use, the doctors’ remarks have helped reignite a long-simmering debate about cellphones and cancer.

To be fair to Parker-Pope, she does say,

“Cellphones emit non-ionizing radiation, waves of energy that are too weak to break chemical bonds or to set off the DNA damage known to cause cancer. There is no known biological mechanism to explain how non-ionizing radiation might lead to cancer.”

Of course she is implying that we don’t know everything, and therefore cancer-from-cell-phones could be a real effect. While it is true that biology is just taking off as a quantitative science, electromagnetic radiation is the best understood physical phenomena in the universe.

If cell phones cause trouble, why not wi-fi? Sure enough, the BBC reported that there is concern about putting wi-fi near schools. Wi-Fi radiation has a frequency of around 2.4 GHz, similar to cell phone radiation, making it about 500,000 times too weak to cause DNA damage.

Tara Parker-Pope of the New York Times continued to stir the pot with a report about a study (Int J Cancer, 120, 769 (2007)) the supposedly provides evidence that cellphones cause cancer. Does it? As you can see yourself by following the link, the authors say straight out:

“We found no evidence of increased risk of glioma [a kind of brain cancer] related to regular mobile phone use.”

Furthermore,

“No significant association was found across categories with duration of use, years since first use, cumulative number of calls or cumulative hours of use.”

Sounds pretty definitive to me! But there is a possible catch:

“For more than 10 years of mobile phone use reported on the side of the head where the tumor was located, an increased OR of borderline statistical significance (OR = 1.39, 95% CI 1.01, 1.92, p trend 0.04) was found, whereas similar use on the opposite side of the head resulted in an OR of 0.98 (95% CI 0.71, 1.37)”

The authors are saying that they saw a slight increase of incidence of glioma on the side of head that the cellphone was used versus the far side. However, the increase is called “borderline” meaning that it could be simply due statistical fluctuations in their data. The way to make a definitive statement is to collect a lot more data (examine a lot more patients). Of course, if the original comparison of cancer incidence in cellphones users and non-cellphone users continues to find
no increase in the rate of cancer for cellphone users, this scenario would be impossible to explain. Why would brain tissue 4 inches removed from a cellphone be less susceptible to cancer than that removed by hundreds of feet? (Here I’m guessing a typical distance between a non-cellphone users brain and somebody else’s cellphone)

Not to be outdone, in July 2008 the CBC reported that Toronto’s Department of Public Health wants teenagers to limit their cell phone use to less than 10 minutes per day. Loren Vanderlinden, the health department supervisor cites an emerging pattern in studies suggesting greater risk of brain tumors from long term cell phone use. And what is Vanderlinden’s expertise you ask? She is an anthropologist.

Shortly after this, Dr. Ronald Herberman, head of the UPMC Cancer Center and the University of Pittsburgh Cancer Institute issued an internal memorandum warning staff members of the health risks of cell phones. This was subsequently picked up by local newspapers, the Pittsburgh Tribune-Review and the Pittsburgh Post-Gazette. NPR followed shortly after.

In the memo Dr. Herberman wisely says

We shouldn’t wait until definitive information comes out. By then, we might have a virtual epidemic on our hands.

Wait. That’s the reason? By this logic we should stop doing anything that has a remote chance of being dangerous – like stepping outside. Admittedly, there is a huge downside to this risk. But we’ve already seen that the physics forbids it and there is no evidence of any change in the incidence of glioma.

Thankfully, other reasons were quoted. For example, Dr. Herberman mentioned that the Toronto Department of Public Health has issued an advisory. As we shall see, the entire alarmist story has a self referential nature, with a few people pushing a viewpoint, others picking it up, and all of them referring to each other for “proof”.

It turns out the Herberman memo was pushed by Devra Lee Davis, who was the director of the UPMC Center for Environmental Oncology at the time.

A professional epidemiologist will surely straighten out this mess! Or not. Davis says,

The question is do you want to play Russian roulette with your brain? I don’t know that cell phones are dangerous. But I don’t know that they are safe.

It must be intuitive to most people, even a doctor, that one cannot prove a negative. Can you prove that Santa Claus does not exist?

Who is our dear Dr. Davis who makes such outrageous statements? Some web searching established that she holds a B.S. in physiological psychology and an M.A. in sociology from the University of Pittsburgh. According to her web site she completed a Ph.D. in “science studies” at the University of Chicago and has a Masters in Public Health. Not bad!

Hang on a sec, what is “science studies”? I’m not sure, but the University of Chicago web site reveals that her Ph.D. thesis was on the conceptualizations of religion and science in the work of Immanuel Kant and Auguste Comte. So she has a philosophy degree. Usually philosophers end up driving cabs, so she’s done quite well for herself.

Dr. Davis is also an author. Her last book is The Secret History of the War on Cancer. A secret history, how exciting! I looked up the review of this book in The New Republic. Here is part of what the reviewer had to say:

“The Secret History of the War on Cancer is a hysterical and exasperating book. It has no overall structure and it repeats itself in the sloppiest ways. It meanders into stories and themes that have no obvious pertinence to cancer and the war on cancer. And worst of all — this being a protest, after all, against the state of medicine and science — are its New Age tones. A well-trained and reputable cancer epidemiologist might be expected to produce a book based on research and data rather than unsubstantiated claims and anecdotes. But throughout The Secret History of the War on Cancer there is a continuing homage to “treatments” that have not even been subject to serious research, much less proved safe or effective.”

Whew! That is not exactly a ringing endorsement of Dr. Davis’s scientific expertise!

I wrote an op-ed about cell phones and cancer a few years ago. As a result Dr. Davis and I exchanged some emails about the issue and she kindly pointed me to the
Bio Initiative web site. What I found there was a dozen angry doctors giving a long list of the woes that low energy radiation causes. One of these was Dr. Lennart Hardell, who we will meet later. At the bottom of the list I found Dr. Louis Slesin.

Slesin, it turns out, is the founding editor of Microwave News, a newsletter and blog carrying on the good fight against the evils of low energy radiation. As usual, his credentials are impeccable: he got his MA in Chemical Physics from Columbia and a Ph.D. in Environmental Policy from MIT. But we have learned that credentials don’t mean much in this game. Slesin is a strong (to this day) proponent of the discredited idea that power lines cause cancer (and Lou Gerhig’s disease and miscarriages). hat does an old anti-power-line warrior do? He upgrades to cell phones!

The French Fatwa

On June 20, 2008 twenty French cancer and public health specialists issued a public appeal in the Journal du Dimanche calling for prudence in cell phone use. The organizing force for the appeal was
Dr. David Servan-Schreiber. Dr. Servan-Schreiber obtained his MD from the University of Laval and a Ph.D in neuroscience from Carnegie Mellon University. He spends part of his time as a
clinical professor of Psychiatry at the University of Pittsburgh. So far so good, but continue reading his biography and you discover that he was director of the Center for Complementary Medicine at the University of Pittsburgh’s Shadyside Hospital. Complementary medicine? Like aromatherapy and acupuncture and coffee enemas?

Questions fly to the mind: why do cell phones attract the New Age crowd? Why does Pittsburgh attract New Agers? Can crystals really cure impotence? Is my crystal more powerful if I form it into a pyramid shape?

Metastudies find Metacellphones cause Metacancer

In 2009 the LA Times breathlessly reported that a new article published in the Journal of Clinical Oncology says that cell phone users suffer 10 to 30 percent more brain cancer than people who rarely use cell phones.

Could our simple physics argument be wrong? Let’s take a closer look at the article.

An analysis published Tuesday of data from 23 epidemiological studies found…

So the research is actually a metastudy, which means the authors looked at a collection of other studies instead of doing their own work. Sounds kinda lazy, but presumably more studies means more data means more reliable conclusions. So what did they find?

… no connection between cellphone use and the development of cancerous or benign tumors.

What? So what’s the stink about? Let’s read on:

But when eight of the studies that were conducted with the most scientific rigor were analyzed, cellphone users were shown to have a 10% to 30% increased risk of tumors.

Wait. What? The authors just eliminated 15 of the studies? Are you allowed to do that? The statisticians have a clear answer to this question: no, unless you have compelling reasons to discount some of your data. So what do the authors say?

“The other group of 15 studies were not as high-quality,” said study coauthor Joel M. Moskowitz, director of the UC Berkeley Center for Family and Community Health. “They either found no association or a negative association or a protective effect — which I don’t think anyone would have predicted.”

Is that it? Data was discounted because it “either found no association or a protective factor”? Is this a joke? I think Bernie Madoff argued this way too: “I always make money on the stock market, except when I don’t”.

Maybe its not quite as bad as it sounds; the article goes on to say that the eight damning studies were singled out because “the researchers were not told which people had tumors and the studies were not supported with mobile phone industry funding.” Fair enough. But then the other shoe drops:

However, seven of those eight studies were conducted by a single researcher, Dr. Lennart Hardell, an oncologist in Sweden.

It turns out Hardell is a professor of oncology for University Hospital in Orebro, Sweden. He says

“By now, evaluating all international studies in this area, there is a consistent pattern of a nearly doubled risk for brain tumors (glioma and acoustic neuroma) on the side of the head where a mobile phone has been used for at least 10 years”

Wow, that’s pretty definitive! But he is talking about part of the famed and flawed Interphone study. The same one that is riddled with recall bias and internal rife.

He is also the principle author of Secret Ties to Industry and Conflicting Interests in Cancer Research. Can we really trust researchers who write hysterical books about how industry is trying to give us all cancer? How “fair and balanced” can he possibly be?

There is a lot of crappy research out there. It is the responsibility of good scientists to filter this out, not help propagate it. This nonsense must stop or we are in real danger of repeating the power lines-cause-cancer fiasco of the 90s.

Politicians Step in

Politicians work in a world where perception means more than truth. So it is no surprise that many have stepped into the cell phone debate. San Francisco city council spent some time considering requiring warning labels on cell phones. Not to be outdone, the Council of Europe called on all 47 member states to “ban all mobile phones, DECT phones or WiFi or WLAN systems from classrooms and schools, as advocated by some regional authorities, medical associations and civil society organisations.” Wow. And D-OH Dennis Kucinich had called for the implementation of warning labels on cell phones.

The World Health Organization Steps In

On May 31, 2011, a committee of the International Agency for Research on Cancer, a part of the World Health Organization (WHO), reversed previous statements and declared cell phone radiation as “possibly carcinogenic to humans (group 2B)”. Now what? Surely an august body like the World Health Organization would soberly assess the situation and not needlessly scare billions of people!

But hang on a second, what exactly does “group 2B” mean?

I went to the WHO web site and found a long list of things that are “group 2B”, including a bunch of funky chemicals, carpentry and joinery, coffee, diesel fuel, gasoline, firefighting, magnetic fields, nickel, pickled vegetables, and talc. “Coffee”? “Carpentry”? Seriously, WHO? Seriously?

Even if I believed their conclusions for a second (and I don’t), it appears that cellphones are not particularly dangerous. So what is the evidence that the WHO is using to frighten five billion cellphone users?

First, the committee did no new research, they simply reviewed previous studies. Their press release said they found a positive association between cellphone use and a type of brain cancer called glioma. For details the committee referred the reader to a report that would appear in a month.

[Let me say right here that this delay is scientifically unacceptable and morally dubious. You don’t go around saying “Danger!” and then say, “I’ll tell you why in a month”. I appeal to the IARC to change their press release policy.]

Waiting a month to find out how I had gone wrong was not easy, but an article finally appeared in the Lancet. In short, their conclusion are based on two things, an article by the notorious Lennart Hardell and one number in the Interphone study. This study reported lower than usual brain cancer rates for all cellphone users except for those ten percent with the highest usage rate, who had a 40% greater than normal rate of glioma (but not another brain cancer called meningioma). The Interphone researchers noted that the statistics associated with this group were likely biased and that the error in the reported 40% increase was huge, ranging from no increase to a 90% increase. It is no wonder they concluded that “overall, no increase in risk of glioma or meningioma was observed”.

Apparently the IARC committee begged to differ. They also chose to ignore dozens of other studies that found no effect (for example one from 2009 in the Journal of the National Cancer Institute).

Let’s look at the Interphone results in more detail. What you see in the chart below is the “odd’s ratio” for getting glioma plotted versus the number of hours people have used their cellphones. An odds ratio of 0.7 means that one has 70% chance of getting glioma versus people who do not use cellphones. The black lines on the bars indicate the errors in the odds rations obtained by the Interphone researchers.

The total odds ratio for glioma was 0.81, implying that cellphones suppress brain cancer. This is an odd result that the researchers admit could indicate a problem with their methodology. Now imagine dividing the population examined into ten groups. Some of them are going to have odds ratios less than 0.81 and some are going to have them greater.

Now comes the important part. The Interphone study had no way to determine cellphone usage rates except asking the study participants. Imagine you have brain cancer and someone tells you they are doing a study and asks you how much you use your cellphone. Could this be the reason why you have cancer? And, yes, you do use your phone a lot! Like we discussed above, this is called recall bias. In fact it was mentioned by the Interphone researchers as a possible distortion of the data because many people reported implausibly intense cellphone usage.

The end result will be a higher odds ratio than expected for the highest usage rate, exactly as seen in the chart. Finally, notice that the errors in the highest usage rate are very large. Because of these effects, the Interphone study did not conclude that cellphones cause brain cancer.

It is important to recognize a fact of life about doing studies: experiments often involve random events (such as getting cancer) and thus have a “significance” associated with them. Saying that an experimental conclusion has a significance of 99% means that the researchers judge that the odds of their conclusions being wrong are one in a hundred. This means that if two hundred similar experiments are performed, on average two of them will yield an incorrect result. Why do we care? Thousands of experiments concerning the interactions of the human body with electromagnetic radiation have been conducted over the past century, some of them are going to (incorrectly) find an effect. That is what statistics tells us. Alternatively, if you do a single experiment but look for hundreds of possible outcomes, some of them are going to (incorrectly) look like a causal effect. And, like our case here, if you divide data into many groups, some of them are going to be high.

It doesn’t take a PhD to understand these simple facts about statistics. But having a PhD doesn’t guarantee they are understood either. The WHO committee has decided to ignore dozens of studies that find no effect due to cellphones and focus on the very few that do. One could argue that this is simply being prudent, but I do not call needlessly frightening five billion people prudence, I call it recklessness.

Doing Studies

I don’t want to give the impression that all studies are bad. There are many cases where epidemiological studies have dramatically improved the lot of mankind. Among them are Percivall Pott’s demonstration in 1775 that soot was causing scrotal cancer in chimney sweeps, John Snow’s proof that cholera is water borne in 1854, and the discovery that cigarette smoke causes lung cancer in the 1950’s. The latter was not easy, in a 1947 conference on the causes of lung cancer everything except cigarette smoke was blamed: influenza, fog, X-rays, road tar, the common cold, coal fires, industrial pollutants, gasworks, and car exhaust. Sometimes we scientists are a little dull, but we get there eventually!

Yet something seems wrong with the medical research system. In fact studies show that 80% of nonrandomized studies turn out to be wrong. And 25% of gold standard randomized studies are wrong. Medical researchers have powerful incentives to make a splashy discovery. And boring null results tend not to be published. When these issues are combined with sloppiness in experimental design and procedures, false results are the inevitable result. This is a regrettable feature of our medical research system and is something we have to live with. Better training, better vetting, and better procedures (such as registering studies before they are done) can improve the situation immensely; unfortunately there appears to be little movement or motivation in the medical community to implement the required changes. So, take everything you read with a grain of salt (but not too many, studies indicate it is bad for you).

Cell Phones and Brain Metabolism

By 2011 the cellphones-cause-cancer crowd was getting a bit desperate and was in need of some “good” news. Fortunately for them, along came Nora Volkow. Volkow is a doctor specializing in psychiatry and is the director of the National Institute on Drug Abuse. She is also an expert on positron emission tomography. One day Dr. Volkow had a great idea: why not use PET to see if brain activity is affected by cell phones.

Here’s how they did it. Forty-seven subjects had cell phones strapped to their left and right ears and were given fluorodeoxyglucose injections — sometimes with the right phone “on” and sometimes with both phones off. Fluorodeoxyglucose is a radioactive chemical that can be detected by the PET imager. It acts like glucose in the body and therefore provides a neat proxy for metabolic activity.

The study was done in due course and published in the prestigious Journal of the American Medical Association. It was also picked up by the New York Times.

Remarkably Volkow and her team found an effect! Scans of people with cell phones “on” revealed higher rates of glucose uptake. It seems that cell phones can affect the brain by changing metabolic activity. As the authors say

“Whole-brain metabolism did not differ between on and off conditions. In contrast, metabolism in the region closest to the antenna (orbitofrontal cortex and temporal pole) was significantly higher for on than off conditions. The increases were significantly correlated with the estimated electromagnetic field amplitudes.”

The authors also kindly include a figure showing the difference between scans in the “on” and “off” states.

It all looks pretty convincing. But look at their own words again. The whole-brain metabolism did not differ between on and off conditions, yet activity (they say) went up near the on phone. So activity must have gone down on the other side of the head. How do you explain that? If it goes up because of the phones electromagnetic radiation then it should go up less further away from the phone, not down.

While we are at it, what do the authors mean by “on”? Further reading reveals that they simply turned the cell phones on. Did they not know that cell phones that are on are mainly quiescent, merely pinging the cell network now and then? They needed to have the phones on and transmitting!

And what does the last sentence mean? Increases were correlated with EM fields? In fact one of their figures shows this correlation.

One sees a positive correlation between glucose metabolism and “E” — the electric field strength (due to the phone) inside the head. But surely Volkow and friends didn’t drill electric probes into the skulls of their volunteers. No, “E” is an estimate of the field strength in the head. Reading on we find that this is computed by assuming an dipole electric field. Hold on there cowboy! Did the authors just assume that a cell phone antenna is a classical dipole? Do they not know that cell phone antennas are fractal? All right that’s a blunder — but assuming that the electric field freely penetrates the brain is an even worse mistake. Human cells are essentially bags of salty water and are excellent at screening electric fields. In fact electric fields barely penetrate past the skin.

Its starting to look bad for our intrepid doctors (I hesitate to call them scientists). But it gets worse. Take a look at the graph again. The solid black circles represent the data points (plotted against the imaginary electric field strength). The lines attached to the circles are called “error bars”. They indicate the statistical error in the measured quantity. If the experiment were repeated many times the points would fluctuate up and down, but would land within the error bars 68 percent of the time. So far so good. But look at how nicely the points line up (the authors provide a line in the graph to show this). They are far closer to the line than warranted by the (authors’ own) error bars! There are 43 points in the figure, so on average 13 of them (43 * (1 – 0.68)) should have been further from the line than their error bars. Not a single point lies this far from the line! Something is seriously wrong with their statistical analysis. Given how sloppy the rest of the experiment is, I am not surprised.

Observations on a Manufactured Crisis

Although the cell phone scare seems to have died down recently, it continues unabated in India (look here or here), Africa (article), or closer to home. It also lives on in the minds of true adherents. Louis Slesin and pals continue their life calling. Ronald Herberman has left UPMC and now works in private industry. Devra Davis has also left UPMC and has set up a private foundation to continue the good fight and sell her books. Dennis Kucinich’s district was redrawn and he lost the primary. Apparently he plans to work for Fox News. It is hard to imagine sinking lower.

In the end this is not a story about a set-to in the scientific community. The science issues have long since been settled. The real issue is how this meme came to reside in so many minds and how it maintains its strength.

The answer lies in a complex mix of bad science and human frailty. Almost all crazy ideas concerning science have originated with scientists or doctors. The problem here is that researchers tend to be attached
to their pet ideas and can be remarkably stubborn in the face of contrary evidence. Doctors are people too, after all, and one must expect this. And in a sense its not bad — perseverance is a good quality to have when doing research.

Bad research is often followed by bad reporting. The media seizes on sensational results, re-interprets them for the public, and spreads the word. The distortion of a distortion is then taken up by a receptive public. If a false idea makes it this far, it is generally impossible to reverse the process. The idea takes on a life of its own as “popular knowledge” and no amount of backtracking or correctives can squelch it. The whole process is indicative of a public with a deep distrust of the establishment. In many ways this is a good thing (We’re spending how much on littoral combat ships? How many phone conversations are being tracked?) But the skepticism has spread to conventional medicine, nutrition, climate, evolution, nuclear power, and countless other scientific activities.

Somewhere a line of trust needs to be drawn. Where it is placed will affect all of us for a very long time. Let’s not screw it up.

Feel Good Moment Turns into Horrifying Indictment of America

What was supposed to be a feel-good story about a kind-hearted waitress has unexpectedly and wretchedly morphed into a damning indictment of inequality in America.

Sarah Hoidahl works at a Ruby Tuesday in New Hampshire where she recently picked up the $27 lunch tab for two furloughed National Guardsmen. Ellen Degeneres caught wind of this, invited her on the show to talk about it, and surprised her with a $10,00 check as a reward. You can watch the heartwarming CNN report here:

After fatuously noting that what she did “turned into some great entertainment”, the CNN reporter asks her what she is going to do with “all that money”. Here’s where things turn tragically, pathetically, execrably damning. She replies that she plans to

• pay off some medical bills
• pay off some student loans
• she has other debt to pay down
• she owes money to her son’s daycare
• she wants to make a donation to charity
• and she wants to buy a car.

Sarah, honey, you need a fuck of a lot more money than $10,000!

Let’s think a minute about her utterly depressing to-do list. She’s likely a college graduate, she owes money to the government for student loans, probably is in hock to credit card companies, had health problems, almost certainly has no health care, is probably a single mom, has no car, and works as a waitress.

I salute you, Sarah Hoidahl, for being a caring person even though America has collectively taken a giant shit on you. Even with a college degree, Sarah has to work as waitress at Ruby Tuesday, condemning her to a hand-to-mouth life in the wretched underbelly of the American economy (to Ruby Tuesday’s credit, they do offer some benefits for their employees). Decent health care for the working poor? Perish the thought! Pure Bolshevism!

Sarah is working at or near minimum wage, which is $7.25 per hour. That is not a lot of money. In terms of purchasing power parity (a way to compare international currency), $7.25 an hour places us behind Luxembourg, France, Australia, Belgium, the Netherlands, Ireland, the UK, New Zealand, and Canada in the OECD countries (all the stats you want are here). When you factor in the cost of living(*) we fall to fifth last in the OECD, ahead of only Mexico, the Czech Republic, Estonia, and Slovakia. This is not the behavior of a rich, “exceptional”, or caring nation.

How hard is it to live on minimum wage? McDonalds teamed up with Visa to create a website that kindly offers financial planning advice for its employees. Their sample budget had the employee working 40 hours week at $7.72 an hour. After allowing a ludicrous $600/month for rent and $20/month for health insurance and other expenses, the imaginary employee was left with $800/month to pay for food, entertainment, travel, Christmas, and all those other expenses that eat into budgets. Here’s the kicker: to get the budget to balance McDonalds and Visa assumed that the employee worked a second full time job! Yes, you need to work 80 hours a week in America to afford your imaginary $600 apartment and $20/month health care.

You don’t need McDonaldian corporate ineptitude to see that something is going horribly wrong in America. Decades of business-friendly legislation, bought legislators, billionaire club lobbying, offshoring, and the slack-jawed lackeyism of our media have brought inequality in America to third-world levels. The gini index for the USA (a measurement of purchasing power inequality) has been climbing steadily since the 1970s. Of the 34 countries in the OECD, the USA stands fourth worst, ahead of only Chile, Mexico, and Turkey.

We all know that it takes money to make money. The natural outcome of this law is that wealth is concentrated into an ever-shrinking tranche of the population. But just how much it has accumulated is shocking and depressing. The top 1% of the population owns 35% of the national wealth; the same proportion as the bottom 95%.

Surely, in our guts, we can’t think this is a good situation. How will it end? Will the top 1% own nearly everything while the rest of us are doomed to groveling servitude? Doesn’t that sound like a recipe for revolution? People will accept a lot of abuse, but at a deep level they have to feel that they are being treated fairly. The ridiculously low minimum wage simply allows McDonalds to earn more corporate profit while the rest of us foot the bill for the food stamps and medicaid that their employees are forced to use. Something needs to change. A simple start would be to raise the minimum wage.

————————–
(*) I computed this by dividing PPP minimum wage by PPP GDP per person.