“Information is power” (not)

As I raced through Waterfront Station last week, late for something or t’other, I overheard a well-dressed young man intone, “Information is power.” The pretty young woman he was with enthusiastically agreed.I moved on, dodging slower pedestrians and trying to figure out why and how such a cockamamie truism had taken such a stranglehold on us all. Information is power? Says who?

Information, aka data, is not even knowledge, never mind power. Without context, without a hypothesis, without a narrative of some kind, simply having access to Google and disparate bits of information means nil.

Anyway, isn’t the usual phrase knowledge is power? And even knowledge, moreover, rarely translates into power- unless you’re a blackmailer.

Unfortunately, the cliché has taken off and far too many people actually believe that having access to information, be it minute by minute stock/business data, medical information or Google (“facts”) actually means something.

Let’s take one example of basic information: observational studies. We observe one thing, see that it seems to happen whenever something else does and presto, we have a correlation on our hands that we conveniently forget are simply disembodied bits of information that probably mean nothing – but which we assume reflect cause.

Epidemiology had one major success with that: cigarette smoking and lung cancer. And they’ve never let us forget it. The problem is that almost no observational studies epidemiology threw at us ever turned out to be accurate when properly scrutinized.

Take the estrogen debacle. For years observational studies and epidemiology insisted that women who took estrogen, particularly at midlife, were healthier, lived longer, had fewer heart attacks and even suffered less from dementia. The problem was that this link came from well-off women who had the time to fill out those surveys and questionnaires, which meant they were better educated and of a higher socio-economic background. This meant that they were also healthier – in fact there’s even a name for it: the healthy user bias. People who are from a higher socio-economic status are healthier. Period. Why? We don’t know.

Perhaps they eat better or have less stress; perhaps they have better genes and it’s that which has led them to be better off in the first place. Maybe they breathe cleaner air and live in nicer areas where they don’t breathe or step in toxic gunk. After all, it’s not the CEO of a company, whether in India or the United States, whose house abuts the factory runoff, it’s the hapless janitor and his family who can’t afford anything better.

The media loves reporting on observational studies, where the inevitable term used in the headline is “linked”. Vitamin D is “linked” to better health. In healthy societies women’s choice of mate is “linked” to more masculine features, which naturally means that evolution has had something to do with the preponderance of older men marrying women young enough to be their daughters.

We forget that correlation has nothing do with cause. The Women’s Health Initiative clinical trial was stopped early because it was the women who took estrogen who were dying in droves from breast cancer and heart disease. All those years researcher upon researcher had insisted that their hypothesis, namely that estrogen was the female hormone, was right and had absolutely nothing to do with the binary nature of our socio-cultural classifications. But gosh, they were wrong.

Failure of imagination follows in the tracks of information – simply knowing that something happens tells us absolutely nothing about why it happens or whether manipulating one factor will have an effect on the other. It could be incidental, an artifact, or just plain wrong.

Still, we walk around, secure in the knowledge that our platitudes, like information, give us an edge.  But it’s not power, just swagger.