The economist Emily Oster, on becoming pregnant, was surprised by how woolly much of the medical advice was – how restrictive, and how many of the restrictions seemed to be without foundation. "The key to good decision‑making," she writes, "is taking information, the data, and combining it with your own estimates of pluses and minuses." Naturally, as anyone who has been pregnant will know, that is the exact opposite of the way things work: you are given no data, you are told which decision is the best, and if you ask about data, the medics give you a look as if to say: "One more smart alec remark and it's the child-protection register for you, chum."
So she sets about compiling the data herself, does so comprehensively and simply, and gives her own conclusion, along with a little box, "The Bottom Line". The old-fashioned reader, who doesn't find a break-out box necessary for the comprehension of material, will find this annoying. But that is a quibble. There's a quote from Steven Levitt, author of Freakonomics: "This may be the most important book about pregnancy you read." The way these things are, with their both being hot popular-economics property, and living in the same country, his judgment could be an act of social generosity ("quote whoring", I believe is the technical term). In fact, by the end, and as unlikely as it sounds, I agreed with him.
New truths are still being published about the mechanics of human gestation, certifiable facts that for some reason modern western medicine will not acknowledge. For instance, alcohol is not as dangerous as you think, and nor is caffeine; antenatal screening tests have advanced considerably over the past 20 years, while the rumours you hear about their risks have not changed; the obsession with maternal weight-gain (more prevalent in the US than in the UK) appears to be an issue of taste among professionals, and has very little to do with the health of the mother or the baby. I have never seen the numbers crunched so satisfyingly, the findings delivered so neutrally, with only the barest whiff of "well, would you look at that? I appear to be right again" to the tone.
Discussing the first two weeks of pregnancy, during which women tend not to know they're pregnant and about which they worry for the next eight and a half months, she writes: "For the period between fertilisation and your missed period, your baby is a mass of identical cells. If you do something that kills one of these cells (such as heavy drinking or some kind of really bad prescription drug use), another cell can replace it and do exactly the same thing. The resulting baby is unaffected. However, if you kill too many of these cells, the embryo will fail to develop and you will wind up not pregnant at all." Perhaps you knew that, and I just run in an exceptionally ignorant crowd. But I've never seen it on a leaflet.
Her section on alcohol will astonish anyone who believes the advice of the UK's chief medical officer (one to two drinks, once or twice a week; reduced to total abstinence for a time in 2007, and then subtly notched back up) or their US counterpart (the Center for Disease Control doesn't just advise, but "urges" abstinence from alcohol in pregnancy). Not only do studies repeatedly fail to bear out any harmful effects of light-to-moderate drinking, they usually show the children of light drinkers to have fewer behavioural problems than those of teetotallers. Oster articulates well the antiscience science of it: "One phrase I kept coming across was 'no amount of alcohol has been proven safe' … this seemed to me to have two problems. First, too much of any goods can be bad. If you have too many bananas, the excess potassium can be a real problem. But no doctor is going around saying 'no amount of bananas have been proven safe!' … Second, what is all this evidence if not proof? It's exactly this type of evidence that leads us to conclude that binge drinking is problematic. But if you are willing to conclude that, why wouldn't you be willing to conclude that light drinking is fine? That is what the evidence shows."
Risk aversion and the "overinterpretation of flawed studies" characterise obstetrics. Oster identifies another aspect of this, in her section on morning sickness: "Dwyer, the friend with the terrible nausea, was told that she could take a prescription drug 'if she really felt like she needed it' … she came away thinking it was dangerous for her baby, but if she cared only about herself, she could take it. Who would be comfortable taking anything at that point?" This is key – the sense so often conveyed by medicine that merely by seeking a solution for a pain or ailment, you are failing to put your baby first; that some fundamental part of motherhood is embracing a problem rather than trying to solve it.
Some centuries ago, perhaps most enthusiastically in Scotland, the idea of predestination took hold. God had decided in advance whether you were saved or damned, there was nothing you could do. But that emphatically didn't mean you could behave however you liked; in manifesting holy behaviour, you alerted the rest of the community to your membership of the elect. I feel that pregnancy has gone that way – the exaggerated presentation of risk takes the place of this toss-of-a-coin God. In other words, there's no way we can really tell you what's safe, there's no way you can really dodge the wrath of the Lord, but by manifesting acquiescent behaviour you show society that you are a good mother, that you'll win at mothering.
It's a creed in which medicine kowtows to superstition, women are redomesticated and stupefied, with the stated intent of protecting the foetus, but in fact to shore up a vision of the world in which, if it looks predestined, tough tits, that's because it is. Emily Oster would never say any of that, of course, because it's wild speculation, and she deals in data. But her book has done something important to counter this creed – indubitably important for pregnant women, but vital, too, for everyone else.
• Zoe Williams's What Not to Expect When You're Expecting is published by Guardian books.
