April 18, 2018 | 17 Comments
I received this request from Steve Blendell (slightly edited for spelling):
How are you friend? Take a look at Prof Cotter’s letter – he’s a physicist. Do the stats stand up?
The referendum is in May – our side have got off to a good strong start with posters.
The referendum is whether to repeal the Eighth Amendment which gives human beings a right to life. ‘No’ voters think killing the lives inside would-be mothers should be illegal, while ‘Yes’ voters want to draw their knives.
Ignore here the conceit, shared by all democracies, that such matters can be put to a (general) vote.
Cotter’s letter to the editor:
Sir, – Posters on my street for the No campaign state that the rate of terminations in England is either one in four (25 per cent) or one in five (20 per cent), depending on which poster I look at. It is also interesting to note that these data only refer to England. The reason for this is that if you include official 2016 statistics for Scotland and Wales, the overall rate drops to 14 per cent. Now 14 per cent is a long way from 25 per cent and doesn’t look good for the No campaign. So voters need to be aware of how statistics are being manipulated to encourage a no vote. – Yours, etc,
THOMAS G COTTER,
Cotter apparently believes the (if true) slightly lower number of killings in England, Wales, and Scotland justify killing multitudes more in Ireland. Which is incoherent. Either the killing is moral and allowable, or it isn’t. If it is, what’s the difference if the entire population decides to kill itself off?
Since that argument goes nowhere, let’s look at the numbers instead. Here is more or less what I told Blendell.
Here are the official statistics: Link (pdf).
They put the abortion ‘rate’ in England and Wales this way: ‘The age-standardised abortion rate was 16.0 per 1,000 resident women aged 15-44.’ That is calculated like this:
number of abortions/number of women aged 15-44 (in thousands).
That’s one definition of ‘rate’, but not the best if I understand them correctly. The best is
number of abortions/(number of births + number of abortions).
An equivalent way to put it is
number of abortions/number of conceptions.
Call this the Real Abortion Rate, and contrast it to the official rate. The Real rate will be higher, and likely much higher, than the number they are touting, which includes all women, whether or not they were pregnant.
Suppose only 1 woman in that age group got pregnant and then killed her child. That’s a Real rate of 100%, but it would be a very small official rate. To find it, take that 1 and divide by all the women (in thousands) aged 15-44. It’s in the thousands of thousands (millions), anyway.
I could not find what the Real rate is for England and Wales, but according to one chart in 2013 there were about 53,900 thousand people (roughly 54 million) in England and 3,100 thousand (3.1 million) in Wales. If women aged 15-44 were, say, 20% of these totals, then the total is 11,400 thousands women aged 15-44, more or less, in 2016.
Now that same report said there were 190,406 abortions in 2016. So that would make my estimate of the official rate per 1,000 women at
190,406/11,400 = 16,
which is exactly what they got, meaning that 20% guess of number of women in that age bracket is pretty good.
But if only 1 women was pregnant and killed her child, the Real rate would be 100% but it would make the ‘official’ abortion rate 1/11,400 = 0.00008, which is mighty small! This is only used to show that the definition of ‘rate’ matters.
More than 1 woman got pregnant. Here’s the official stats for England and Wales: Link.
Extrapolating would make about 900,000 conceptions in 2016, maybe slightly higher, maybe lower. They do not account for multiple births per woman, nor are miscarriages counted. But 900,000 is in the ballpark. That would makes the Real abortion rate about
190,406/900,000 = 21%.
That 21% is NOT per 1,000 women like the 16 above is, so be very careful making comparisons. This says (roughly) 1 out of EVERY 5 ‘conceptions’ are killed. Which is huge. That varies by age group, with (as the official report says) the highest rates around 22, i.e. the most fecund years.
Therefore this is how I would do the posters:
ONE OUT OF FIVE BABIES ARE KILLED IN ENGLAND & WALES.
Maybe accounting for uncertainties it’s 0.5 out of 5, or 1.5 out of 5. But 1 is a reasonable guess. I didn’t do Scotland, but you get the idea.
The numbers will all be meaningless. Statistics are (almost) useless. Those who want to kill do not care how many are killed. They just want to kill. Polls and bookies are predicting bloodlust wins, incidentally.
Image grabbed from here. Notice the hilariously inept ‘Trust us.’
Post corrected of my innumeracy. Bonus pic.