It’s not at all clear who actually coined the aphorism that there are ‘lies, damned lies and statistics’ but whoever it was was wrong.
Statistics are nothing more than meaningless figures until someone sits down to analyse and interpret them and it’s there that the problems arise. In, fact it would be more accurate to say that there are liars, damned liars and people who either lie about statistics or abuse them to lend an unmerited degree of degree of credibility to their lies.
The most prolific serial statistics abusers in modern society tend to be politicians and journalists and it’s to latter that we turn our attentions in this article.
Yesterday, the Office of National Statistics released it’s latest report on trends in conception rates in England and Wales, which include the most recent annual figures for 2011. The big, and very welcome, news from this report is conception, birth and abortion rates in teenagers have fallen yet again and are now at their lowest level for 30 years.
Now that’s good news…. unless, of course, you work for a newspaper that trades heavily on scaring the bejeebus out its readers with tall tales of hyper-sexualised feral teenagers shagging like rabbits, in which case you have a bit of problem because telling your readers the truth also means telling that pretty much everything you’ve been telling them for years about teenage pregnancy and abortion has been a complete and utter lie.
So what do you do?
Well, if you’re Steve Doughty, the social affairs correspondent at the Daily Mail, you carefully cherrypick the new ONS statistics in order to give your reading the wholly misleading impression that paints a picture of abortion rates in the England and Wales that is pretty much the exact opposite of what the data actually says.
So, this is what Steve has to say about abortion rates amongst women over 40:
Abortions among older women are falling fast as growing numbers choose to become mothers.
But figures published yesterday showed that, at the same time, abortion rates have been going up among younger women as more pursue higher education and careers.
The findings from the Office for National Statistics suggest that women over 40 are increasingly unlikely to regard pregnancy as a mistake, while many are actively trying to start a family.
The abandonment of abortion by the great majority of women who conceive over the age of 40 was revealed by records of conceptions in 2011 in England and Wales.
Conception rates among women over 40 went up 3.7 per cent in a year and have now more than doubled in two decades.
In 1990, there were 12,032 women aged over 40 who became pregnant. By 2011 that had reached 28,747. But in 2011 just over a quarter of those women went for an abortion, while a decade ago the proportion was almost 35 per cent, and 20 years ago it was nearly 42 per cent.
And now, here are the two graphs that expose Steve as liar. The first shows the real trend in conception, birth and abortion rates for women over 40 since 1998, as rates per 1,000 women, while the second shows the indexed rate for that same period (1998=100).
Women over 40 clearly aren’t ‘turning their backs on abortion’ because as you can clearly see the abortion rate for women over 40 has actually risen by around 17% since 1998, and around 20% since 2001, it just hasn’t risen as fast as either the conception or birth rates in that age group over the same period.
So, more women over 40 are choosing to get pregnant and have children when they’re over the age of 40 than was the case in 1998 AND more women over the age of 40 are terminating unwanted pregnancies today than they were in 1998, and is solely because the number of women choosing to have babies at that age has increased faster that than number of unwanted pregnancies that the overall proportion of pregnancies ending in a termination has fallen.
What about younger women, especially teenagers?
Remember, Steve’s already made the claim that abortion rates are going up amongst younger women as more pursue higher education or a career. Is that true?
Let’s look at the trends for under 16s, under 18s and under 20s, and on these graphs the solid lines shows actual annual data, again in rates per 1,000 women, while dotted lines show the underlying trajectory of the overall trend.
And would you look at that, conception and birth rates have both been falling since 1998-2000 and have accelerated downward since 2007, and even the abortion rate has taken a fairly significant downward turn since 2007.
Now, you may not be aware of this but the earliest published variants of the aphorism about lies, damned lies and statistics that anyone’s every managed to track down don’t actually use the word ‘statistics’ at all. What they say instead is that there are ‘liars, damned liars and experts’ and its just about at this point in Steve’s article that he decides to wheel out his chosen ‘expert’.
Earlier this month the ONS said that unemployment and worklessness are the key reasons influencing teenage pregnancy, and played down lack of free contraception or sex education as a cause.
Professor David Paton of Nottingham University Business School said yesterday: ‘Teenage pregnancy rates have only been going down since 2008, by which time the Teenage Pregnancy Strategy had already spent a lot of money.
‘The rate has still been going down since the TPA ended in 2010.
‘It confirms what we already know from the scientific literature – reducing teen pregnancy is not a matter of more contraception and sex education.’
He added: ‘One factor is that more girls want to stay on at school.’
Yes, it’s David Paton, the man from SPUC, not that Steve bothers to mention to his readers that Paton is a longstanding anti-abortion activist:
Dr David Paton, a Nottingham SPUC member, spoke at the SPUC 2002 National Conference last September, in his capacity as Head of the Economics Division at the Nottingham University Business School.
Paton’s been pitching the line about sex education allegedly not working for quite a while, in fact I dealt in some depth with one of his published ‘studies’ here and his comments in the Daily Mail are pretty much typical of his ‘academic’ work on teenage conceptions and abortions, i.e. for a Professor of Industrial Economics at a well-regarded university he exhibits rather a worrying lack of regard for the concept of long-term trends.
If you look at the underlying long-term trends on all three graphs you’ll see that Paton’s claim that abortion rates have only been falling since 2008 is only true for under 16s and that, on all three graphs, there’s a relatively sharp off-trend rise in conceptions and abortions between 2005 and 2007.
For a couple of years, in the mid 2000s, conception and abortion rates in under 16s, under 18s and under 20s went up and the reason for this is easily explain for all that Paton chooses not to comment on it.
It is a well-established and well evidenced fact that conception, birth and abortion rates are sensitive to rates of contraceptive use, and particularly to the numbers of women using oral contraceptives and this is, in turn, also sensitive to women’s perceptions of the safety or otherwise of oral contraceptives. In simple terms, whenever there’s a health scare surrounding the use of contraceptive pill, both conception and abortion rates will rise, usually for at least a couple of years after that scare hit the press, because some women will switch to less reliable forms of contraception in order to avoid the health risks they’ve been reading about in the press.
In 2005, there was just such a health scare, linking the use of oral contraceptives to an increased risk of deep-vein thrombosis and, consequently, an increased risk of strokes, pulmonary embolisms and heart attacks. The actual risk, itself, was minimum and focussed primarily on women taking long haul flights but, as usual, the media grossly overstated the risks and went into full scaremongering mode, because that sells newspapers, and the result of that can be clearly seen in those three graphs in the figures for 2006 and 2007.
As for Paton’s comment about rates continuing to fall despite the cancellation of the Teenage Pregnancy strategy in 2010, as you can see that’s nothing more than a continuation of the overall trend which was created, at least in part, by the work carried out under this strategy and any actual effects arising from the cancellation of the strategy will, in any case, take at least 2-3 years to filter through into the conception, birth and abortion data, so it’s far too early to say what impact, if any, that might have had.
What we can, however, say is that during the period that the Teenage Pregnancy Strategy was operating, the overall trend in conceptions in women under 16, 18 and 20 turned around from what is was prior to 1997, i.e. a rising trend, and the underlying long-term trend has been downwards in all three age groups since the turn of the millennium.
On top of the false claims about women over 40 and under 20, we’ve also got a claim about women in their 20s to tackle:
But abortion rates among under-30s have been rising as more pursue education and careers, and as women put off marrying or decide they cannot afford children.
Again, we have another graph showing conception, birth and abortion rates in the 20-24 age group which shows that claim to be utter rubbish.
Look carefully at the abortion trend (in green) and you’ll see that its not rising it all. It’s actually fallen slightly since 2007, although realistically you’d have to say that the long term trend since 1998 is pretty much a stable one, and the trend for women in late 20’s looks much the same as this one.
In light of Steve’s concerted efforts to show his readers that black is in fact white and that abortion rates are falling in women over 40s and rising in women under 20, when the opposite is true, I’d like to propose a modest alteration to the Daily Mail’s famous masthead, one that I think captures the essence of its approach to reporting facts.
I’ve done a bit of mock up, so see what you think…
As for Steve’s article on the ‘great abortion divide’, I suppose I should give it some sort of rating for factual accuracy, in keeping with the practices of established fact checking organisations, but rather than trying to give it marks out of 5, or maybe 10, I think Steve’s efforts can be summed up very nicely in just one word:
I’ve had a request from Mark (see comments) asking for the data on the trends for other age groups (25-29, 30-34 & 35-39) that I didn’t include in the original article and I am, of course, happy to oblige.
So, starting with the 25-29 age group, the trend looks like this:
Okay, so overall the abortion rate amongst women aged 25-29 is a bit higher in 2011 than it was in 1998 but also a little lower than it was in the mid 2000s, all of which adds up to a broadly stable trend lacking any significant changes over the entire period. As for conception rates, these are up since 2001 but have pretty much stabilised since 2007 although there is a small drop in 2011, just as their is in the data for the 20-24 age group.
Now remember, the Daily Mail’s claim is that abortion rates have been rising in women under 30, which they attribute to more women pursuing education and careers and to women putting off marriage and having children for financial reasons. What the graphs actually show us is something that all statisticians know very well but which many journalists routinely fail to understand; one data point does not equal a trend.
That leaves us with women in their 30’s to consider and the data for 30-34 and 35-39 age groups looks like this:
So, we have a pretty steady rise in conceptions and a very modest rise in the abortion rate in both age groups between 1998 and 2011 which is, again, indicative of a stable trend.
If we look this in terms of the average annual increase in the abortion rate then for women age 25-29, it’s 10 additional abortions per year for every 100,000 women in that age group, 17 additional abortions per 100,000 women in the 30-34 age group and 8 additional abortions per year per 100,000 women aged 35-39. It’s a rising trend, yes, but a very modest one and because of changes in these size of the female population in these age groups over time, what it translates into in terms of raw numbers is a fall in the actual number of abortions in each age group over time. In the 25-29 age group, the actual number of abortions carried out in 2011 was down just over 800 from the 1998 figure, while the 30-34 age group saw a fall of over 2200 abortions and the 35-39 age group a fall of 1300 abortions.
Now, if you’re not well-versed in the ins and out of handling statistics then this might all seem rather complicated, so to help out I’ve produced this handy infographic which summarises the differences between the Daily Mail’s interpretation of the latest statistics on abortion rates and the actual figures reported by the Office for National Statistics.
5 thoughts on “The Great Abortion Divide”
From the figures in this article it does seem like a completely wrong statement of the situation has been given by the Mail writer for the over-40s, but in (the extract you provide of) his article he appears to be contrasting over-40s with under-40s, with some additional reference to under-30s.
Your article provides figures for over-40s and under-24s, so the Mail writers statement about the younger age groups can’t be verified (or not!) based on what your have presented. Can you add in the data for 24-40 age groups?
These statistics mask the real conception and abortion rates because (unfortunately) the pill and more than likely the implant are in themselves abortifacients. And that’s without going into the stats about increased STIs
Err, no Ed. Neither oral contraceptives nor implants are abortifactants.
Biology 101, pregnancy begins with implantation not fertilisation.
It really comes down to what you mean by “rate”.
If by “rate” you mean “the actual number of abortions” then clearly for the over 40s that number has not gone down. However, if by “rate” you mean “the proportion of over 40s’ conceptions which are aborted” then clearly that rate has declined significantly. This is probably the (common) understanding of “rate” that the Daily Mail is referring to.
In fact this pattern of a declining abortion rate (rate as proportion of all recorded conceptions) appears to be declining in all the over 30 age groups. The under 30s show static or slightly increased rates.
I have some sympathy with Ed’s critique below (the figures don’t indicate the effect of abortifacients) but unfortunately there’s very little way to quantify that effect.