Abortion statistics
Nadine Dorries, Conservative MP for Mid Bedfordshire, says on her website1 that
“The figures released this morning by the DoH show that there is a 23 per cent increase in abortions in girls under 14.”
The figure for abortions for under-fourteens in 2006 and 2007 are 135 and 163 respectively (see Statistical Bulletin: Abortion Statistics, England and Wales 2006, table4a pp13 and Statistical Bulletin: Abortion Statistics, England and Wales 2007, table 4a pp13, respectively).
163 less 135 is 28. 28 over 135 is 0.21, and so it’s easy to see that the figure is 21%, not 23%, but that small difference is understandable if worked out on the back of an envelope, so I’m not going to make much of it.
Dorries goes on to say
“The figures today show that children are aborting babies, which is a stark indicator of how the government has presided over a continuous and steady increase in abortion rates in ever younger girls.”
Fortunately, statistics are available for 1991 to 2001 from the National Statistics website and from 2002 to 2007 from the Department of Health website. Statistics for abortions under fourteen years were not collected prior to 2002 (it would seem), but for under fifteen years were collected.
| Under 14 | Under 15 | |
| 1991 | 935 | |
| 1992 | 941 | |
| 1993 | 1015 | |
| 1994 | 1080 | |
| 1995 | 946 | |
| 1996 | 1098 | |
| 1997 | 1020 | |
| 1998 | 1103 | |
| 1999 | 1066 | |
| 2000 | 1048 | |
| 2001 | 1066 | |
| 2002 | 168 | 1075 |
| 2003 | 148 | 1171 |
| 2004 | 157 | 1034 |
| 2005 | 137 | 1083 |
| 2006 | 135 | 1042 |
| 2007 | 163 | 1171 |
There isn’t data for rate, and so it’s hard to do any discrete probability distributions but you can do a quick standard deviation2; I’ve put data that vary by more than one standard deviation below the mean in italics and data that vary by more than one standard deviation above the mean in bold.
Using a single year’s change to describe a trend is largely meaningless. Looking over the available data, the increase over time may not be statistically significant3. The trend is neither continuous nor steady and possibly not there at all. In other words, Dorries is scaremongering.
xD.
1 - It’s not a blog if you can’t comment on it or trackback to it.
2 - For 14, mean is 151.33 and ? = 12.45 while for 15 mean is 1052.59 and ? = 66.76
3 - My memory of stats isn’t very good, but a Student’s t-test suggests it’s insignificant at a 5% confidence level
UPDATE 1335: The BBC agrees with my figure of 21%.
UPDATE 1617: Unity weighs in over at the Ministry of Truth.
June 26th, 2008 at 12:10 am
A quick couple of updates. In footnote 2, the question marks should be sigmas, and thanks to Cath Elliott, who writes for CiF, for linking to this post.
xD.
June 26th, 2008 at 11:11 am
[...] order, but I wanted to quickly thank Cath Elliott, who writes for Comment is Free, for linking to my post on Nadine Dorries’ use of statistics1 from this post of hers on CiF, entitled ‘It takes two to make a teenage pregnancy‘. [...]
July 9th, 2008 at 1:48 pm
[...] is a misleading sentence. The key word is ‘now’; this is nothing new. I showed (using clever sums and everything!) that there is no statistically significant rise in the number of abortions by people under sixteen. [...]