Teenage Pregnancy: Fact or Fiction?

One of the most persistently popular articles I’ve published over the last year or so was a compendium of statistical information relating to rape, which I put together last November in an effort to put some reasonably reliable figures under this very important and, at times, highly contentious debate.

With the issue of teenage sexuality featuring heavily in the news as a consequence of, amongst other thing, the Bailey Review and Nadine Dorries’s ‘Back to the 50s’ campaigns on the teaching of sexual abstinence and abortion, I think its time for another statistical compendium, this time looking at a range of statistics relating to conception, abortion and STI rates amongst teenagers. All the data here relates either to England or to England and Wales – Scotland produces and publishes its own public health statistics which, unfortunately, are often rather less accessible than those published by the Department of Health and the Office of National Statistics based on data from England/Wales.

In terms of format, I’ll be adopting a myth-busting approach to at least some of the data as, to be frank, a lot of nonsense is written about teenage sexuality and sexual behaviour with simply isn’t supported by the actual evidence.

So, let’s get started by tackling the biggest myth of them all, as expressed in this Daily Mail headline from last February.

We’ll never end our teenage pregnancy epidemic until we admit what’s REALLY causing it

The article itself (link via IstyOsty) which is attributed to Brenda Almond, Emeritus (i.e. retired) Professor of Moral and Social Philosophy at Hull University, takes the Peter Hitchins line of blaming sex education for teenage pregnancy and provides an prime example of the reason why aging, reactionary. moral philosophers should steer well clear of the social sciences:

A staggering 40,000 — or 40 per 1,000 — under-18s still fall pregnant in Britain each year, a pitiful improvement on 1998, when the figure was 46.6 per 1,000…

Historically, of course, this is an obvious truth. Before sex education became a central part of the national curriculum in the modern era, rates of sexual activity were far lower among young people and teenage pregnancy was exceedingly rare.

And now, the facts…

Fact No 1. The National Curriculum was introduced into England, Wales and Northern Ireland by the Education Reform Act 1988.

Fact No 2. Only the science component of Sex and Relationships Education, which deals with anatomy, puberty, biological aspects of sexual reproduction and use of hormones to control and promote fertility, is actually a compulsory element of the National Curriculum. All the other important stuff, which covers relationships, social and moral attitudes and values, risk reduction, etc. is optional both for the school and for young people who can be withdrawn from SRE lessons by their parents.

Fact No 3. The conception rate for teenagers, i.e. under 20s, did not alter significantly between the mid 1960s and the early-mid 2000s. The birth rate for under 20s in 1966 was 47.8 births per 1,000 women under 20. In 2004, the combined birth and abortion rates for women under 20 came out at 50.5 per 1,000. Allowing for miscarriages and stillbirths, where the rates for the latter have declined since the 1960s as medicine and medical technology has improved, the conception rates for women under 20 were much the same in 2004 as they were in 1966. What did change significantly over that period was the birth rate, which had more of less halved by 2004 as a result, primarily, of access to legal abortion, and the birth rates for married women under 20 and, in particular, for children born to women under 20 within 8 months of marriage.

During the 1960s and early 70s, no one really cared very much about the teenage pregnancy rate, just so long as pregnant teenagers were either married or shoved unceremoniously up to the registry office before the child was born. It was only after the legalisation of abortion effectively killed-off the shotgun wedding that teenage pregnancies, generally, became a matter of political – and often hypocritical – concern.

Fact No 4. The big rise in births – and, by extension, conceptions – to under 20s occurred between 1951-52, when there were only a little over 29,000 live births to mothers under the age of 20, and 1966, when the number of live births peaked at 86,528, with the most rapid increase taking place from 1956 onwards.

Fact No 5. The actual relationship between teenage conceptions as the introduction of sex education as a National Curriculum subject can be seen in this graph, which plots the actual trends in conception, abortion and birth rates for 13-15 year olds between 1955 and 2009. Birth and abortion rates/trends  are for conceptions occurring before teenagers’ 16th birthday.

The blue trend line shows the conception rate trend, the green line, which starts in 1968, shows the abortion rate and the red line shows the birth rate, with the yellow area indicating the period during which the National Curriculum was – and still is – in effect. The y-axis scale runs from 0 to a rate of 10 per 1,000 women aged 13-15.

We are looking here at a period during which its estimated that the number of young women whose first experience of vaginal intercourse predated their 16th birthday increased from 5%, in 1964 to 38% in 2008, an increase of 660%. Over that same period, the conception rate increased from just under 3 per 1,000 to its current, declining, rate of around 7.8 per 1,000, and increase of around 160%, while the birth rate has been kept in check by the legalisation of abortion in 1967 and has never once exceeded 2 per 1,000.

The introduction of the National Curriculum and, for the first time, a set sex education curriculum in the sciences coincides with a marked decline in the rate of increase of underage conception during the 1990s leading to an actual decline in conception, birth and abortion rates over the last 10 years. The improvements, to date, are modest but – at least until the last General Election – things were moving in the right direction.

We can see the post National Curriculum trend a little more clearly in this graph, which looks at the indexed trend in teenage conception rates since 1990 (1990 = 100) by age group (under 16, 16-17 & 18-19).

If you click the graph, you’ll get a larger image in which you’ll be able to clearly see the annotations, however, the first thing we can say for certain is that although there are a few bumps in the road, the overall trend since 1990 is downwards, particularly for under 18s.

Moreover, the annotations I’ve added provide a few clues to understanding some of the key movements in these trend lines. The graph starts in 1990, and falls fairly steeply in all age groups for the next three years, a period during with the ongoing HIV/AIDS public health campaign continued to receive significant government support. That more or less ended in 1993 with the launch of John Major’s ‘Back to Basics’ campaign, which shifted the emphasis in schools from providing high quality public health information to providing low-rent moralising and, within a year, the conception rate in the 13-15 age group rose by 7%.

‘Back to Basics’ squandered many of the gains made by the HIV/AIDS campaign of the late 80s and early 90s and it followed, in 1995, by a health scare related to the use of oral contraceptives, putting everything pretty much back to square one.

In 1998, the new (and New) Labour government introduced a teenage pregnancy strategy with an unrealistic target of cutting conception rates for under 18s by half by 2010 but an otherwise welcome injection of money and resources into providing high quality public health-based sex and relationship education – which includes the promotion of abstinence, contrary to anything Nadine Dorries might claim – and, by 2009. conception rates had fallen in both under 18 age groups by 20%.

This brings us to our next critically important fact…

Fact No 6. Teenage conception rates in the UK are not rising, they’ve actually been falling over the last decade and the conception rate for under 16s is now the lowest it been for 30 years.

While its true that the UK has the highest conception rate in Western Europe, comparisons to other European countries are far from straightforward as there is not one single European trend or pattern for teenage conceptions. Rather its the case, as what was previously the Health Education Authority (now part of NICE) noted in this report, that there are five very distinct international groupings within Europe, each of which has its own distinct pattern/trend for teenage conceptions. Trends in the UK over the last 40-50 years most closely resemble those in Germany, France and (before it banned almost abortions) Poland although by far the best international comparators for the UK are the United States, despite its teenage conception rates being double those of the UK, and the ‘Old Commonwealth’ countries, Canada, Australia and New Zealand, all of which have very similar conception, birth and abortion rates to the UK.

To illustrate just a couple of the problems that arise when making comparison to Europe, historically, the Catholic countries of Southern Europe had much higher teenage conception and birth rates than the UK in line with the once-common practice of young women marrying earlier in life in much large numbers than the UK. Scandinavian countries, such as Sweden, might have had a lower overall birth and abortion rate than the UK but the ratio of abortions to conception was considerably higher than the UK – over 65% in Sweden and Denmark in the mid 1990s compared to 40% in the UK at the same time.

Having touched on abortion, its worth throwing another interesting graph in to the mix, one that plots the relationship between conception and abortion rates, and the ratio of abortions to conceptions, for local authority areas in England against the deprivation indices for those same areas.

This particular graph shows the relationship to area-based education deprivation. I actually ran a complete series of graphs for a range of metrics including income, employment and health deprivation and the main multiple deprivation index and the all look pretty much same, differing only the strength of the correlations indicated by the three linear trend lines – blue is conception rates, red is abortion rates and green the ratio of abortions to conceptions with the most deprived areas appearing at the left-hand side of the scatter plot.

From both the conception and abortion rates, the trend is pretty much what most people would expect. Both are highest in the poorest areas of England and lowest in the most well-off areas. What many people might not expect is the green trend line which shows, quite clearly, that amongst teenagers who do fall pregnant, its those living in the most well-off and economically advantaged areas who are most likely to have an abortion. In some of the most well-off areas of South-East England, a pregnant teenager is twice as likely an abortion as a similar teenager living in one of the most deprived areas of the North-East or an inner-city area outside London.

What’s most interesting here is the strength of the correlations. Teenage conceptions rates are very strongly correlated to multiple deprivation (0.842) and to all the other deprivation metrics (0.7-0.83) and the correlation to abortion rates is somewhat more variable, particularly in relation to income. But when we come to look at the ratio of abortions to conceptions, by far the strongest correlation lies between educational disadvantage (or rather, advantage) and the percentage of conceptions ending in abortion (0.67) with employment deprivation – or lack thereof – the next highest while the correlation to the income metric, and to multiple deprivation, is only a moderate one (0.38 and 0.44 respectively).

This, of course, makes perfect sense when you consider that for many pregnant teenagers the issue that will most heavily influence their decision as to whether or not to have an abortion or continue their pregnancy to term will be that of the impact of impending motherhood on their ongoing education and, beyond that, on their future employment prospects. So, this bring us to out next important fact.

Fact No. 7 Although conception and abortion rates are typically highest in the poorest areas of England, it is relatively well-off, middle class young women who, because they see a university education and a career ahead of them, who are most likely to choose to have an abortion if they fall pregnant. If, on the other hand, you live in a derprived inner-city area and what you see the future holding out for you is a few pretty average GCSE and a job on the checkouts at Asda – if you have any employment prospects at all – then you’re twice as likely to give early motherhood a go as this will appear, if nothing else, to be the best of bum deal all around.

For any fans of the ironic, it should be noted that in Central Bedfordshire, the area which includes Nadine Dorries’s consitituency of Mid-Narnia, sorry Mid Bedfordshire, the most recent teenage conception figures (2007-2009) show that almost two-thirds of teenage conceptions (64%) result in an abortion.

And that bring us neatly to out final run of facts:

Fact No 8. There is no teenage pregnancy epidemic.

Fact No 9. High quality, comprehensive sex and relationships education is effective in reducing teenage conceptions, and would in all likelihood be much more effective if self-appointed moral campaigners and religious organisations would simply get out of the way and allow professional sex educators to get on with their job without continual interference.


Fact No 10. Nadine Dorries, Peter Hitchins and the like not only has no effective solutions to offer anyone – least of all a young women growing to adulthood in 21st Century Britain – but they are, in fact, part of the problem.

  • cim

    Why the use of polynomial fit lines for the graph in section 5? I think they’re giving a somewhat misleading impression about the very recent trend.

  • Anonymous

    Not really.

    Using second order polynomials provides a reasonable fit for illustrating the long term trends, particularly for conception rates where I had only a very limited amount of early data points to work with.

    For the birth rates the variation is arguably small enough to have warrented a linear fit line, while for the abortion rate I could perhaps have made use of a higher order polynomial to get a slightly more accurate picture of the rise in the abortion rate over the first 10 years or so after legalisation.

    So far as the most recent trend is concerned, this is covered in detail by the follow-graph of conception rates since 1990, which clarifies the recent trend and most importantly, I think, clearly illustrates the sensitivity of the trend to the kind of public health information that’s kicking around at any given time.

    A while ago, I put up a post at LibCon which looked in some detail at media influence on the trends in conceotion and abortion rates over time.In that graph, which got lost after Sunny did some mucking about with the back-end of the site, there are some very obvious trend movement linked to high profile media activity.

    Since the late 70’s there have been three periods during which there has been a marked drop in conception rates which was was sustained over a minimum of a two year period, each of which coincides with a major spluge of publcity relating to a ‘new’ STI. The first of these, in the early 80s, coincides with/follows the media’s focus on the herpes virus, the second – and largest – with the HIV/AIDs campaign and the msot recent at the beginning of the 21st century which links to both the govenrment’s investment in its teengage pregnancy strategy and the publixity afforded to chlamydia which followed the introduction of a new test in the late 1990’s which help to expose just how widespread it was.

    On the other side of the coin, there are similar but opposing movements in these trends linked to media coverage of health scares relating to the use of, particularly, oral contraceptives and, on the abortion trend, some evidence of smaller movements linked to periods of heighten activity and publicity by anti-abortion campaigners. On the latter, the movements are small – a bit spluge of anti-abortion material in the press seems to influence a few hundred women to choose not to have an abortion, but these are visible because they show up as periods during which the abortion trend runs counter to the overall trend in conceptions, i.e. years in which conception rates rise but abortion rates remains static or show a slight fall.

    What there is here is sufficient to counter the big myths of the supposed teenage pregnancy epidemic and, in particular, the absurd claim that this is linked to the inclusion of sex education in the National Curriculum when, in fact, the rising trend in conception rates, etc. has slowed and, more recently reversed, since the introduction of the National Curriculum at the end of the 1980s.

    Cut me a little slack here – its a blog post, not a peer-reviewed research study, where I would of course be much more exacting in my visual presentation of the statistical evidence.

  • Pingback: Flake()

  • cim

    I think the problem with polynomials is that while they may provide a better fit for the majority of the points in the middle of the trend, the way the line is trending at the end of the data ranges is massively exaggerated. If you don’t have a reason to believe that the data is really following a polynomial trend – which it isn’t here – then a moving average is probably far better for picking up multiple-region trends. (Which the detailed graph, I agree, does a good job of showing)

    So for instance the birth rate is basically flat across most of the time period, but the quadratic fit makes it look like it’s falling recently. Conversely the conception rate curve is quite heavily skewed by the inclusion of the 1964 value – I’m not sure it makes sense to include the pre-90s conception rate values in the curve at all.

    (I don’t actually believe that 1964 conception rate value is true, either, since it’s such an outlier from the rest and implausibly close to the birth rate)

  • Pingback: Marianne()

  • Pingback: David McMillan()

  • Pingback: Martha Lawton()

  • Pingback: Veronika Thiel()

  • Pingback: CF()

  • Pingback: Prym face()

  • Pingback: Education Technology()

  • Pingback: Carly blm()

  • Pingback: Education Technology()