The impact of the English Teenage Pregnancy Strategy: pregnancy and sexually transmitted infections, David Paton
The impact of the English Teenage Pregnancy Strategy: pregnancy and
sexually transmitted infections
David Paton
presented at: 
50 ANOS DE EDUCAÇÃO SEXUAL:
balanços e perspectivas
Lisbon 13th November 2004
1. Background 

There continues to be considerable disagreement amongst academics as the likely impact of many policies aimed at improving the sexual health of teenagers.  As I have discussed in my earlier paper, the existing body of research on, for example, the provision of family planning, access to abortion, sex education and abstinence education has provided conflicting results.  In 1999, the UK Government announced a major initiative aimed at reducing the rate of teenage pregnancies in England.  The focus of this Teenage Pregnancy Strategy was based on improved contraceptive services for young people, more and better sex education and measures to re-integrate young parents into society.  The strategy has included initiatives based both at national local levels.  These strategies have been accompanied by significant injections of new funding.  The Teenage Pregnancy Strategy has been supplemented by the National Strategy for Sexual Health and HIV (Department of Health, 2001).  This has the aim of stemming sharp increases in rates of sexually transmitted infections (STIs) amongst young people in recent years (Public Health Laboratory Service, 2002) and proposes similar policy solutions to those put forward by the Teenage Pregnancy Strategy. 

The policies adopted in England provide a social experiment which is likely to provide a wealth of information and data that will be useful to other countries also concerned about sexual health amongst young people.  In this paper, I present some evidence from the early years of the policy, focusing particularly on the impact of improved contraceptive services for young people on rates of teenage pregnancy and STIs. 

2. Family planning & Teenage Pregnancy Policy in England 

2.1 Early Years 

Family planning provision for young people was almost non-existent in the UK until Helen Brook opened a centre aimed directly at young unmarried people in 1964.  During the late 1960's, the Family Planning Association also began to provide some services for young people.  However, the proportion of clients below the age of 16 was extremely low until the early seventies (Leathard, 1980).  In 1974, the then Department of Health and Social Security (DHSS) issued guidelines advising that contraceptive advice could be given to girls under the age of 16 without parental involvement, advice that was reissued in 1980.  Since the mid-seventies, teenagers in England have been able to access contraceptive advice and services free of charge from a network of family planning clinics since the mid-seventies. 

In Figure 1 I show the pattern of attendance at family planning clinics for under-16s and 16-19 year olds since the 1970s.  From this can be seen that the rate of attendance by under-16s in England at these clinics increased from 7.5 per 1000 in 1975 to more than twice that figure in 1984. 

During the 1980s, family planning clinics suffered from funding cuts associated with the Conservative Government and this lead to a modest decline in attendance rates at least for older teenagers.  In late 1984, Victoria Gillick won a court ruling that family planning could not be supplied to girls aged under-16 (the age of consent) in England and Wales.  Following this, attendances by under-16s decreased by over 30%.  The ruling was overturned towards the end of 1985.  Family planning attendance by under-16s recovered to above its previous level by 1988.  From 1989, teenagers were allowed to access emergency birth control (morning after pill) from some family planning clinics again without parental consent in most cases. 

Since the late 1980s, family planning attendances by young people have continued on an upward trend.  Latest figures suggest that over 17% of 15 year olds and 25% of 16-17 year olds attend family planning clinics at least once in a year.  This figure excludes those who obtain family planning from other sources, e.g. family doctors, condom machines etc. 

2.2 The 1999 Teenage Pregnancy Strategy 

Despite the family planning measures, teenage pregnancy rates in the UK remained the highest in Western Europe.  In 1992, the Conservative Government produced a report titled “The Health of the Nation” in which they pledged to cut conception rates amongst under-16s by 50% by the year 2000.  At the forefront of their strategy for doing this was an expansion of family planning services for young people.  When the Labour Government under Tony Blair was elected in 1997, teenage pregnancy rates were as high as ever and this prompted the Government to instigate a major inquiry into the issue.  The report of this inquiry was published in April 1999 (Social Exclusion Unit, 1999).  In the report, it was argued that an important factor contributing to high teenage pregnancy rates in England was a lack of knowledge of and access to family planning services aimed specifically at young people.  In June 1999, the UK Government officially adopted the recommendations of the report by launching the Teenage Pregnancy Strategy for England.  Contained within this Strategy was a commitment to reducing under-18 conception rates in England by 50% by the year 2010, with an interim target of 15% by 2004, and to establish a downward trend in under-16 conception rates (Social Exclusion Unit, 1999).  In a parallel development, the Government has also adopted a National Strategy for Sexual Health and HIV (Department of Health, 2001) with the aim of stemming sharp increases in rates of STIs amongst young people in recent years, (Public Health Laboratory Service, 2002).[1] 

Two key policy initiatives to achieve these aims are relevant to this paper.  The first is the expansion of community based family planning services aimed specifically at adolescents, a policy that is common both to the Teenage Pregnancy Strategy and the National Strategy for Sexual Health.  Responsibility and funding for implementing this policy were devolved to local areas, and there is evidence that the rate of expansion has shown considerable regional variation (Wellings et al, 2002).  The second policy initiative was a nationwide shift in emphasis towards the provision of the morning after pill or emergency birth control (EBC).  Regulations that came in at the start of 2000 made it much easier to dispense the hormonal version of EBC without a doctor’s prescription at family planning clinics and other sources and there have been many initiatives to promote this form of birth control to young people.  Pharmacies were also permitted to supply EBC without a prescription from 2000.  However, this services was restricted to those over 16 and was subject to a fee, whereas provision at family planning clinics is available free of charge without age limit.  Many commentators and family planning providers felt that the fee was a major deterrent to teenagers accessing emergency birth control from pharmacies.  It was further felt important to offer this service to under-16s.  As a result, many local authorities have used part of their funding from the Teenage Pregnancy Strategy to set up schemes in which young people of any age can access EBC free of charge from pharmacists.  This service has also been provided in some schools.  

3. An Economic Framework for Analysing the Teenage Pregnancy Strategy 

In my earlier paper I used an economic approach to demonstrate why providing family planning services may have an ambiguous impact on teenage pregnancy rates and I briefly review this approach here.  Consider, for example, teenagers who choose between sexual activity and abstention based on their expected utility of each choice (Oetinger, 1999; Paton, 2002).  Those who choose sexual activity must further choose whether or not to use contraception to protect against pregnancy and/or STIs.  The expected utility of each choice is a function of the utility of sexual activity, of outcomes (for example, pregnancy or an STI), of the perceived probability of each outcome and of the relative costs of each choice. 

Now think about the impact of a policy that reduces the marginal cost of family planning for adolescents.  Those who choose sexual activity will be more likely to use some method of family planning and (to the extent that their method of choice is effective) will face a lower probability of pregnancy.  At the same time, for adolescents who would prefer not to get pregnant, the decrease in the probability of pregnancy will lead to an increase in the expected utility of sexual activity relative to abstinence.  As a consequence, we would expect some adolescents who would otherwise have chosen abstinence to participate in sexual activity (and others to choose more sexual activity more of the time).  Some of this group will get pregnant due to contraceptive failure or mis-use. 

Applying such a model to the English Teenage Pregnancy Strategy yields some interesting theoretical predictions.  Consider firstly the uniform decrease in the absolute costs of all methods of family planning implied by the recent expansion of youth-oriented family planning clinics in England.  As we have seen, sexual activity is expected to increase in response to greater access to family planning and the impact on pregnancy rates is ambiguous.  However, only some forms of family planning offer any protection against STIs.  As a result, whatever the impact of greater access to family planning is on pregnancy rates, the impact on STI rates is likely to be less beneficial. 

Secondly, consider the reduction in the cost of EBC relative to the price of other forms of family planning.  It is known that a large proportion of teenage pregnancies result from contraceptive failure (see, for example, Churchill et al., 2000).  EBC provides a post hoc intervention whereby pregnancy can still be averted even after contraceptive failure or non-use.  Under the economic model, the availability of EBC enables young people to reduce the risks of pregnancy even more than in the presence solely of other methods and, thus, will be predicted to lead to an increase in rates of sexual activity.  This effect may be reinforced if the knowledge that EBC is available weakens a woman’s bargaining power at the time when effective decisions over sexual activity are taken (Akerlof et al, 1996).  The overall impact on pregnancy rates is impossible to predict.  On the other hand, as EBC offers no protection at all from STIs, the relative reduction in its cost would be predicted to result in an increase in STI rates.  A clear empirical consequence of this discussion is that we would expect the relationship between family planning and STIs to have worsened from 2000 relative to the relationship between family planning and pregnancy rates.[2] 

In fact, it is notable that the impact of access to emergency birth control has to date received no specific attention in the economic literature.  To my knowledge, the only relevant research is that of Churchill et al. (2000) who find that adolescents prescribed with EBC were more likely than others subsequently to be referred for abortion and that of Gold et al (1994) who find that teenagers provided with advance supplies of EBC were no more likely to engage in risky sexual behaviour than a control group.  Similarly there is very little evidence on the determinants of STI rates, the main exception being the 2004 paper by Klick and Stratmann.  The main finding of this paper is that the incidences of gonorrhea and syphilis in the USA are positively and significantly correlated with abortion legalization (see also Singh, 2003).  The authors also include the earliest legal age at which contraceptive services can be obtained without parental consent as a control variable.  For some specifications, they find that a younger legal age is associated with higher rates of gonorrhea.  Note, though, that Klick and Stratmann (2003) use data on STI infections to people of all ages and, to date, there is no work at all that focuses on infections amongst teenagers. 

4. Evidence from the Teenage Pregnancy Strategy 

In Table 1, I report national data on conception rates, STI rates and the number of youth-oriented family planning clinic sessions for each year between 1998 and 2002.

The direct effect of the policy of increasing clinic-based family planning services for young people is clear.  Between the start of the teenage pregnancy strategy in 1999 and 2002, the number of clinic sessions for young people rose by some 37.9%, whereas the number of GPs offering services to any person rose by just 2.6%.   Conception rates amongst under-18s fell from significantly between 1999 to 2001 and this fall was used by the Government as evidence that the Teenage Pregnancy Strategy was working.  Unfortunately in 2002, pregnancy rates amongst under-18s rose slightly.  Further, inspection of Figure 2 which shows longer running trends in conception rates amongst under-16s shows that the decline in pregnancy rates began in 1996 for under-16s and in 1998 for under-18s, both prior to the start of the Strategy.  Between 1999 and 2002, STI diagnosis rates have by about 25% for both under-16s and 16 to 19 year olds.  The increase in STI diagnoses pre-dates the Teenage Pregnancy Strategy (see Figure 2), but an interesting feature of recent years is that whereas before 1999, trends in STI followed trends in pregnancy rates very closes, from 1999, the trends diverge significantly.  This is in line with the predictions of the economic model discussed above. 

Clearly, however, a range of longer term trends and influences may have affected these national figures and this makes it very difficult to draw inferences about causal relationships.  In a paper presented to the 2004 Royal Economic Society Conference (Paton, 2004), I performed a detailed panel data regression analysis, relating teenage pregnancy and STI rates amongst 99 health authorities between 1998 and 2001 to a range of variables including provision of family planning clinics, GP contraception, unemployment, educational achievement and children in care. 

Formally, I estimated two models.  Firstly, to test for the overall impact of family planning access on pregnancy and STI rates, I estimated the following model: 

pregnancyit = a0 + a1 FPit + gxit + hi + nt + mit                             (1a)
                        STIit = b0 + b1 FPit + dxit + hi + nt + wit                                  (1b)
where FP = some measure of access to family planning;

            x = vector of other variables likely to affect pregnancy and STI rates;
            h = region-specific effects;
            n = time-specific effects;
            m and w are classical disturbance terms.
Secondly, in order to test whether the impact of family planning access has changed with the shift in emphasis since 2000 towards emergency birth control, I also estimate the following variant of the econometric model: 

         pregnancyit = a0 + a1FPit + a2FP*1999 + a3FP*2000 + a4FP*2001 + gxit + hi + nt + mit    (2a)
         STIit = b0 + b1 FPit + b2FP*1999 + b3FP*2000 + b4FP*2001 + dxit + hi + nt + wit             (2b)
where FP*1999 is an interaction term between FP and a dummy variable for the year 1999. 

In this specification, a2 is the differential impact of family planning on teenage pregnancy in 1999 compared to 1998, a3 is the differential impact for 2000, whilst a4 is the differential impact for 2001.  The coefficients b2 - b4 can be interpreted similarly.  If a3 and a4 are significantly negative, this would imply that the relationship between family planning and teenage pregnancy has improved since the adoption of the teenage pregnancy strategy.  b3 and b4 can be interpreted similarly. 

Paton (2004) discusses several methodological issues.  The first one is that of correctly identifying the family planning impact.  Family planning services are more likely to be set up in areas where pregnancy rates (and perhaps STI rates) are high.  Thus, we may observe a spurious positive correlation between the family planning and teenage pregnancies.  Put another way, unobservable high rates of sexual activity in an area due, for example, to socioeconomic factors, are likely to lead to a high demand for family planning services as well as high pregnancy and STI rates.  This problem is alleviated in Paton (2004) by the use of a 2-way fixed effects estimator, sometimes called the difference-in-differences approach.  Essentially this controls for different rates of pregnancy (or STIs) in each area and means that the coefficients are identified by changes in time.   

One disadvantage of this approach is that if there is insufficient variation in a variable over time, then it will be difficult to identify coefficients.  Put another way, with the difference-in-differences estimator, there is a high chance of a Type 2 error, unless there is significant variation across regions in changes in variables over time.  This is unlikely to be a serious problem for the key family planning variable used here (free clinic sessions for young people) due to the significant variation over time caused by the major policy shift in England.  In the paper, there are also controls for the possibility that teenagers resident in a particular region may use family planning services in an adjacent area and for a number of other statistical issues. 

The results from this analysis are very clear.  Just as earlier research has found, access to family planning is not associated with lower pregnancy rates.  On the other hand, family planning clinic sessions for young people are significantly and positively associated with STI diagnoses rates.  Put another way, areas that have increased the number of family planning clinic sessions for young people the most have not seen bigger decreases in pregnancy rates than other areas, but they have experienced the biggest increases in STI diagnoses.  Further, the (adverse) impact of family planning on STIs seems to be significantly higher after 2000 than before, in line with the predictions of the economic model that increasing use of emergency birth control would worsen the relationship between the two variables. 

One issue with the English STI data is that information is only collected on diagnoses at specialist Genitourinary Medicine (GUM) clinics.  No information is available on the true rate of infections.  This is particularly a problem for diseases such as chlamydia which is largely asymptomatic.  Thus, one explanation of the results in Paton (2004) is that family planning clinics are raising awareness of chlamydia and encouraging young people to go to GUM clinics, hence increasing diagnoses.  In fact the significant relationship between family planning clinics holds even when only diagnoses of chlamydia are excluded from the data.  Thus, this explanation does not seem sufficient to explain the results.  Weaknesses in the data, together with the lack of much previous research on the impact of family planning on STIs suggest that these results be treated with some caution.  They do, however, raise concerns that measures used by the UK Government (and in many other countries) to lower teenage pregnancy rates may in fact be having an adverse impact on sexual health of young people.  

5. Conclusions 

In this paper I have shown how the 1999 Teenage Pregnancy Strategy in England provides a useful natural experiment for testing predictions regarding the impact of access to family planning services on teenage pregnancy and STI rates.  Although teenage pregnancy rates in England have decreased in recent years, the trend downwards pre-dated the start of the Pregnancy Strategy and rates continue to be very high compared to other countries in Western Europe.  Indeed, the most recent evidence is that the rate has stabilised.  In contrast, rates of STI diagnoses have continued to increase dramatically amongst young people over the course of the Strategy.  

Looking more closely at regional variations in England, the evidence suggests that those recent increases in the number of youth family planning clinic sessions are not associated with reductions in teenage pregnancy rates, but are associated with higher rates of diagnoses of STIs amongst teenagers.  Further there is evidence that the link between family planning and STI rates has worsened significantly since 1999 with the most likely reason being the shift towards greater promotion of emergency birth control.   

The experience English Teenage Pregnancy Strategy is a salutary one and demonstrates an important lesson for policy makers.  It is not enough to introduce a measure aimed at a specific outcome without considering the endogenous response of agents to the policy itself.  In the case in question, there is at least some evidence that some measures aimed at reducing adolescent pregnancy rates induced changes in teenage behaviour that were large enough not only to negate the intended impact on pregnancy rates but also to have a possible adverse impact on another important area of adolescent sexual health - sexually transmitted infections.  Above all, the UK experience should cause policy makers in other countries to think carefully before emulating the strategies on family planning and emergency birth control that appear not to have been successful in England. 

References 

Adams, John G U (1994), ‘Seat Belt Legislation: the evidence revisited’, Safety Science, 18: 135-52.
Akerlof, George, A, Yellen, Janet, L and Katz, Michael, L. (1996), ‘An Analysis of Out-of-Wedlock Childbearing in the United States’, Quarterly Journal of Economics, 111 (2, May): 277-317.
Department of Health (2001), The National Strategy for Sexual Health and HIV, July, London: Department of Health.
Gold, Melanie, et al. (2004), ‘The Effects of Advance Provision of Emergency Contraception on Adolescent Women's Sexual and Contraceptive Behaviors’. Journal of Pediatric and Adolescent Gynecology, 17, 87-96.
Klick, J. and Stratmann, T. (2004), ‘The effect of abortion legalization on sexual behavior: evidence from sexually transmitted diseases’, Journal of Legal Studies.
Oetinger, Gerald S. (1999), ‘The Effects of Sex Education on Teen Sexual Activity and Teen Pregnancy’, Journal of Political Economy, 107 (3): 606-44.
Paton, David (2002), ‘The Economics of Abortion, Family Planning and Underage Conceptions’, Journal of Health Economics, 21 (2, March): 27-45
Paton, David (2004), ‘Random Behaviour or Rational Choice?  Family Planning, Teenage Pregnancy and STIs’, Royal Economic Society Annual Conference, Swansea, April, 2004.
Peltzman, Sam (1975), ‘The Effects of Automobile Safety Regulation’, Journal of Political Economy, 83 (August): 677-723.
Public Health Laboratory Service (2002).  Trends in Selected STIs: 1991-2001, London: PHLS.
Richens, John, Imrie, John and Copas, Andrew (2000), ‘Condoms and Seat Belts: the parallels and the lessons’, Lancet, 355 (9201, January): 400-403.
Singh, B (2003), ‘A Preliminary Investigation of The Effects of Restrictions on Medicaid Funding for Abortions on Female STD Rates’, Health Economics, 12(6), 453-464.
Social Exclusion Unit (1999), Teenage Pregnancies, London: HMSO CM4342.
Wellings K., Wilkinson P., Grundy C., Kane R., Jacklin P., Stevens M. and Gerressu M. (2002), An audit of contraceptive service provision in England, with special reference to service provision for young people. London: Teenage Pregnancy Unit.
 
 
 
 
Table 1: Trends in Family Planning, Sexual Activity, Teenage Pregnancy & STI rates

 

Year

Variable

1998

1999

2000

2001

2002

Adolescent family planning clinic sessions

28,115

27,075

30,710

34,192

37,329

GP contraception provision

23,547

23,873

24,065

24,299

24,466

Conception rates:

 

 

 

 

 

all teenagers

65.43

62.94

62.45

60.71

n/a

16 – 19

74.84

72.33

72.22

70.31

n/a

under-18

47.03

45.35

43.86

42.48

42.52

under-16

8.88

8.19

8.28

7.94

n/a

STI rates:

 

 

 

 

 

all teenagers

93.1

103.0

114.6

119.3

127.9

16-19

110.3

122.2

137.5

142.5

152.2

under-16

7.76

8.70

9.15

10.13

10.91

Source: Adapted from Paton (2004) 

Notes

(i) Adolescent family planning clinic session numbers include estimates for Brook clinics, as described in the Appendix.

(ii) GP contraception provision is the number of GPs offering a contraceptive service to any patient, not just to those patients on their list.

(iv) Conception rates for all teenagers, 16-19, under-18 and under-16 are per 1000 of the 15-19, 16-19, 15-17 and 13-15 female population respectively.  STI rates for all teenagers, 16-19 and under-16 are per 10,000 of 15-19, 16-19 and 13-15 total population respectively.

[1] Part of the increase may be due to greater awareness and diagnosis of STIs.  However, the PHLS attribute a significant proportion of the increase to a rise in risky sexual behaviour amongst young people.

[2] There is also a complementary and extensive literature on risk displacement in the context of automobile safety.  Authors such as Peltzman (1975) and Adams (1994) argue that the beneficial effects of a technical improvement in road safety may be offset by an increase in dangerous driving.  Richens et al. (2000) explicitly link these ideas to the issue of condoms and protection against STIs.
 

ASSOCIAÇÃO FAMÍLIA E SOCIEDADE
Horário: 9h30 - 13h30
Rua do Lumiar 78 / 1750-164 Lisboa
Tel/Fax: 21 314 95 85 - familiasociedade@sapo.pt