The Economics of Secret Abortions and Emergency Birth Control
Professor David Paton FAITH Magazine July-Aug 2007
Economics has had something of a bad name among pro-lifers recently. This is largely due to the efforts of the American economist Steven D. Levitt, who claimed, in Freakonomics, that the legalisation of abortion has led to a reduction in American crime rates. The reader might be surprised to discover that I believe his research was excellent and deserves to be welcomed by the pro-life movement – and this is why.
Most economists are trained at an early stage to have a professional disregard for what is right or wrong, and focus, rather, on the facts. Levitt’s hypothesis was not concerned with the moral issues surrounding crime and abortion but with whether in fact there is a link between them, and this is where the pro-life movement should keep an open mind. Their openness will be amply rewarded since research of this kind – from the viewpoint of social science – tends to support the pro-life cause. On that basis, my intention is not to discuss the rights or wrongs of abortion, but to examine the research evidence on issues such as secret abortions for teenagers, schools offering the morning-after pill, etc.
Let us begin with the ubiquitous word confidentiality. Whether it be used in clinics, pharmacies, youth groups, by GPs or school nurses, there is no avoiding it. The policies that are routinely sold to our children rest on the contention that all people have the right to confidentiality and therefore we must have confidential access to sexual health services, to abortion and so on. One might associate the word with another – secret – but ‘confidential’ is a far easier word to sell, to patients, to parents and to children. Yet when we think about what it really means, that we are providing these services without parental knowledge to children as young as twelve, then perhaps the word secret is more appropriate. And the question is, Why? Why are we being besieged by policiesthat provide abortion free of charge without parental knowledge or the morning-after pill at pharmacies and schools?
The argument is perfectly simple and very easy to sell. The Government has said it wants to cut teenage pregnancies, and aims to do so by fifty per cent by the year 2010. Obviously, so the argument goes, if we want to cut teenage pregnancies and abortions we must have access to sexual health services – in other words, teenagers are less likely to get pregnant if they are using contraception; failing contraception, then we should give them access to the morning-after pill, which may be seen as preferable to a twelve-year-old getting pregnant. Whether we like it or not, this argument is very attractive to parents, even to those who are instinctively against abortion and the morning-after pill.
The next stage of the argument is that, in order to maximize the use of these services, confidentiality must be guaranteed. A twelve-year-old will naturally be afraid of parental backlash, so while we would rather encourage her to tell her parents she has got the morning-after pill, the decision is ultimately hers and she has the right to keep it a secret. Likewise with abortion. Without that guarantee, youngsters are less likely to access “services” and it is argued that this will result in more pregnancies at an early age.
For a lot of people, this argument is compelling. Many parents are genuinely worried: certainly they would rather know when their young people are involved in sexual activity and accessing the morning-after pill but, if this is what it takes to avoid their thirteen year-old daughter getting pregnant, then so be it. I believe this attitude is why many parents go against their instincts and accept this sort of policy. The question that now lies before us is: do these policies actually work?
Let us step back a little and look at the question in context. Figure 1 shows a summary of the teenage pregnancy strategy introduced in England and Wales in 1999. It shows what needs to happen for the Government to meet its interim target of reducing pregnancy rates by fifteen per cent by the year 2004 and the final target of reducing rates by fifty per cent by 2010. There is no specific target for under sixteens.
As the graph illustrates, the Government has already failed to achieve its interim target for 2004. Teenage pregnancy rates have in fact fallen by ten per cent from the Government’s base year of 1998. However, 1998 was a peak year for teenage pregnancies and, as the graph shows, pregnancy rates had started to fall before the Government announced its Teenage Pregnancy Strategy during 1999 (indicated by the vertical line on the graph). Most of the funding only started to filter through to the local authorities from about 2000 onwards. If the Strategy was working, we would expect to find the downward trend in pregnancy rates (which started before the Strategy begun) accelerating in recent years. What we actually find, however, is that the graph flattens out, and, over the past few years,there has been very little change in the under-eighteen pregnancy rate. Hence, despite spending well over £100 million since 1999, there is very little evidence that the Government’s Teenage Pregnancy Strategy has caused any reduction in pregnancy rates at all.
As things stand, even the Government admits there is no hope of achieving its desired target for 2010 (indicated by the dotted line in Figure 1, which is why Hazel Blears, the Minister responsible for this at the time, announced in 2005 that the Government had done enough and that it was now up to parents to bring pregnancy rates down.
Figure 2 highlights the progress of sexually transmitted diseases among teenagers over the same period of time – an issue which is seldom mentioned by the Teenage Pregnancy Unit in all its talk of pregnancy rates. Notice that only recorded infections are represented here, and not, for example, cases of Chlamydia where there are no symptoms. Nevertheless, thre is general agreement among those who work in the sexual health area that, even taking on board changes in recording and greater knowledge of the likes of Chlamydia, there has been an explosion of sexually transmitted infections, particularly among young people.
The economists’ approach Economists approach these issues of sex, abortion and family planning as they would everything else. We build up a simplified model of reality by making assumptions about how the world works, and then we try to assume everything away apart from that specific issue we are focusing on. Some may object to this approach, but in fact we all use simplified models of reality in every day life. For example, a road map is an extremely simplified version of reality, dependent on many assumptions. It ignores schools, pubs and trees; all it has is the key roads. What is important is that the map helps you get you from A to B. Whether or not the simplification from reality is justified depends on whether or not it succeeds in this goal. The same goes for an economic model. You make allthe assumptions necessary in order to make it simple, to highlight more clearly the area of interest.
Let us examine, then, the economic approach to the issue of secret abortion. What happens when you introduce a policy of confidential access to abortion services for minors or, alternatively, you introduce a policy asserting parents’ right to know? This is very important because, although the remit of the Teenage Pregnancy Unit is to cut pregnancy rates, one of its main policy recommendations that it emphasises year after year is to provide youngsters with easier and better access to confidential abortion.
An economist would typically model this issue just as he might model insurance problems. We can think of abortion acting as a type of insurance policy for teenage sexual activity in the same way that a car insurance policy covers theft etc. Consider a youngster deciding whether or not to start engaging in sexual activity with a boyfriend or girlfriend. Some of the factors in this decision-making process will be: What if I get pregnant? What if my mum finds out? How will it affect me going to university or to college, getting a job?
Now economists know that people don’t just respond passively to changes in policy. Rather, we tend to change our behaviour in response to the incentives with which we are faced. When we have an insurance policy that protects us against the risks of, for example, our car being stolen, we tend to be a little less careful about preventing the risks from happening. This phenomenon has been the subject of endless studies and is termed ‘moral hazard’ by economists. Insurance companies are well aware of the problem and that is why they try to take steps to force us to be careful. If you take out a higher excess on your insurance, you will generally get a lower rate.
Now how does this relate to abortion and confidentiality? If abortion is easily accessed, then it acts to alleviate some of the risks of having sex. For those youngsters who are not opposed in principle to abortion, it provides a way in which, if pregnancy occurs, birth can be avoided, i.e. if pregnancy occurs either through failed or non-use of contraception, there is a possible let out clause. This does not mean that all youngsters would start taking more sexual risks when abortion is easily available. Some youngsters would never dream of having an abortion in any case and would not be affected by any policy change. Others actively want to have a baby and will similarly be unaffected. But, at the margin, there are likely to be some youngsters who perceive abortion as alleviating some ofthe risks of having sex and/or having sex without contraception. When abortion is made easier to access, e.g. by assuring youngsters that their parents need never know, we would predict more youngsters to engage in risky sexual activity. Some of these youngsters will get pregnant when they would not have done otherwise and the overall teenage pregnancy (births and abortions combined) rate is likely to increase. It may be that, within the higher pregnancy rate, the proportion of abortions will increase and proportion of births decrease. However, economic models show that when abortion is easier, it is possible that even births may either not change or could even increase. This is because some youngsters who think they will opt for an abortion if they get pregnant, may not do so when theyare faced with the actual decision.
A key plank of the Teenage Pregnancy Strategy – indeed its main aim – is to cut teenage pregnancy rates. Yet if the economists are right then this aspect of the policy should lead to an increase in pregnancy rates.
Putting it into practice Ideally, of course, we would hold up this economic theory against the data. The best way of testing this (putting aside ethical concerns for a moment) would be to conduct an experiment in which one group did not have access to confidential abortion and another did, and then we could monitor the differences between them. Better still would be to offer confidential abortion to two similar groups of people, then remove confidentiality from one of those groups and then assess the relative change in pregnancy rates in the between the two groups.
Quite rightly, ethical demands prohibit such experimentation; but we may turn to the US which has run exactly that experiment for us. Over the past twenty years, individual States in America have been allowed to introduce Parental Notification or Parental Consent laws before minors are allowed abortions. Some States have taken up this option, whilst others have continued to operate a policy of assuring confidentiality. From the point of view of testing the economic theory, it has been a first-class experiment, since it has allowed researchers to control for pretty much every other factor, from nationwide trends in abortion to socio-economic effects.
The results are very clear. Many States in the US have introduced some sort of parental-involvement law since 1984. Some of these laws have not been enforced, but over thirty States have had these laws in place for an appreciable period of time.
To date, over sixteen studies have looked at this issue and the impact on abortion, birth or pregnancy rates, the best of which is probably a study by economist Philip Levine. His conclusion is very clear: that in those states with a parental-involvement law, abortion rates decreased on average by between fifteen to twenty per cent. At the same time birth rates remained fairly unchanged. Clearly, then, it was not simply a question of people declining abortion and opting for giving birth instead; rather, fewer people were getting pregnant. He estimated that parental involvement laws caused pregnancy rates amongst minors to go down by between four and nine per cent.
Virtually all of the other research papers back up this finding. They have used a range of research methods, some better than others, but none has found an increase in pregnancy rates. A couple of research papers indicated only a decrease in abortion rates that is statistically insignificant (i.e. it could have been due to chance rather than the change in the law), but these have tended to analyse data on abortion rates for all ages, not specifically teenage abortion rates. The research evidence on this question is hard to deny and right in line with the predictions of the economic models: prohibition of secret abortions cuts both teenage abortion rates and teenage pregnancy rates. Perhaps the most effective measure that British Government could use to help achieve its target of cutting teenage pregnancy rates is to introduce a rule giving prohibiting confidential abortions for minors, such that parents would have to be informed or, better still, would have to consent to their youngsters having an abortion. Yet, for some reason, the Government has chosen to do the exact opposite.
Parallels with contraception We might also look at the parallels with contraception – how does the availability of contraception at school without the requirement of parental consent impact on teenage abortion and pregnancy rates?
Following a similar argument to the above, we might expect two contrasting effects. Firstly, sexually active youngsters who were not using contraception would now be more likely to use contraception, and fewer would get pregnant. Although contraceptives have a very high failure rate amongst teenagers, on average, contraceptive sex is less likely to result in pregnancy than sex with no contraception at all.
Nevertheless it is also true that the ‘insurance policy’ argument of before still holds. Since sex is now less likely to result in a pregnancy, those who previously avoided sexual activity due to concerns about getting pregnant are now more likely to engage in sexual activity. Some of this group would still get pregnant, despite using contraceptives. Whether or not confidential family planning leads to lower teenage pregnancy rates depends on whether the first effect (which lowers pregnancy rates) is larger than the second effect (which increases pregnancy rates). The Government’s approach has been just to ignore the second effect and claims repeatedly that introducing contraception in schools will increase the use of contraception without increasing sexual activity and that pregnancyrates will certainly fall.
Given that the economic theory is inconclusive as to what the overall outcome will be, it is important to look at the empirical evidence. There has been far less research on confidential family planning than on secret abortions. To date, only two research papers in refereed academic journals have investigated the impact of confidentiality/no confidentiality on abortion or pregnancy rates. One paper, published by myself focused on the Gillick ruling in the UK and found very clearly that when, in 1985, Victoria Gillick managed to get a temporary ban imposed on confidential family planning for minors, pregnancy rates did not increase. In fact, closer examination of the evidence reveals that, if anything, pregnancy rates decreased in 1985relative to where they were and relative to Scotland where that policy was not introduced. The difference is not large enough to exclude chance, but we can fairly safely conclude that removing confidentiality from family planning services for minors did not lead to an increase in teenage pregnancy rates.
The second piece of research is from the US – the so-called ‘McHenry county’ experiment. The original paper, published in the American Journal of Public Health in 2004, suggested that the introduction of a parental consent for contraception ruling in McHenry County, Illinois led to more teenage births, fewer abortions and to a significant increase in pregnancies overall. In fact, an error in the data later came to light and, once this had been corrected (published in an erratum in a later issue of the Journal), it became clear that there was actually no significant change in teenage pregnancy rates.
Subsequent work carried out jointly by myself and Dr Zavodny examining a longer time period suggests that there was actually no significant impact either on abortions or births. In short, to date, no research has found that cuts in confidential access to family planning lead to a significant increase in either abortion or pregnancy rates. Parents who support confidential access on the belief that their daughters will be less likely to get pregnant are simply being misled.
Emergency birth control What about emergency birth control (the morning-after pill)? Apart from the fact that it may cause a very early abortion (which of course would not get picked up in the official teenage pregnancy figures), it differs from common methods of family planning in several ways. Most importantly, it can be used either after ‘unprotected’ sexual activity has taken place or after contraception has failed. For this reason, when the morning-after pill was introduced, it was hailed as a medical advance that would virtually end the problem of unwanted pregnancies. There are published papers in academic journals claiming that greater access to the morning-after pill would cut unwanted pregnancies by fifty per cent. I will now go on to explain what the actual impact of greater access toemergency birth control (EBC) upon teenage pregnancy has been.
The UK provides a very good case study. Since the end of 2000, EBC in England and Wales has been available over the counter at pharmacies, but only to people over the age of eighteen and at a fee. However, since the start of 2000, some local authorities have introduced schemes to provide EBC free of charge and confidentially to young people including those under the age of 16 at pharmacies. Other authorities have decided not to spend money on these schemes.
When these schemes were introduced there was virtually no research evidence demonstrating whether or not they were likely to work. The Government simply did not know what the outcome would be. In other words, it has been nothing less than a social experiment carried out on your children. What has been the result of this experiment?
Analysing the dataDr Sourafel Girma of Nottingham University and myself have carried out an extensive analysis of the impact of these schemes on teenage pregnancy rates. The results were published recently in Health Economics, a highly rated peer reviewed journal. We looked at every local authority in England, using what is called matching difference-in-difference estimation. This is useful technique that allows the researcher to isolate the impact of the EBC scheme from other possible confounding factors.
Let us assume we want to compare teenage pregnancy rates in those areas that have introduced the policy and those that have not. There are a whole range of other factors that may affect pregnancy rates, including unemployment, poverty, family stability and so on. For this reason, we examine changes in pregnancy rates before and after the policy was introduced relative to changes in areas that did not introduce the scheme. This comparison is called difference-in-differences.
For example, assume that the EBC scheme was introduced in Nottinghamshire in 2001 but was not introduced in, say, Warwickshire. What we want to know is not whether pregnancy rates are lower in Nottinghamshire, but whether they went down in Nottinghamshire more or less than they went down in Warwickshire.
The strength of this difference-in-difference approach is that, with enough areas in both categories, we can get a very powerful test of the average impact of the scheme. However, in the past, this type of test has been criticised on the grounds that those areas introducing the EBC pharmacy schemes may be fundamentally different to those areas that do not. For example, areas with the scheme might be those in with high levels of deprivation and where teenage pregnancy rates were increasing. We might not observe a decrease in pregnancy rates when the pharmacy scheme is introduced, but rates might still be lower than had the scheme not been introduced. To get around this problem, Dr Girma and I used the matched difference in difference estimator. This involves making sure that the areas ingroups with and without the schemes are similar in other respects unemployment, educational levels, ethnic mix, starting teenage pregnancy rates and so on. Once we do this, we have a very robust test indeed.
The paper concludes: ‘Irrespective of either the matching or the adjustment procedure, we are unable to find evidence that schemes allowing emergency birth control leads to reductions in teenage pregnancy rates’ – in other words, whichever way we looked at the data, there was no evidence that confidential pharmacy EBC schemes lead to reductions in teenage pregnancy rates. The experiment on your children has failed.
Our research is backed up by all the other academic papers which examine the impact of easier access to EBC on unwanted pregnancy or abortion rates. Many articles have been published on this topic, using a variety of methodologies and different data. Some are what we call randomised control trials in which two groups of people are selected: one group is given advance supplies of the morning-after pill, while the other is not. Others are observational studies which examine data in the whole population. Each article (published in reputable journals such as New England Journal of Medicine, the Journal of the American Medical Association, Contraception, Sex Education) comes to the same conclusion: easier access to EBC seems to have no significant impact on teenage or unwanted pregnancy orabortion rates.
The journal Sex Education published an earlier article of mine in which I found some evidence (though not conclusive since the data on STIs available to us is far inferior to that on pregnancy rates) that STI rates amongst teenagers have increased fastest in those areas promoting the EBC the most. Again, this would appear to be common sense. We know that pregnancy rates increase because availability of the morning-after pill makes people take greater risks in terms of riskier sexual activity; and we also know that the morning-after pill does not offer protection against STIs. Research on this point is by no means conclusive yet, but there some evidence which complements my findings. Two economists, Jon Klick and Thomas Stratmann found conclusive evidence (again published in a peerreviewed journal) that legalisation of abortion in the US contributed directly to an increase in sexually transmitted infections.
Using the facts Even Anna Glazier, a health expert and a strong proponent of greater access to the morning-after pill, stated in early 2006 in an editorial in the British Medical Journal that greater access to emergency birth control has failed to cut pregnancy and abortion rates. Despite this, she along with many other advisers to the Government, continue to advocate still wider access to emergency birth control. At this point you may ask “Do the facts matter?” In one sense evidently they do not! One interpretation of what has happened in the UK over the past few years is that certain groups were determined that the policies I have talked about here would be introduced irrespective of the facts. The Government tells us that we need to introduce confidential access to EBC to cut teenagepregnancy rates; they introduce the policy; we now discover that this policy does not work, yet there is no sign at all of the policy being withdrawn as a result. In fact, we continue to be bombarded by the same old slogans about young people’s rights to access confidential services. There is a retreat from the facts, but not from pushing the policy.
However, this is not good enough. Too many ordinary people on the street, who are frankly worried and concerned to do the best for their children, would not support these policies if they were aware of the evidence. It is absolutely crucial that people of good will examine the scientific evidence and disseminate the facts through to schools, local authorities and health trusts. Those groups trying to further this social experiment on your children must be challenged at every stage. If a health trust or local authority tries to use your tax money (money that could be spent elsewhere) on one of these policies, then demand to see the evidence that it will result in lower unwanted pregnancy rates. If they introduce the policy anyway, then don’t let up. As time goes on, demand to see theevidence on whether or not it has actually worked.
By way of summary, the graph in Figure 3 shows what has happened to family planning clinic attendance and take-up of EBC at these clinics for under sixteens in England since the early 1970s. The bottom of the graph indicates what has happened to conception (abortion and birth) rates to under 16s over this time.
Clearly family planning attendance has increased dramatically over this time, but we have to be careful in interpreting this. There have clearly been significant social changes, independent of Government policy, that have contributed to this rise. The thing to focus on is what happens to the under 16 conception rate when there is some sort of policy change that affects the supply of family planning. Some of these are marked on the graph. Take the Gillick ruling: the Government was forced to allow parents to know when their young people got access to family planning clinics, and we saw a decrease in family planning attendance for under sixteens. What happens to the under 16 conception rate at this point? Pretty much nothing at all.
In 1992, the Conservative Government introduced the Health of the Nation Report in 1992. The goal was to cut teenage pregnancies by fifty per cent by the year 2000 through greater access to family planning clinics, better access to sex education and earlier sex education in schools. What happened to the conception rate? Nothing much. What about when emergency birth control was first made available to under 16s at family planning clinics in the early 1990s? Again no obvious change in the conception rate. In 1999, the new Labour government introduced its own teenage pregnancy strategy – a brand new policy aimed at cutting teenage pregnancy rates by fifty per cent by the year 2010, through more access to family planning clinics, family planning in schools, better and earlier sex education.What has happened to the conception rate? Again not much at all. Perhaps everything is about to change between now and 2010 and when teenage pregnancy rates will suddenly plummet. History tells us that we should not be too optimistic.
On a more positive note, I have noticed recently that groups like the Teenage Pregnancy Unit are complaining about financial cuts. Generally speaking, local family planning clinics and morning-after pill-type schemes are among the first casualties in any financial cuts carried out by health authorities. Of course the special interest groups, represented by the Teenage Pregnancy Unit object. But money matters, particularly to health trusts and local authorities. When money gets tight, policymakers in a local area naturally look for the areas of expenditure that are proving to be least effective. So even when the political situation of changing hearts and minds in Government is very difficult, as the evidence becomes clearer that policies are not having an impact, money can be the decidingfactor. Every local authority has a teenage pregnancy advisor – ask them what money has been spent, how it has been spent, what evidence is there that these policies work and what is the progress in your area. All of us can and should be active in holding our local health trusts, PCTs and local authorities to account every year on how they are spending our money.
In conclusion, you may object to Steven Levitt’s suggestion that abortion leads to lower crime, but I hope the reader is convinced that there is some worth in the research carried out by economists. If you are prepared to examine evidence on these type of questions fairly and without favour, the answers you get are simply invaluable in holding your local representatives to account and, ultimately, getting policy changed.
This paper, written in 2001, features in Steven D Levitt and Stephen J Dubner, Freakonomics: A Rogue Economist Explores the Hidden Side of Everything. Ref?, William Morrow.
Levine, Phillip B (2003), ‘Parental involvement laws and fertility behavior’, Journal of Health Economics, 22 (5, Sept), 861-78.
Paton, David (2002), ‘The Economics of Abortion, Family Planning and Underage Conceptions’, Journal of Health Economics, 21 (2 March):27-45.
Madeline Zavodny & David Paton (2006), ‘Teenage Pregnancy Risk: the impact of parental involvement for contraception’, Occasional Papers 18, Industrial Economics Division., available at www.nottingham.ac.uk/~lizecon/RePEc/pdf/18.pdf.
Girma, Sourafel and David Paton (2006), ‘Matching Estimates of the Impact of Over-the-Counter Emergency Birth Control on Teenage Pregnancy’, Health Economics, 15 (Sept): 1021- 32