In defense of IVF subsidies

Why I think it's not great that the US administration has backtracked from its campaign promise to mandate insurance coverage for IVF

In the last few days there has been quite a bit of discussion on twitter about the US administration’s decision to not mandate insurance coverage of IVF¹. Well known pro-natalist Lyman Stone has been arguing that it is indeed good not to cover IVF, with others chiming in as well:

I’ve been working on IVF-related topics for a while and thought I’d push back a bit on some of the arguments I saw. Partly this will be data-driven and partly my own opinion. I will make it clear for each argument whether I’m arguing back on a more factual or more subjective basis.

Argument 1: IVF patients are really old

One of the main criticisms is that we shouldn’t be subsidising people at very advanced maternal ages, with critics making it sound like that is basically the majority of IVF patients. I agree that we have to be careful about subsidising treatments at older ages but I’m very confident that it’s just a mischaracterisation to say that most IVF patients are very old and the patterns in the data clearly show this.

It’s important to acknowledge that age is a crucial factor when it comes to infertility and also IVF success. IVF success rates fall very strongly with age, at least when using a patient's eggs. I have written extensively about this in my post about egg freezing, go check it out if you want to know more! One of the graphs I showed was this:

Source: HFEA register data, own calculations

We can see that when the patient uses her own eggs, success rates fall very very steeply with age. This means that it’s probably not a good use of public money to subsidise IVF for older women given that success rates are so low that the patient would need to undergo lots and lots of rounds for a decent chance of success and that in most cases, the treatment would remain unsuccessful.

The thing is, this is not what most countries do when they subsidise IVF! There are usually very clear age limits and often also other criteria that women have to fulfil in order to be eligible for subsidies. For example in the UK, the official recommendation by the National Institute for Health and Care Excellence is that 3 cycles of IVF should be offered to women below 40 and 1 cycle should be offered to women aged 40-42 if they have never had IVF before and if there is no evidence of low ovarian reserve².

From a quick search, it seems like some US states that mandate insurance coverage for IVF currently don’t have age limits but for example Rhode Island only mandates insurance coverage for diagnosis and treatment between 25 and 42. It would definitely be possible to have mandated insurance coverage in the US with age limits in place.

Additionally, the age distribution of patients who undergo IVF is not what you might think. Let’s look at comprehensive data from the UK³ for who actually does get IVF treatment. Roughly half of all initial IVF treatments⁴ today are women in the 18-34 age group, which is also the age group in which most non-IVF births occur. The 45-50 age group is a very small portion of all treatment initiations and the vast majority of first treatments happen before age 40.

Source: HFEA register data, own calculations

It is true that on average, IVF patients are older when they initiate treatment compared to the general population at birth. Let’s look at ages at birth in the UK and the average age at first IVF treatment:

Source: ONS statistics on births by parents’ characteristics, HFEA report on age at first IVF treatment, own visualisation

The average age at first IVF treatment has just surpassed 35 years. This is around 6 years older than the average age at first birth in the general population which is 29. Higher parity⁵ births happen on average between 31 and 34 years old. The takeaway here is that while there is an age difference it is not as dramatic as you might initially think.

Given these numbers, it seems uncharitable from Lyman when he says “being 47 is not a disease”. The average IVF patient is nowhere near 47 and most countries where IVF is covered by insurance have age limits so that those who are 47 wouldn’t actually be getting public subsidies. If you oppose IVF subsidies going to women who are too old then you can advocate for stricter age limits instead of no insurance coverage for IVF at all.

Argument 2: IVF isn’t healthcare

Another main argument is that wanting a child when you’re infertile for whatever reason is merely a preference and that healthcare shouldn’t be treating preferences. This is one of the areas where I think it gets a bit trickier and I don’t have a clear answer. I personally don’t think this is a great argument for several reasons. First of all, health insurance covers lots of things that you can frame as preferences. In general, we don’t really see healthcare as something that should do the bare minimum but something that should help us improve our quality of life.

Additionally, I don’t really understand why being infertile should be that different from other things that we classify as diseases. Some infertility is not age-related at all but caused by underlying conditions like endometriosis or PCOS on the female side and things like poor sperm quality on the male side. If you have blocked Fallopian tubes, then IVF circumvents the tube and completely solves that problem. If you have PCOS or for some other reason you don’t ovulate, then the hormonal stimulation from IVF makes you produce eggs. If you have poor sperm quality, then manually injecting a single sperm directly into an egg during the IVF process means that you can still have children. How is having any of these conditions and IVF solving them different from having a non-infertility-related disease that healthcare treats? In both cases, these are things that we associate with normal bodily function which impact your quality of life.

And then yes, some of the infertility will be age-related. But this is also different from person to person. Lots of people wait until 35 to have children and do so without problems. If you happen to be unlucky and already run into problems at 35, is it then your fault for having waited until 35? Again, with other health conditions we usually don’t ask how much of the condition is your own fault and then refuse to pay for your treatment because you had an unhealthy lifestyle beforehand.

Argument 3: IVF incentivises postponement and has no effect on the birth rate

I think the strongest counterargument to subsidising IVF is that it could incentivise fertility postponement. The idea here is that if people know that free IVF is an option later in life, they will be more likely to wait to have children and rely on this backup option if natural conception then fails. Maybe if IVF wasn’t an option or was less appealing because of the high costs, these people would have just had children earlier in life without running into fertility problems.

There is some evidence that this might be true: Machado and Sanz-de-Galdeano 2015 use the staggered rollout of insurance mandates for IVF in different US states to identify the effects of coverage. They find that insurance coverage led to slightly higher average age at first birth of about 3 to 5 months in the treated states and that there was no significant positive effect on completed fertility.

It’s true that this might be a problem and that we should keep this in mind when thinking about promoting and subsidising IVF. Unfortunately, IVF is currently far from perfect and it can’t fully make up for the increase in infertility that you get when postponing childbearing⁶. I hope that one way to mitigate this problem is to make sure that people understand age-related fecundity decline and know about the limitations of IVF. Currently, knowledge about IVF is limited and it is not widely known that IVF is really not a miracle cure, especially at older ages. It does seem unlikely to me though that the second-order effect of incentivising slightly more postponement exactly compensates or even overcompensates for the first-order effect of helping people conceive.

Given these arguments, I still come out in favour of covering IVF. I think there are lots of reasons why people postpone childbearing. Stopping to subsidise IVF certainly wouldn’t fix those, for example the fact that education timelines have gotten longer and that motherhood penalties are smaller for older mothers. Of course it would be great to fix those things as well but in the meantime I would prefer if people who happen to run into difficulties don’t have to financially ruin themselves to have children or forgo treatment altogether because they can’t afford it.

I agree that IVF subsidies are in general not a very effective way of pushing up birth rates and that IVF births will probably remain a small fraction of total births in the near future. However, I care about raising birth rates for several reasons and one of them is simply that I want people to have the children they want. If someone has decided to postpone childbearing for whatever reason then I still want to help them have a child, which might mean subsidising IVF.

Why I think IVF should be covered

Apart from the arguments that I’ve given in this post, in the end I think the main reason why I feel strongly about this topic is more emotional: For many people it just seems really really hard to be involuntarily childless and being able to do something about that shouldn’t depend on whether you can afford IVF or not. I find the overall discussion about this slightly callous. Obviously there are always trade-offs and unfortunately we can’t cover every treatment that slightly improves someone’s quality of life. But what we’re talking about here is a really fundamental thing in people's lives. Within reasonable age limits, providing a certain number of IVF treatments has a good chance of giving someone the child that they really want and I think we shouldn’t lose sight of that.

1. The general context here is that currently there are some US states that mandate that insurers have to cover infertility treatment whereas other states don’t. The campaign promise had been to mandate coverage everywhere in the US so that it wouldn’t depend on which state you live in whether you get IVF coverage or not.

2. This would be measured by the level of anti-Mullerian hormone (AMH) which indicates the quality and quantity of the egg reserve and also by the antral follicle count (AFC) which is a simple count of the number of follicles that are currently present in the ovaries.

3. I keep using UK data because that’s what I have good access to. It’s reasonable to expect that the numbers in the US won’t be that different. Unfortunately, the only direct information that I could find for the US is that about 5% of treatments happen in women older than 44. Based on this data point the fraction of older patients is somewhat higher in the US than in the UK.

4. Note that I focus on first treatments because I can’t link several treatments to the same person based on the data that I have. By only looking at first treatments, I can actually characterise the patient population.

5. Birth parity is just a fancy word for the number of previous live births that a woman has had.

6. And there is also the fact that it is a physically taxing, invasive procedure and nobody really wants to go through IVF instead of conceiving naturally.