Many of today’s Conservatives have a taste for bans and mandates rivalling that of politicians on the Left, as we saw repeatedly during Mrs May’s administration. The latest example comes from Health Secretary Matt Hancock. His idea, floated at the party conference, is to ban children from attending school unless they have been vaccinated. Downing Street has distanced itself, but I wouldn’t count on the plan being abandoned. Several other countries, including France and Italy, have already gone down this road.

Mr Hancock’s concern is that, despite vaccination rates being high by international standards, they have fallen in recent years (as they have been in many developed countries) and this is threatening to lead to increased incidence of disease.

For example measles, which had been virtually eradicated in the UK, seems to have returned. The uptake of the controversial MMR (measles, mumps and rubella) vaccine fell sharply below the target of 95 per cent coverage for several years. Although it has recovered recently, it is still below target and the UK has lost its official “measles-free” status from the World Health Organization.

Economists often see vaccines as an example of market failure. The argument focuses on the concept of “herd immunity”, which occurs when a large percentage of the population has become immune through vaccination and an unvaccinated person is therefore at reduced risk of getting the disease. Such an individual gets an external benefit from the vaccination of other people.

Herd immunity is then a kind of “public good” – non-rival and non-excludable – which may be underprovided if parents decide not to vaccinate their children. They can be seen as taking a “free ride”.

There are qualifications to this argument. In the past some vaccines have had serious side-effects, and such a risk is a rational reason to avoid vaccination for some children who may be susceptible, for example those with weakened immune systems. The danger is that most parents are non-experts, and may be swayed by scares such as claims that the MMR vaccine predisposes to autism. There is also the possibility that some vaccines may become ineffective through serotype replacement, which occurs if the prevalence of a virus or bacteria declines because of immunisation and allows another, perhaps more dangerous, variant to emerge.

But the medical consensus is that the 13 vaccinations currently offered to UK children up to the age of five are safe and effective, and that ideally all but a small number of children should receive them. And there is evidence that a high take-up of vaccination is cost-effective, in that the financial savings on treatment usually outweigh the costs of the vaccination programme.

So we can certainly justify some government intervention to encourage vaccination. Does this mean, though, that vaccination should be compulsory? I’m not sure whether Mr Hancock knows this – I didn’t until recently – but we have been here before.

Vaccination against smallpox was compulsory in this country for almost a hundred years from 1853. Children had to be vaccinated within three months of birth, or their parents faced fines which cumulated if vaccination wasn’t completed. Ultimately prison was a sanction.

This created considerable opposition, with the creation of the Anti-Vaccination League and huge popular demonstrations on an Extinction Rebellion scale, for more than 40 years. Partly this was because of fear and some degree of ignorance, but it was also in part a classical liberal objection to the growing power of the state. This eventually led to a Royal Commission which recommended that conscientious objection should be permitted.

From 1898 onwards parents could get exemption certificates, on increasingly broad grounds, and tens of thousands did so. The law was scrapped with the arrival of the NHS in 1948. With free universal health care the need for compulsion was seen to have disappeared.

Given the current divides within our society, I would expect compulsory vaccination to create similar if not greater opposition. It could deter many from putting their children into nurseries before the law requires they attend school, or even persuade some to go down the home schooling route after the age of five.

If parents argued the toss it could take months before children were able to start school. In extreme cases children might have to be removed from non-compliant parents, no doubt to popular uproar. If all 13 vaccinations were required, a child presenting for school with an incomplete record could take months to meet this criterion (jabs can’t all be given at once) and would presumably have to miss out on school during that time.

Does Mr Hancock really want to go down this route? It would be a bureaucratic nightmare. The problem is that once you have a law, everybody has to conform and be pursued by the state: on current figures this would suggest around 100,000 parents would need to be approached by medical professionals, social workers and ultimately the police. Even with compulsion you wouldn’t get 100 per cent compliance, as international comparisons show.

But we don’t actually need everyone to be vaccinated to acquire effective herd immunity. Rates of 95 or 96 per cent would suffice.

I always like to look at the data. When we do so, we find that Scotland, Wales and on some measures Northern Ireland already have acceptable vaccination rates of around 95 per cent. The problem lies mainly in England, and within England there are big variations. If we look at the figures for the 12-month uptake of the 6-in-1 (diphtheria, polio etc.) vaccination, for example, the North-East has an uptake of just over 95 per cent. London, however, only has 87.4 per cent.

This is not because London is packed with ideological anti-vaxxers and the BMJ is avidly read in Newcastle and Sunderland. Much has been said about the spread of anti-vaccination ideas through social media, but it’s not at all clear that this is a major factor in differential uptake.

It’s in large measure a demographic thing. The North-East has a much more settled population compared with London, where mobility in and out is much greater. This has consequences for the accessibility of GPs, for example. In some parts of London getting a GP appointment is very difficult and inconvenient for working parents to manage. It is easy to see how vaccinations can be missed.

The North-East is the region with the lowest proportion of births to non-British mothers, whereas in London a majority of new mothers are now non-British. Many of these are from continental Europe, and interestingly the Nuffield Trust points out that most outbreaks of measles now come from Europeans who were not immunised.

There is clear evidence that children from some ethnic groups are much less likely to be immunised than others.

Most Asian heritage children are highly likely to be vaccinated: in London those of Indian background are more likely to be covered than white British children. African and African-Caribbean children are much less likely to be covered, as are White Irish and White Polish. Somalis and some small minority groups have particularly low uptake.

Explanations for these ethnic variations include differential access to GPs, but also in some cases cultural and religious factors and unfounded beliefs about the effect of vaccination on different body types.

There is no easy answer to this, but it suggests that strategies to increase vaccination uptake need to improve access to primary care, perhaps to offer vaccinations in pop-up centres away from doctors’ surgeries, and to be targeted on hard-to-reach groups through education, advertising and community initiatives. Compulsion is certainly not the answer.