Research on how real juries decide cases is illegal. That is a bizarre and unsustainable law in the age of openness. Taxpayers fund more than 800,000 juror sitting days a year: why should we not be allowed to discover how juries operate?

The 21st-century society is an open one. We no longer accept that important parts of government should be operated in dark secrecy. We have a Freedom of Information Act and we expect openness in all parts of the justice system, unless there is some compelling reason, such as national defence.

There is no more important feature of our democracy than the jury, the forum in which it is decided which of us are convicted and punished with up to life imprisonment. In 1957, Sir Patrick, later Lord, Devlin observed that one of the first things a dictator taking over Britain would do was abolish the jury, because “no tyrant could afford to leave a subject’s freedom in the hands of 12 of his countrymen”.

With anything that important, there is a need to know how it works. In so far as Section 8 of the Contempt of Court 1981 prevents research on how juries work, it is an anachronistic law and should be reformed urgently.

In the case against The Times, Lord Justice Pill said that the strength and value of the jury depend on “the open and frank expression of views between 12 people in the secrecy of the jury room” and that “confidence to express views in that way depends on the juror’s knowledge that the views will not be revealed outside the jury room”.

The reason for that principle is that by keeping secret what any named juror says, or how he or she votes, the system prevents blackmail and bribery. But that protection remained inviolate after the Times article because only a volunteer, Michael Seckerson, had been identified. No juror who wanted to remain anonymous had been exposed.

There is now a compelling argument to change the law. The jury is probably a good institution but we do not really know how it works. Experiments with simulated juries have generated some curious results. In 2007, a study suggested that jurors are less likely to return guilty verdicts against attractive defendants. In an American experiment involving 715 people, a truthful speaker was judged to be lying by 74.3 per cent of the subjects and a lying witness was judged to be truthful by 73.7 per cent.

A Crown Court prosecution can be proven only if it is made “beyond a reasonable doubt”. So it is remarkable that if we ask ourselves whether we are sure, beyond a reasonable doubt, that the jury system itself works well, the answer is one worthy of Alice in Wonderland: we do not know and it is illegal to find out.

The author is director of the Open University law programme

Explore Law

• Columnists

• Law Panel

• Law Reports

• Student Law

Times Recommends

• BGC 'set aside £40m to lure Tullett brokers'
• Legal aid review to ensure value
• Fears as 21 magistrates' courts face closure