What is the criminal law for? That deceptively simple question was addressed recently in a masterly paper by Professor Jeremy Horder, issued just a few days before he completed his term as the commissioner responsible for advising ministers on reform of the criminal law.

As a subject of academic study, criminal law has always been regarded as something of a soft option. When I was at university, it was a topic you got out of the way in your first term before moving on to something more intellectually stimulating – such as criminology or jurisprudence.

And, until very recently, it was possible to reach the summit of the criminal bar without having taken a degree at all. The Supreme Court is notoriously short of criminal specialists and the law lords were renowned for producing unintelligible or unworkable rulings in criminal appeals.

The last judge with extensive criminal experience to sit as a law lord for any significant length of time was probably Lord Edmund-Davies, who served from 1974 to 1981. Since he was one of the most distinguished lawyers to graduate from King’s College London, it is fitting that the chair of criminal law which Jeremy Horder has just taken at the college is named after him.

Horder’s consultation paper, published by the Law Commission under the suitably serious title Criminal Liability in Regulatory Contexts, attracted publicity for its recommendation that low-level criminal offences should be removed from the statute book. By a happy coincidence, the coalition government is also currently appealing to the public’s appetite for scrapping unnecessary legislation and quaint bylaws.

But there is rather more to the Law Commission’s proposals than simply finding laws to abolish. If implemented, they would amount to a wholesale shift from criminal punishment to civil penalties – not so much a bonfire of the regulations as an incineration of the statute book.

The proposals originated in a move by Labour to reduce the burden on businesses. Horder’s paper is therefore restricted to activities that are subject to regulatory enforcement – in other words, areas in which a government department or agency has been given powers to enforce standards of conduct. These range from regulating solicitors to waste disposal, from farming to food safety and from banking to workplace health and safety. So the commission’s proposals would apply only to licensed traders who were part of a pre-determined group: chemists, for example, but not drug abusers.

But the power to make secondary legislation is not confined to local authorities. ‘There are now over 60 national regulators with the power, subject to certain limitations or checks, to make (criminal) law,’ Horder notes.

And yet they rarely use the crimes they create. For example, legislation passed in 1996 made it an offence to employ illegal migrant workers. There was, on average, only one prosecution a year before a new agency was created with power to impose fixed civil penalties instead.

Of course, the scarcity of prosecutions may simply mean that everybody is obeying the law. But it is much more likely that prosecutions are simply not seen as cost-effective. As Horder says, ‘ordinary people and businesses are being subjected to ever-increasing numbers of what, in all probability, will turn out to be illusory or empty threats of criminal prosecution’.

He is also concerned that ministers are taking powers that are far too wide. Under the Animals Act 2006, for example, a secretary of state could make it an imprisonable offence for a pet owner to tether a dog outside a shop in wet weather or to stroke a cat’s fur in the wrong direction.

So the Law Commission’s primary provisional recommendation is that the criminal law should be reserved for the most serious wrongdoers: those who deserve the stigma of a criminal conviction because they have engaged in ‘seriously reprehensible conduct’.

In the commission’s view, acts or omissions should not come before the criminal courts if the outcome is going to be little more than a fine. For behaviour to merit criminalisation, even a first offender should justifiably face imprisonment.

On a broader level, Horder argues that it has become far too easy for government ministers and agencies to create new crimes. In his view, this should be regarded as a step of almost constitutional significance and one that should be reserved to parliament.

But do we really want to lose the safeguards of the criminal process and replace them with a form of administrative justice? What if officials are insufficiently independent or fair? Criminal courts may be ponderous and expensive but at least they embody safeguards against wrongful convictions.

Horder’s answer is that there must be an unfettered right to challenge any civil penalty in the courts. He might also have added that a civil penalty carries little stigma these days: for example, being caught for speeding does not disqualify you from sitting as a judge. On the other hand, more serious regulatory breaches may lead to the loss of one’s livelihood.

Although this is only a consultation paper, its promise of £11m in annual savings should appeal to ministers. But they would do well to look beyond the press release: the report itself makes it clear that the actual savings could be anything between £500,000 and £52.8m – a pretty broad spread.

Finally, Horder’s answer to the question I posed at the outset. ‘The criminal law should be employed only when engagement in the prohibited conduct in question warrants official censure because it involves a harm-related moral failing, not just a breach of a rule or simple departure from a standard.’ As the exam papers used to say, ‘discuss’.