The social costs of fraud and its prevention
By Bruno Prior
The social costs of fraud and its prevention
The IEA is an educational charity in the classical liberal tradition, whose mission is to improve understanding of the fundamental institutions of a free society by analysing and expounding the role of markets in solving economic and social problems.
It might seem strange, therefore, that we have published a paper that discusses how to use the law rather than markets to address an economic and social problem (fraud). But classical liberals recognise the key role that the institutional framework (of which the law is a key part) plays in the allocation of resources. Property rights are core to that institutional framework. Fraud undermines the voluntary exchange of property.
The important role of law and property rights in classical liberal economics does not mean that any proposal to tighten the law to protect property is inherently classical liberal. Law and property rights play an equally important role in other schools of economic thought.
Within the classical liberal fraternity, there have been differences of perspective on the relationship of law and economics. Four of the most influential figures in the IEA’s history came at it from different (but not antithetical) angles:
Friedrich Hayek (following Leoni) developed a sophisticated insight differentiating between law and legislation.
The (Milton) Friedman Doctrine (the social responsibility of business is to increase its profits) assumed (and promoted) the effective legal protection of property rights.
Ronald Coase emphasised the role of law in achieving efficient outcomes through voluntary agreement.
James Buchanan thought economics should play a stronger role in the process of legislation than in the application of the law, but recognised that legislators’ incentives would be a bigger influence than economic understanding on the legislative process.
But on related points, there would have been little difference of emphasis between them:
Freedom is the absence of coercion.
There are no costless solutions, only trade-offs.
Legislation is likely to be imperfectly written and implemented, whether because of limited knowledge or the incentives on those involved.
Legislation and the state should be no more extensive than required to carry out the essential functions of limited government.
Legislation must be clear, objective, and practical to comply with for all who are covered by it.
The law constrains law-abiders more than law-breakers.
Insights such as these (not simply support for the rule of law) are the common thread that binds varying classical liberal perspectives on law and economics, and (cumulatively) differentiates them from other perspectives.
* * * * *
In “Fixing Fraud”, Dr Cronin outlines some changes to legislation that she believes would reduce the incidence of fraud. They extend and amend measures in the Economic Crime and Corporate Transparency Act 2023, which institutes strict corporate liability for a failure to prevent fraud (tempered by the provision of a due diligence defence). Senior officers of the corporation are identified with the corporation and face criminal sanction if the corporation fails in its duty.
For Dr Cronin, this does not go far enough. She is disappointed that it is limited to companies “with at least two of the following criteria: a turnover of at least £36 million; a balance sheet total of at least £18 million; a total of at least 250 employees”. She regards the prevalent use of deferred prosecution agreements (DPAs) to convert effectively a criminal into a financial sanction as a “deterrence deficit”. She wants to see companies of all sizes prosecuted for failing to prevent an employee from committing fraud, and their senior officers taken to criminal court without the option of a DPA, pour encourager les autres.
With government tax and spending at record levels and heading further upwards, minimal growth, businesses weighed down by an ever heavier legislative and regulatory burden, persistent large deficits and escalating debt, you would think that classical liberals would be more focused on reducing the burden of government than extending it. But let’s concede that the need for smaller government in general does not preclude the possibility that specific legislative additions may enhance our institutional framework.
And let’s also concede that stronger and wider sanctions probably have an effect at the margins and will reduce to some extent the offence at which they are targeted. This is thus a proposal that increases the protection of property rights. Why should classical liberals not embrace it?
Why stop there? Let’s make fraud a capital offence. That would provide a stronger disincentive, and greater protection for property rights.
I hope you baulked at that. Most reasonable people would regard that as disproportionate. We understand intuitively that legislation is not justified simply because it would have the effect intended. We do some sort of subconscious cost/benefit analysis and discount draconian measures. People have different tolerances for state action, but we all understand that it is a trade-off at some level.
Classical liberals’ tolerance for state action should be low. Freedom is the absence of coercion. Other than anarchists, we accept that some amount of state coercion is necessary to prevent coercion by individuals. But we believe it should be kept to the minimum necessary.
Is it a minimal use of state coercion to criminalise business managers for failing to prevent an employee from committing fraud?
Dr Cronin regards as an efficiency (not a burden) that the law uses the threat of criminal sanction to turn business managers into law enforcement agents.
By affording a central role to the internal compliance system, corporations [and, if Dr Cronin had her way, all businesses] are delegated responsibility for policing the behaviour of individuals within their sphere of influence, and criminal law enforcement is, in effect, privatised.
My classical liberal philosophy regards the co-opting of private organisations to act as agents of the state, extending its tentacles well beyond the reach of even its current over-grown condition, as one of the worst developments of the past 30 years, not an efficiency. Would it be the ultimate efficiency for all government actions to be carried out through private organisations, and for those organisations to be burdened with so many state responsibilities and such severe sanctions that their managers can barely think about their primary purpose? That is a new kind of classical liberalism to me. There should probably be a different name for it.
Part of the “efficiency” that Dr Cronin admires is that:
there is no need to prove the culpability of individual high-ranking officials, as is the case with the common law identification principle, or to establish a causative link between the deficiency of the internal system and the commission of the predicate offence by the company’s employee or associate. All that is required of the prosecution is evidence that the base offence has been committed by an employee or associate of the company, although this person need not be prosecuted as an individual.
In the name of efficiency, the burden of proof is reversed. The senior officers are guilty unless they can prove that they had implemented measures that should have been sufficient to prevent the employee from committing the offence. As they would be in court for having failed to prevent the offence, that will presumably be quite difficult.
(It is worth noting that Dr Cronin regards this as a virtue of codification over common law. Some might regard this as evidence of the opposite – that common law is more consistent with the Anglo-Saxon, Lockean conception of the relationship between the state and the individual, and that codification is more suited to the continental, Rousseauian conception.)
Under Dr Cronin’s favoured extension of the law to cover businesses of all size, there would be some kind of sliding scale for what might be considered sufficient measures:
the requirements of the compliance system are proportionately tailored to what can be reasonably expected in the company’s individual circumstances, if required at all.
How that scale might be implemented in practice is left as an exercise for the reader – a high-stakes exercise if the reader is a business manager trying to anticipate what might or might not be an adequate defence in a criminal court. This is not a proposal for legislation that is clear, objective, and practical to comply with for all who are covered by it.
The proposal assumes that people respond to incentives in one specific way, but are otherwise rather non-human idealised agents.
In that world, faced with a duty and an incentive, they implement systems that successfully fulfil their duty. In reality, unintended consequences are commonplace because legislators tend to make these assumptions that people respond uniformly in the way they want them to respond. Legislation is likely to be imperfectly written and implemented, whether because of limited knowledge or the incentives on those involved.
In practice, people’s responses would vary. Officers of large businesses generally have the awareness of legislation and resources to deal with it, and will do what is required. But their incentive is not, as Dr Cronin imagines, to prevent fraud. Their incentive is to implement systems so that they cannot be prosecuted if fraud occurs.
They are practiced at this. Corporations are weighed down with the associated procedures and HR resources, because of the quantity of legislation that already places analogous burdens on them. This would add more bureaucracy. It would be very unusual for corporate leaders to be successfully prosecuted. It might prevent some fraud, but you could expect fraudsters to be inventive in adapting to the systems, and for officers not to be aware of these adaptations, nor consequently to hunt for them much further than any impact on the bottom line, so long as they are legally insulated. The law constrains law-abiders more than law-breakers.
Mid-sized companies would do likewise, but the cost would sit more heavily on them. At some level, you cross a line where the company is small enough to lack the knowledge or resources.
Some managers of small companies would inform themselves and struggle manfully to comply, but not have the skills and resources to systematise efficiently to minimise the burden. It would be dealt with on a more ad hoc basis, and mistakes would be made. Some of these would be prosecuted.
Managers of other small companies would be so stretched by their existing responsibilities that they would not fully inform themselves, and would fail to implement appropriate measures. Some would be unlucky enough to employ a fraudster, and would end up in court.
Dr Cronin believes there is no concern about the ability of smaller companies to comply with the legislation because of the “proportionate tailoring” quoted above. But it is easy to assume reasonableness without setting out how it would work, especially when you are not the potential victim if the reasonableness is less than certain. “First, assume a can-opener.”
The assumption that the legal system would be fair, proportionate and sympathetic to the limited capabilities of small players seems particularly unsafe in two-tier Britain. You may disagree with that characterisation, but it is harder to deny that it is widely perceived and would affect people’s expectations, justifiably or otherwise.
A further tranche of managers of small businesses would therefore simply not exist. Against their hopes for the business, they would weigh the risk of prosecution and the additional burden of avoiding that risk on top of their many other responsibilities, and judge that it wasn’t worth starting a business or taking a management position.
Management costs should certainly increase to compensate for the increased risk of criminal liability and workload to try to reduce that risk. The risks of taking on an employee would be increased. Some sole proprietors would decide to stick to a workforce of one that they trust not to commit fraud.
At the margins, some startups and growth prospects would be deterred or killed. Although most businesses remain small, every large company started small. If you deter the establishment or expansion of small businesses, you reduce the pool from which growth emerges. You could call it the European model.
* * * * *
There are no costless solutions, only trade-offs. The proposal weighs a figure for the incidence of fraud on the benefit side of the equation, and nothing on the cost side.
For a variety of reasons, the figure cited for the incidence of fraud is not an adequate measure of the benefit of the proposal. It claims that fraud is “extraordinarily prevalent”, based on a single figure from a Home Office report: “Fraud accounts for 41% of all offences committed in England and Wales” (footnote 8, p.12). That is indeed what the cited report says, but let’s have a closer look at what that figure actually means.
The Home Office report relies on statistics for “Crime in England and Wales: Appendix tables: Table A1” for the year ending December 2022. Worksheet A1 reports figures from the Crime Survey for England and Wales. From a sample of 26,621 adults, it estimates that fraud constituted 3,651,000 of a total of 8,985,000 offences, scaled to the population. That is indeed 40.6%.
But that is not actual offences. Rather, it is people’s recollection of their experiences. Offences actually recorded by the police are found in Summary Table 1. Of those, “Fraud and computer misuse” constitute 1,167,492 out of a total of 6,595,116 recorded offences, i.e. 17.7%.
The figure of 40.6% should not be described simply as fraud’s proportion of offences. It is the estimate from a survey of people’s impressions of fraud’s proportion of domestic offences. Non-domestic offences are not counted. The proportion of recorded offences is much lower. And that includes computer misuse, so fraud’s proportion is smaller still.
Of that number, which Dr Cronin does not attempt to assess, one has to judge what proportion would be affected by the proposal, because not all fraudsters will be committing their crimes at work. Of that subset, one then has to estimate what proportion would be prevented by legislation that incentivises primarily the covering of backsides.
Perhaps this is 99% of 99% of 99%. And perhaps people’s perceptions of crime are more accurate than the police’s records. And perhaps non-domestic crime is insignificant and the domestic figures are almost representative of the total, even of the proportion of fraud that is committed at work (as though fraud is primarily a B2C activity).
But if you propose to increase coercion and the bureaucratic burden on an economy and society already teetering under the weight of its legislation, regulations and other public-sector costs, it is incumbent on you to try to be as accurate and realistic as possible about the benefit. One estimate of the frequency of the offence is not sufficient, unless you show that it is reliable, and that a large proportion of those offences will be addressed. Without it, you cannot say whether the undoubted infringements on liberty and the economy are worth the benefit.
And let’s be realistic. It isn’t 99% of 99% of 99%, is it? How is there such a discrepancy between people’s estimates in the survey and recorded offences? Perhaps incidents of fraud are under-reported. But perhaps people only report serious incidents, and the discrepancy reflects the minimal subjective valuation of the unrecorded proportion.
It wouldn’t be surprising if it were the latter. I am quite surprised that the survey figure is as low as 41% (or 50% if we include computer misuse). Who nowadays does not regularly experience an attempt at fraud on the phone or by email? It is almost a daily occurrence for many of us. How many of those calls, many of which are evidently from the sub-continent despite their UK Caller ID, will be deterred by the proposal?
How much liberty would I sacrifice and bureaucracy would I tolerate to implement a measure that would have no impact on this common form of these offences? Absolutely none. Could one reasonably judge otherwise?
Instead of criminalising managers for failing to prevent employee’s crimes, why don’t we keep clamping down on foreign calls with UK Caller IDs? The work is already in hand , and has a more direct impact on one of the most common experiences of attempted fraud, without any meaningful infringement on our liberty.
Something similar could be done for phishing emails and the like, via a voluntary system that incurred a tiny payment on each sent email as evidence of good faith, because the high-volume-low-success-rate spamming model is only viable at zero marginal cost. Again, more effect for less infringement of liberty.
Once we have reduced these offences, what other forms of fraud and computer misuse would remain, how prevalent would they be, and what would be the most effective ways of dealing with them? We ought to know their nature and scale to understand whether threatening British managers with criminal proceedings in order to force them to police their employees is an effective and proportionate measure. Legislation and the state should be no more extensive than required to carry out the essential functions of limited government.
* * * * *
Sadly, Dr Cronin’s paper attempts no such exercise. The sum of the justification is that people think they experience a lot of fraud, and this is a way that we could try to force companies to police one form of fraud. That is not sufficient. It is little more than the Politician’s Syllogism: Something must be done. This is something. This must be done.
It is a substantial extension of state coercion, and therefore loss of liberty.
The trade-offs are not considered. It is treated as a costless solution, when the costs are actually substantial, while the benefits are uncertain.
It assumes that legislation will work as intended and people will behave as expected. There is no acknowledgement of the knowledge problem nor public choice issues.
It is not demonstrated that the proposed extension of legislative powers is essential nor even a good way to carry out the essential functions of limited government. Not only the trade-offs, but also alternative options are not considered.
It assumes that the legislation will be both adaptable to each person’s circumstance, but also clear, objective, and practical to comply with, without showing how. Experience suggests otherwise.
Like so much modern legislation and regulation, it will impose costs on the majority of law-abiders, without necessarily preventing the minority of law-breakers from adapting to it.
From a deontological perspective, it is not just inconsistent with classical liberal principles, but contrary to them in many ways. From a utilitarian perspective, there is no meaningful attempt to weigh all the factors necessary to judge the welfare impact.
Dr Cronin views a move from the limited scope of common law to the more comprehensive scope and specificity of legislation as a virtue. It is doubtful that Hayek would have agreed. The longer forms of his distinction between law and legislation were “Nomos: The Law of Liberty” and “Thesis: The Law of Legislation”. Common law as the embodiment of liberty and legislation as its antithesis? That is perhaps too simplistic, but one ought at least to explain why human order is superior to emergent order in this case, and not simply treat it as self-evident. This analysis seems uninformed by Hayekian insights.
That Milton Friedman believed that the prevention of fraud was an important role for the government does not imply that he would have supported a measure simply because it aimed to strengthen that protection. He was even more concerned about government over-reach than he was about preventing fraud. He was clear where the balance of risks lay:
Many people want the government to protect the consumer. A much more urgent problem is to protect the consumer from the government.
He disliked intensely the criminalisation of people who are not themselves harming others, and was clear that economics could not be an adequate justification for doing so. Classical liberal principles of limited government and individual liberty outweighed any utilitarian economic argument:
I think it's absolutely disgraceful that our government, supposed to be our government, should be in the position of converting people who are not harming others into criminals, of destroying their lives, putting them in jail. That's the issue to me. The economic issue comes in only for explaining why it has those effects. But the economic reasons are not the reasons
The Chicago School does not, at more than a superficial level, provide a conceptual framework that enables Dr Cronin’s proposal to be viewed positively from a classical liberal perspective.
Coase would also have been interested in the enforcement of property rights. But would he have believed that the right way to determine the most efficient method of enforcement of one aspect and who should bear that cost was for legislators to impose the responsibility on certain people on pain of criminal charges? Probably not. There is not much voluntary exchange in that allocation. He might have highlighted that the minimal attempt in this proposal to take account of the costs and benefits to all parties illustrated why legislative imposition is inferior to market allocation. It is doubtful that he would have regarded acting as involuntary law enforcement agents as part of the nature of the firm.
It is an economic insight that property rights underpin markets. But that is far from a sufficient understanding of economics to form the basis of a legislative proposal. That the proposal consequently overlooks some important considerations is an accidental illustration of the wisdom of Buchanan’s emphasis on the value to the legislative process of a deep understanding of economics. One such lacuna is a more realistic assessment of how the legislative process would function and how people would react to the incentives thus created, informed by the public choice school that Buchanan helped to found.
An exploration of the deontological and utilitarian pros and cons around this proposal would have helped to improve understanding of the fundamental institutions of a free society. Markets would have made a more prominent appearance, not just as a beneficiary of the proposal, but also as a yardstick for the justice and efficiency of the proposal. Sadly, as it stands, that is not what Dr Cronin’s proposal does.