Law - compliance and incentives
Justice versus Social Engineering
Lesson from "Sun Tzu on The Art of War"
Related topic - Marginal increases in liability as net income increases
Related topic - Marginal increases in total deductions as gross income increases
Related topic - Case Study 3 - No incentive for the NRP to earn more
Home & weblog
Blog archive & site history
Site map & search

Justice versus Social Engineering


Sometimes the verdict in a a particular case (court case or otherwise) appears harsh, perhaps not appearing to match the circumstances. But then you ask "if the law were different, how would people tend to behave?" Or perhaps you don't ask this, and continue to miss the point. Single-issue activists, and many politicians, often miss the point.

This article makes some fairly obvious comments, including the observation that the dilemma becomes more acute when children are involved.

Examples of compliance

Sun Tzu showed the way 2400 years ago.

Non-resident parents who don't pay

Regulations now "give power to magistrates courts in England and Wales, and the sheriff in Scotland, where there has been willful refusal or culpable neglect to pay child support maintenance, to commit the liable person to prison, or make an order for him to be disqualified from holding or obtaining a driving licence, as they consider appropriate".

This has received various comments:

  • "a distraction from the real issues"
  • "pointless because they will become unable to pay child support"
  • "it would be OK if contact-thwarting PWCs were also suitably punished"

Since 28% of NRPs are "nil compliant", does this mean that 280,000 NRPs are going to be sent to prison or lose their driving licences? Are vast numbers of self-employed NRPs going to lose their livelihood and become unable, instead of just unwilling, to pay? I suspect not. Throw a suitable number in prison, put publicity about this in envelopes sent to non-compliant NRPs, and there should be far less non-compliance. It is worth a try at this little bit of social engineering.

It is important to distinguish between the effect on the ones sent to prison, who may well become unable to pay child support, and the vastly greater number who may see their aspirations in different directions and so start paying! If prison was a certainty for noncompliance - I suspect there would be little noncompliance and few sent to prison. (I wonder how many of the objections about harsh penalties are from people who want to retain the option of being non-compliant?)

There is a proviso. If harsh penalties are to be imposed, the facts of the case must be accurate. The CSA's legendary incompetence is such that the very words "here is evidence from the CSA" must surely be "reasonable doubt" in the eyes of UK law! (The defence lawyer rises to his feet. "My Lord, my client has just one piece of evidence that proves him innocent - the CSA says he is guilty").

Resident parents who thwart contact

If a resident parent, typically a mother, refuses to allow the contact that a court has allowed the other parent, how can she be persuaded to do so?

Courts tend to be reluctant to impose fines on lone mothers, because it will impact the children. If she is thrown into prison, the children will be without a mother for a while. (There is, of course, a father still outside prison). Many people focus on the individual case, and become constrained by the impact in that single case. Very few lone mothers appear to be thrown into prison for thwarting contact. So what incentive is there for them to stop doing so?

As with achieving compliance in nonresident parents, this needs policy changes and the intent to change mass behaviour. It can't be left to officials in individual cases to try to change mass-behaviour. They haven't the mandate to do so, they need to focus on the individual case. A policy is needed to cure this problem, and if necessary to throw lone parents into prison with suitable publicity so that the intent is taken seriously.

Examples of incentives

These are significantly different, but are still examples where laws that may appear OK for given sets of circumstances people to behave in undesired ways. The rules may be OK for circumstance X, and also OK for circumstance Y, but they may unintentionally cause people to change their circumstances (or not change their circumstances) against the intentions of the law-makers.

The UK's child support law is riddled with such cases. The current scheme is very bad. Even the reformed scheme has features which make it worth while playing games at the expense of all the other stakeholders. Several such cases are analysed on this web site.

Sharp increases in deductions as income rises

Here are extracts from an article that looks at the marginal increase in all deductions as gross income rises. Here are the approximate deductions for the current scheme for an NRP with low housing costs and just one qualifying child:

Gross income Net income Child support Retained income
£200 pw about £163 about £5 about £158
£250 pw about £197 about £29 about £168
£300 pw about £230 about £57 about £173
£350 pw about £264 about £79 about £185

For an NRP initially on £200 gross, earning an extra £150 would leave an extra £27 after all these deductions. Indeed, increasing earnings by £50 from £250 to £300 would leave just over an extra £5! These earnings cover the range from well under to well over average NRP earnings ( which is £207 net). They represent earning at which in other circumstances NRPs would probably be trying to "better themselves".

Here is a similar table for the reformed scheme for an NRP with 3 qualifying children:

Gross income Net income Child support Retained income
£150 pw about £130 about £19 about £111
£200 pw about £163 about £33 about £130
£250 pw about £197 about £49 about £148

Earning an extra £100 gross leaves the NRP with an extra £37 of spending money. (The reformed scheme is better than the current scheme. This leaves him with a higher proportion of his extra gross earning - about one-third here, instead of about one-sixth in the earlier table. It is also only as bad as this if there are 3 children).

These deductions must be particularly galling if the PWC is on Income Support so the children don't benefit from any of the extra money paid. They certainly make it worth while doing everything possible to hide any increased earnings! Who can blame a nonresident parent for saying "stuff it, why bother to try to earn more?"

Has the government done such analysis? If so, did it really mean what it said in the CSA Reform White Paper: "At least 75 per cent will be retained by the nonresident parent. This gives nonresident parents a clear advantage from working the longer hours that overtime pay covers while at the same time giving a share of the increased prosperity to the children." If the government really wanted to provide the incentives they claim, they would "band" child support like income tax. When you cross a tax band boundary, you don't suddenly pay tax at that rate on all your income, only on the part which has entered that band.

Interactions with means-tested benefits & tax credits

Means-tested benefits (and tax credits) are provided by a "benevolent" government as a safety net. As your income disappears, you still survive. Unfortunately, the necessary consequence of this is that as your income rises, the benefits and tax credits disappear. This results in well documented "poverty gaps" and "employment gaps" where there are few incentives to work and/or work harder. Where there are bad interactions, people can be worse off working, for example when their passported benefits disappear, or their cost of working isn't covered by the difference between their income and their previous benefits.

Apart from the straight financial effect, there are poisonous social effects too. Many people see benefits at a certain level as their right, not as a top-up for poverty relief. Since declaring income reduces their benefits, they have incentives to hide income, or play other games. Some even claim that they have earned the right to do so because they have paid taxes for years. (This is approximately as silly as saying that because they paid for insurance on the house for years, they have a right to get money back even if they haven't a legitimate claim, so they make up a claim. What they were paying for was "peace of mind", not a savings scheme).

The current CSA scheme caused a lot of fraud because of the way the child support payment interacted with a lone parent's Income Support. The Income Support wasn't seen by many parents as a top-up for poverty relief, it was seen as a right that was being interfered with by the CSA. In principle the government has a perfectly plausible case for reducing the Income Support £ for £. After all, Income Support is to overcome a need - and those parents didn't have the need. (If they did, then so would those not receiving child support, and then the whole rate of Income Support would have to rise accordingly).

But ... at least two things went wrong. First, very many people don't see things in that same way, and will always thwart government efforts to make the rules work precisely. Second, the government, in effect, lied to the public that child support was about helping the children, when in fact it was making them worse off by simultaneously reducing their social security payments and eliminating any private arrangements they had (aka "benefits fraud").

Means-tested benefits (and tax credits) are a close fit to the statement "the road to hell is paved with good intentions". It has been demonstrated that however good the intentions, they mix badly with child support. Does the fact that people play games and cheat and behave badly mean that government should simply let them do so? No - but government should design systems which simply avoid such incentives! However "right" and "valid" it was to reduce Income Support £ for £, it was also careless to defy human nature. Yes, people should respect the law - but lawmakers should make laws that are capable of being respected. Those game players are giving the government a message - "if you don't want us to play games, don't give a goal to shoot at".

A theme of this web site is that, as far as possible, all interactions between child support and means-tested benefits and tax credits should be eliminated. Such interactions will always cause people to play games at the expense of the other stakeholders. The lesson from the avoidance of interaction between child support and the PWC's Working Families Tax Credit (WFTC) is - it can be done, and it doesn't cause the Treasury to run out of money!

The reformed scheme is risking disrespect because of the interaction between child support and the NRP's WFTC. "Bad" behaviour is inevitable, and not all of it is against the law. This interaction could easily be avoided, and make the reformed scheme even simpler in the process.


The history of child support appears to be that the lawmakers focus on making a particular circumstance "right", and lose sight of the resultant incentives for people to move into or out of that circumstance. Although people are not purely rational beings, vast numbers of separated parents will behave in ways that improve their retained income, even if they act against some aspect of government policy to do so.

The single lesson here is that for every bit of legislation, the analysis "how will people behave to improve their finances as a result of this legislation" should be done and acted upon. The ideal is that separated parents should "do right" because the incentives are there to do so, at least to the same extent that penalties are there if they don't.

JUDGE: "I must ask you to banish all present information and prejudice from your minds, if you have any".
Page last updated: 7 July, 2004 © Copyright Barry Pearson 2003