The benefit labyrinth: a cornerstone of the poverty trap

Allister Heath once labelled Gordon Brown “a man who seems to love making simple things as complicated as possible”. A recent analysis of the benefit system by the Centre for Policy Studies shows that in this respect, Brown is in good company.

Suppose the government considers societal group X, recipients of X-benefits, as materially disadvantaged, and wants to give them a hand. It would then probably raise the level of X-benefits, right?

Wrong. In the UK benefit system, things are not that simple. To support group X, the government would probably add an “X-premium” to Benefit I, raise the income limit beyond which members of X face withdrawal of Benefit II, and grant them a special lower withdrawal rate for Benefit III. However, the higher amount of Benefit III would be partially offset by a reduction in the amount of Benefit IV, while Benefit II would be liable to income tax and National Insurance (NI). Benefit I is tax-exempt and not counted against other benefits, but recipients of Benefit I lose entitlement to Benefit V, which is a necessary precondition for the receipt of Benefit VI and VII. Needless to say, the agencies administering these benefits use differing definitions of terms like “additional earnings”, “income” and “gainful employment”.

The CPS report argues that apart from the time and effort it takes to find one’s way through the labyrinth, the uncertainty created by the complex pattern of benefit interaction contributes to trapping people in poverty. Consider the case of a single mother in part-time employment, who receives multiple benefits, say Working Tax Credit, Child Tax Credit, Housing Benefit and Council Tax Benefit, and who is at the same time liable for income tax and NI contributions. For her, increasing work effort slightly can have repercussions on each of these variables. If the net effect is not straightforward to see, and if she may well be indiscernably better off than before – why bother?

The report proposes a four-step-plan for a radical simplification of the system. Variables like the definition of income, the tax treatment of benefits and personal allowances would be standardised. The chaotic relationship between benefits would be replaced by a situation in which one redistributive instrument serves one redistributive purpose. Benefits with similar aims would be merged, withdrawal rates aligned, and all benefits would be administered by the same agency.

Not bad at all for a start. But if we really want to break up the poverty trap, simplification can only be the trailblazer (as the CPS report readily acknowledges). Even a very simple benefit system can make it economically irrational for its dependants to seek work, and discourage those who have taken the first step to go further. Even a very simple system can be hugely wasteful, by redistributing large amounts among the middle classes instead of targeting the poorest.

A full-blown reform would clear the way out of the pit. It would end the current no-strings-attached welfare provision along the lines of the Wisconsin model, stop paying benefits to reasonably well-off people, and substantially raise the personal income tax and NI allowance to boost work incentives. Simplification, and the transparency that comes along with it, may well be the conditio sine qua non.

Kristian assumes the objective is to allow individuals to make sensible economic choices, based on their own circumstances and preferences; hence he recommends simplicity and transparency in the benefit system.But we are not politicians. What if the real objective is to make everyone dependent on the state? Then there is much to be said for complexity and obliquity. Moreover, with that different objective, producing a system that ‘traps’ people in poverty is not a disaster but a triumph!

Some excellent points here, but the crux to benefit reform is the point made at the start of the last paragraph, namely conditionality. The key problem of the UK benefit system is that it places no expectations or responsibilities onto recipients, but rather encourages them to maximise their ‘entitlements’. The result is passivity and the absurdity of over 5 million people of working age outside the labour force during the last economic boom, with most new jobs being taken by migrants from the accession countries.

As well as simplification, I think there may be a case for absolute reductions in certain benefits in order to improve work incentives. Many workless families receive a similar amount to an employee on the median wage, once housing costs, travel and taxes are taken into account.

If one wants to reduce the extent of state interference (and one does!), surely there must be a more significant role for voluntary charities. But it’s not easy to see how one could make the transition from where we are to where we might want to be. Abolishing the Charity Commission might be a useful start.

Kris, This sentence – “Even a very simple system can be hugely wasteful, by redistributing large amounts among the middle classes instead of targeting the poorest” – is inconsistent with the rest of your well-argued post. Means-tested withdrawal is the main reason why benefits create a strong disincentive to work, by creating high marginal effective tax rates.

Bruno,
that’s true for the present situation, but I guess that could be reformed. At present, the withdrawal rate for working tax credits is 39% – that’s a fairly high effective marginal tax rate, but not excessive. It’s the interaction with the tax and NI system that raises EMTRs to 70% and beyond. Raising the personal allowance would eliminate such overlaps.
The withdrawal rates could be lowered further if the intital benefit level is lowered alongside, and conditions are attached (e. g. time limits).
But there remains a conflict between targeting and incentive-compatibility.

Kris, Mechanistic solutions, whether Basic Income + Flat Tax (BI+FT) as I favour, or increased personal allowances as you prefer, are not the real determinant of the level of marginal effective (ME) taxation. BI+FT could produce high ME tax rates, simply by setting the FT at a high rate. So could increased personal allowances, by setting the rates of withdrawal of benefits below the allowance and the rate of tax above it quite high. What really determines the impact of our proposals is that we recognize the importance of ME tax rates, and therefore propose that the systems be structured to produce a profile of disposable income vs earned income that keeps ME tax rates low.

But we must also acknowledge that, unless we are prepared to reduce the level of support provided to people with zero earned income, this change in the profile is achieved by increasing the disposable income of low-to-middle-earners relative to their earned income. And while that is undoubtedly a good thing, we have to recognise that a side effect is to substantially increase the net cost to government of income taxes and welfare provision. We may believe that in the long-term, the Laffer Effect will close the gap, but in the short-term, we have an obligation to set out how we would increase other taxes, cut spending or increase borrowing further to compensate for this effect.

That is not to deny the value of the mechanistic solutions. The virtue of both your and my proposal is that they simplify the system substantially, which has benefits in reduced bureaucratic costs, and for the ability of people caught in the welfare net to focus on getting out of it rather than on understanding the system to maximise their income from it. In my opinion, the benefits of BI+FT in this regard exceed the benefits of any system that retains means-tested benefits, whilst achieving the same effect if we agree that a straight and steep profile for the curve of earned income vs disposable income is optimal to incentivize work.

Bruno-
I should have added, I merely thought of raising the personal allowance as a first aid measure, to bring down EMTRs quickly and noticeably for a large number of people. I fully agree on the general importance of low and constant EMTRs. Ideally in the form of a a single low rate, which would be at the same time the withdrawal rate of benefits below the personal allowance, and a flat tax above that. One problem with the present system is the variability of EMTRs. You could pick any random number between 0 and 100, and would almost certainly find enough people who would lose precisely that share of a hypothetical extra £1 earned.

Having said that, the loss of benefits for those who switch from not working at all to a low-skilled job would have to be quite steep, to make sure benefit entitlement does not extend too far up the income distribution.
Unfortunately, that could mean that for some people, the difference between a low-skilled job and not working at all would not be that large.
Here, I guess, conditionality comes into play. Living fully on benefits should simply not be a long-term option for healthy people.

Kris, Your last comment demonstrates exactly the dilemma of any system relying on means-tested benefits. Basic Income + Flat Tax is the only logical conclusion to that dilemma. Any other system is trying to achieve the same result through more bureaucratic means. Think of it this way. Taxation as a percentage of income is already applying means testing to the calculation of how much the government gives or takes. Why double up? If we want the relationship between Earned Income and Disposable Income to be a straight line with DI > 0 where EI = 0, it’s a simple y = ax + b formula, where a is 1 minus the tax-rate and b is the Basic Income. Anything else is just making things complicated.

Bruno – There are many advantages to a basic income, but I’m not sure how housing benefits can be incorporated. Currently, housing benefit is often the largest single component of claimants’ income and families in London may receive over £30,000 per year. How could workless households pay rents out of the basic income in the South-East (unless planning controls and building regulations were radically liberalised)?

Richard, My suggestion would be to incorporate a modest value for rent (£1,000 p.a. per person?) and pair that with an obligation on local authorities to provide minimum-standard accommodation to eligible people (those with a connection to the area, according to clearly-defined criteria). The accommodation standard would have to be low enough that people chose to get out if they possibly could. They would not be trapped, unlike housing benefit and social housing. The difference between rent and cost will be made up out of local taxes, and will depend on the cost of property in the area. That provides an incentive to local authorities to permit sufficient development of low-cost housing.

Richard is correct here, and this is precisely the reason for a separate housing allowance system. The issue, however, is largely restricted to London, in that rent differentials across the rest of the UK are much smaller. Dealing with this issue is not predominantly a problem of system design, but rather a matter of housing markets and the planning system. The best we might be able to achieve in terms of benefit reform is to move towards notional payments based on average market rents that therefore break the link between a tenant’s benefit entitlement and rent levels.

Bruno,
I’m not opposed to such a solution. But it would not automatically solve the basic dilemma. If b = £1000 per month and a = 1-0.33, then anyone with a gross income below £3000 would be a net transfer recipient. And from the perspective of the individual it makes no difference whether they receive a means-tested benefit that is withdrawn, or of a basic income that is taxed.
Richard rightfully mentions the incorporation of HB: If b was lowered to, say, £500, it would not even cover the rent in the South-East.
The only short term solution I see is to attach conditions to the receipt of b, lower it drastically in one step as the recipient enters work, but tax the remainder at a low rate.

I am entering this debate rather late and I have not been able to read everything fully. However, we should not necessarily assume that it is best to have a uniform withdrawal rate for benefits plus marginal tax rate right across the income span. There might be certain points in the income range where incentives matter relatively little. It is not clear to me, for example, that moving to a situation in old age where two thirds of pensioners are on a marginal ‘tax’ rate of (say) 60-80% is better than the system where a very small proportion had a 100% tax rate.

I cannot see any justification for higher levels of housing support in London or anywhere else. London differentials preserve the London bubble. If people want to live and work in London, let them pay the cost. That will include the cost to wealthier Londoners of paying for those who provide their services to live there or commute in. As we’ve all pointed out, these costs can be kept down by permitting more development. But I seem to be the only one who has suggested how the system could be constructed to give authorities (and their voters/taxpayers) an incentive to permit development. Saying it should be done is not enough.

Kris, 1. The basic dilemmas I thought we were talking about were (a) keeping METR as low as possible across the curve, and (b) avoiding bureaucratic inefficiencies. You can achieve (a) in other ways, but all at a greater cost of (b) and no additional benefit.2. It makes a big difference to the efficiency and intrusiveness of the system whether the individual receives a means-tested benefit or pays tax on his income and receives a basic income. For all but zero-earners, BI+FT can be handled through PAYE. I am not proposing that the basic income be taxed, by the way ( y = ax + b, not y = a(x+b) ). The difference is largely academic, but Occam’s razor should apply once more.

“The only short term solution I see is to attach conditions to the receipt of b, lower it drastically in one step as the recipient enters work, but tax the remainder at a low rate.”…which creates a massive disincentive to enter work. My priority (I believe it is the right one) is to minimize disincentives to work at all points along the curve (of Earned Income vs Disposable Income). It is abrupt or aggressive withdrawals that create the many poverty traps in the current system. There is no point at which creating a high hurdle to self-improvement is a price worth paying for looking tough.

Philip,(a) I think it’s unlikely that strong disincentives are irrelevant at any point on the curve,(b) If (a) exists for some individuals, I think it is unlikely that each person’s subjective evaluation is the same and that we can therefore identify a point at which strong disincentives are universally insignificant,(c) A Flat Tax has many efficiency-benefits, including the thorny issue of exchangeable allowances. My relationship status and partner’s income are irrelevant to the taxman under a BI+FT system. With multiple tax-rates, a single-earner couple pays more tax than a two-earner couple on identical combined income.Complexity. We have got to kill complexity.

Bruno, I agree with you about the situation in London – we all choose where we live, and even though it is a constrained choice that does not mean that we shouldn’t bear the consequences. After all this is what most working homeowners in London actually do. The problem, however, is to get a common acceptance of this position in conditions of an elective democracy. The ‘London problem’ has allowed serious reform to be put off for decades. This is because it is not an issue of system design but rather of a cultural shift about the role of the state and responsibility of individuals. Only once you have achieved this can you seriously start to design new systems of welfare.

Peter, To be brutal, London’s population is only around 12% of the total population. Even if we include the Home Counties as being in a similar situation, it is still a small minority. If those outside London understood a change to be both fair and in their interests (and some in London will also understand this point), it is not necessarily the case that a democratic mandate could not be obtained. It’s more a question of whether politicians had the stomach for the fight. The London bubble is a serious problem for the country, which, if not corrected by removal of distortions like this, will make the UK an ever less pleasant place to live, until the imbalances correct themselves painfully.

We ought also to be careful to compare apples and apples. The £30,000 extreme case that Richard mentioned would not be for a single person. I don’t know exactly what he has in mind, but that probably represents a larger household. We should be careful not to compare support per person with support per household.

Bruno – You are, of course, correct that the £30,000 case is for a three bedroom property. Nevertheless, I believe single claimants are entitled to around £20,000 per year housing benefit in some of the most expensive London boroughs. And even in the cheapest parts of London a one bedroom flat rents for about £7,000 per year. A single person could struggle to get cheaper social housing, unless he or she had some kind of disability. I completely agree that people who wish to live in London should bear the costs. Accordingly, the ‘local connection’ criteria that entitle families to be housed in a particularly borough should be reformed to encourage people to move to cheaper areas.

Richard, I agree with you regarding ending ‘local connection’ (which incidentally is to be strengthened under new rules announced recently). However, the major way into social housing, and what excludes most households, local or otherwise, is ‘priority need’. This too would need to go allowing for an open choice based system to be developed.More generally social housing is a hidden benefit which has similar disincentive effects to welfare benefits in terms of economic dependency.

Bruno,
as a third priority, next to simplicity and constantly low EMTRs, I would add that the circle of net recipients should not be too large. I see the merits of your proposal in terms of the first 2 points. But BI+FT could mean that a lot of people, who could easily be completely self-sufficient, would choose to live on a combination of BI and a few hours of extra work.
In-work benefits, with their contingency on a minimum number of hours, avoid this to some extent. They are better targeted towards those who do their best, but simply do not have the skills to earn more than the minimum wage. I’d rather support those than high-skilled people with a high preference for leisure.

Richard,I’d suggest that the “local connection” be retained in terms of responsibility and funding, but the local authority have some flexibility over the location where the actual housing is contracted/provided, so it need not necessarily be within the boundary. There’d have to be a limit, so Windsor could house people in Slough, but not in Skelmersdale. If you don’t have “local connection”, you will have a centralized system. In theory, there ought to be more pressure on people to get out and earn if they are supported by people they know and who know them. But I don’t have strong feelings on that point – it may be more theoretical than meaningful in modern “communities”.

Peter’s point about social housing being another poverty trap is important. It’s the 100% security of occupation and risk that this can never be regained once sacrificed (i.e. once you’ve moved out of social housing) that keeps people in these properties rather than risking getting out and looking for something better, work- and accommodation-wise. That’s why a system that you could always fall back on, but whose cost to the occupier could instead be used to contribute towards something better, would reduce the disincentive effects.

Kris, The radius of the circle of net recipients will be determined by the level of support for zero-earners and the combined effect of the tax rates and withdrawal rates of benefits. Whether a conventional, benefits-based system or BI+FT, you are still stuck with the unavoidable choice between high METR and lower numbers within the circle, or lower METR and higher numbers within the circle. The system only affects the bureaucratic overhead. To me, it is clear that high METR is a greater disincentive than low or negative effective tax rates. Our current system is based on exactly the fear that you express. And it doesn’t work.

The choice is whether it is better for people to work for however many hours of work that they can find and manage, or whether it is better for people to be unemployed unless they work a minimum number of hours. For me, that isn’t a difficult choice. As taxpayers, we are better off if the unemployed work a few hours than none. And the unemployed are better off. A few hours’ work may lead to more work, and more ambition.Yours is a pessimistic view of human nature – that people will scrounge unless driven to work. Mine is optimistic – that people will take the opportunities they are capable of if obstacles are removed. The pessimistic view has created the mess we’re in. Targeting doesn’t work

Incidentally, Kris, I was looking at some of your old posts last night, and I see that you agree with Mises and me that targeting doesn’t work. http://www.kristian-niemietz.net/wordpress2/?page_id=82

“Yours is a pessimistic view of human nature”
Not at all. I’m merely saying that people respond to incentives, and as far as that relates to EMTRs, we are obviously on the same page. But EMTRs are not the only incentive variable. EMTRs affect the substitution effect, but what about the income effect?
Under your proposal, the government would start handing out substantial lump-sum transfers to all sorts of people who simply do not need them. Suppose someone earns £1,400pm net – he’s not rich but clearly not ‘in need’. Suddenly, he’s given a BI of, say, £700pm. One option would then be to scale back the work week to three days, and have more leisure AND a higher income than before.

“the circle of net recipients will be determined by the level of support for zero-earners”
There’s a limit to how low you can go. The BI would, among others, be paid out to people who genuinely cannot work and who would stay on BI for a long time. I guess you would not want them to live on a mere subsistence level (ignoring the potential role for private charity, which is a different chapter). But I guess you don’t want a host of complex BI top-ups and premiums either.
Hence, I rather go for a BI (or whatever it’s called) which is not too low but which comes with strings attached for those who could work. High EMTRs would only apply over a very narrow range.

Kris, You are forgetting that BI goes with Flat Tax. You haven’t described his domestic arrangements, which are critical to the current system, but let’s assume he is single. His annual gross pay is £16,800. Currently he pays £2,065 of income tax and £1,219 of employees’ NI. He would be eligible for Working Tax Credit of £477, and quite possibly some other benefits too. That gives him a disposable income of around £14,000 against an earned income of £16,800, making an effective tax-rate of around 17%.

Sorry, just noticed you said £1400pm _net_ (£16,800 p.a.). So that’s around £21,500 p.a. gross earnings under the current system (no WTC at that level) – around 22% effective tax-rate.I am not proposing £700pm because that is more generous than the current system and unaffordable. But let’s take it for the sake of example. I will need to set a high FT to match – let’s say 58%, though it’s probably higher for that BI. So he has £12,470 of FT to set against £8,400 of BI, giving a disposable income under the equivalent BI+FT system of £17,430 – around 19% effective tax-rate. It’s not very different. I say again, the determinant is the profile, not the system. The system determines efficiency.

Kristian assumes the objective is to allow individuals to make sensible economic choices, based on their own circumstances and preferences; hence he recommends simplicity and transparency in the benefit system.But we are not politicians. What if the real objective is to make everyone dependent on the state? Then there is much to be said for complexity and obliquity. Moreover, with that different objective, producing a system that ‘traps’ people in poverty is not a disaster but a triumph!

Some excellent points here, but the crux to benefit reform is the point made at the start of the last paragraph, namely conditionality. The key problem of the UK benefit system is that it places no expectations or responsibilities onto recipients, but rather encourages them to maximise their ‘entitlements’. The result is passivity and the absurdity of over 5 million people of working age outside the labour force during the last economic boom, with most new jobs being taken by migrants from the accession countries.

As well as simplification, I think there may be a case for absolute reductions in certain benefits in order to improve work incentives. Many workless families receive a similar amount to an employee on the median wage, once housing costs, travel and taxes are taken into account.

If one wants to reduce the extent of state interference (and one does!), surely there must be a more significant role for voluntary charities. But it’s not easy to see how one could make the transition from where we are to where we might want to be. Abolishing the Charity Commission might be a useful start.

Kris, This sentence – “Even a very simple system can be hugely wasteful, by redistributing large amounts among the middle classes instead of targeting the poorest” – is inconsistent with the rest of your well-argued post. Means-tested withdrawal is the main reason why benefits create a strong disincentive to work, by creating high marginal effective tax rates.

Bruno,
that’s true for the present situation, but I guess that could be reformed. At present, the withdrawal rate for working tax credits is 39% – that’s a fairly high effective marginal tax rate, but not excessive. It’s the interaction with the tax and NI system that raises EMTRs to 70% and beyond. Raising the personal allowance would eliminate such overlaps.
The withdrawal rates could be lowered further if the intital benefit level is lowered alongside, and conditions are attached (e. g. time limits).
But there remains a conflict between targeting and incentive-compatibility.

Kris, Mechanistic solutions, whether Basic Income + Flat Tax (BI+FT) as I favour, or increased personal allowances as you prefer, are not the real determinant of the level of marginal effective (ME) taxation. BI+FT could produce high ME tax rates, simply by setting the FT at a high rate. So could increased personal allowances, by setting the rates of withdrawal of benefits below the allowance and the rate of tax above it quite high. What really determines the impact of our proposals is that we recognize the importance of ME tax rates, and therefore propose that the systems be structured to produce a profile of disposable income vs earned income that keeps ME tax rates low.

But we must also acknowledge that, unless we are prepared to reduce the level of support provided to people with zero earned income, this change in the profile is achieved by increasing the disposable income of low-to-middle-earners relative to their earned income. And while that is undoubtedly a good thing, we have to recognise that a side effect is to substantially increase the net cost to government of income taxes and welfare provision. We may believe that in the long-term, the Laffer Effect will close the gap, but in the short-term, we have an obligation to set out how we would increase other taxes, cut spending or increase borrowing further to compensate for this effect.

That is not to deny the value of the mechanistic solutions. The virtue of both your and my proposal is that they simplify the system substantially, which has benefits in reduced bureaucratic costs, and for the ability of people caught in the welfare net to focus on getting out of it rather than on understanding the system to maximise their income from it. In my opinion, the benefits of BI+FT in this regard exceed the benefits of any system that retains means-tested benefits, whilst achieving the same effect if we agree that a straight and steep profile for the curve of earned income vs disposable income is optimal to incentivize work.

Bruno-
I should have added, I merely thought of raising the personal allowance as a first aid measure, to bring down EMTRs quickly and noticeably for a large number of people. I fully agree on the general importance of low and constant EMTRs. Ideally in the form of a a single low rate, which would be at the same time the withdrawal rate of benefits below the personal allowance, and a flat tax above that. One problem with the present system is the variability of EMTRs. You could pick any random number between 0 and 100, and would almost certainly find enough people who would lose precisely that share of a hypothetical extra £1 earned.

Having said that, the loss of benefits for those who switch from not working at all to a low-skilled job would have to be quite steep, to make sure benefit entitlement does not extend too far up the income distribution.
Unfortunately, that could mean that for some people, the difference between a low-skilled job and not working at all would not be that large.
Here, I guess, conditionality comes into play. Living fully on benefits should simply not be a long-term option for healthy people.

Kris, Your last comment demonstrates exactly the dilemma of any system relying on means-tested benefits. Basic Income + Flat Tax is the only logical conclusion to that dilemma. Any other system is trying to achieve the same result through more bureaucratic means. Think of it this way. Taxation as a percentage of income is already applying means testing to the calculation of how much the government gives or takes. Why double up? If we want the relationship between Earned Income and Disposable Income to be a straight line with DI > 0 where EI = 0, it’s a simple y = ax + b formula, where a is 1 minus the tax-rate and b is the Basic Income. Anything else is just making things complicated.

Bruno – There are many advantages to a basic income, but I’m not sure how housing benefits can be incorporated. Currently, housing benefit is often the largest single component of claimants’ income and families in London may receive over £30,000 per year. How could workless households pay rents out of the basic income in the South-East (unless planning controls and building regulations were radically liberalised)?

Richard, My suggestion would be to incorporate a modest value for rent (£1,000 p.a. per person?) and pair that with an obligation on local authorities to provide minimum-standard accommodation to eligible people (those with a connection to the area, according to clearly-defined criteria). The accommodation standard would have to be low enough that people chose to get out if they possibly could. They would not be trapped, unlike housing benefit and social housing. The difference between rent and cost will be made up out of local taxes, and will depend on the cost of property in the area. That provides an incentive to local authorities to permit sufficient development of low-cost housing.

Post new comment

The content of this field is kept private and will not be shown publicly.
Type the characters you see in this picture. (verify using audio)
Type the characters you see in the picture above; if you can't read them, submit the form and a new image will be generated. Not case sensitive.