Interpretation of terms used in relation to targets

I beg to move amendment 24, in clause 6, page 3, line 13, at end insert—

‘(f) The publication of analysis of the sensitivity of the estimates of child poverty under sections 2, 3, 4 and 5 to the methodological assumptions which underlie those estimates.’.

With this it will be convenient to discuss the following: amendment 25, in clause 6, page 3, line 13, at end insert—

‘(f) How poverty among children who are not included in household surveys is to be assessed.’.

Amendment 26, in clause 6, page 3, line 13, at end insert—

‘(f) How the statistical surveys used to calculate poverty rates are to be re-weighted to reflect under-reporting of specified groups.’.

I apologise to the Committee if it might be thinking that we are entering what might be regarded as a slightly “anoraky” phase in our discussions, but what the heck? The three amendments are all fairly self-explanatory, so I will not dwell on them at length. They raise three separate issues. Clearly, as amendment 24 points out, we have four different definitions of poverty, but each of them involves some very complex assumptions. For example, let us consider our discussions this morning about the 21 things it takes not to be materially deprived, and recognise that they are all given a weight. We then draw a line of 25 against that, and people are poor if they are above it, but not if they are below it. The fact that that has to be combined by 70 per cent. of median income on a certain equivalent scale and a certain definition of treatment for housing shows us that the whole thing is built on a mass of assumptions and methodological decisions. I do not doubt for a second that the statisticians who put such things together do so in good faith, but they will know far better than I do that the assumptions we make determine the number we get at the end.

Some of the assumptions might be entirely marginal and mild, but others might have a huge impact on the numbers that result. Such action therefore ends up not being a statistical point, but a real one. If we choose a definition that gives us half a million more children in poverty than another definition, the price tag attached to that probably runs into billions of pounds. We therefore have to be confident that the number we are looking at is not a figure within a very broad range that could have  come out differently for a couple of assumptions that might be entirely reasonable. We do not necessarily want to take such action each year, but for the base year figures, we might need to see more evidence of the impact of the key assumptions on the numbers.

To give one example, as the Minister said, the statistics can be adjusted for household size in several different ways. The Department has traditionally used the McClements equivalence scale and now it proposes to use the OECD scale. To give the Government credit, they have thereby increased the child poverty numbers by a few hundred thousand. So they have not done this out of malign intent; they have made their job more difficult by that assumption. I only know that because there is a table in one of the documents to show the impact of that assumption.

That is just one of a very large number of assumptions. It would help if we knew where the numbers are coming from, how sensitive they are to the assumptions that have been made and how much of a mountain we are being asked to climb because of a particular set of assumptions. We need to know which are key and how much impact they have had. Amendment 24 simply requires the publication of this sort of sensitivity analysis. To some extent this sort of thing goes on anyway. The “Household Below Average Income” figures used to do a bit of this from time to time. Given the pivotal importance of these statistical measures for the future, some initial grounding for these measures would help our scrutiny.

Amendment 25 came up in our evidence sessions and we made some progress on that issue, but it is worth remembering that these figures for children in poverty are based on surveys of private households. People go out, knock on doors and sit in people’s front rooms with a laptop. That is essentially how these data are collected—or it certainly was when I went out with one of these surveyors, although not in the emotional sense. We sat in someone’s front room. There are an awful lot of children who will not be identified through that method.

We have talked about children in local authority care. The Minister’s response was, “Yes, but they are not poor.” The local authority ensures that they are not poor, but they may be deprived. If we are only interested in income poverty, fair enough, but if we are interested in some slightly broader concept of child welfare—without going all the way through our earlier discussions—we need to know what is going on with those children. Some of them will be permanently in care, but many will be transitorily in care and coming from poor households and going to poor households. How will they be picked up? The briefings we have had mention children in the criminal justice system, young offenders institutions and all the rest of it. Again, at that point they are being fed, clothed and housed so they are not poor, but if we are interested in children thriving and prospering, they are being missed out.

Some of the children of Gypsy and Traveller families, in theory, could end up in these surveys. I am thinking of Traveller sites in my constituency that are run by the local authority and that would count as private households  in the sampling frame for the surveys. I am not sure—they have a postcode so that is probably all it takes. We are not excluding all Gypsy and Traveller children, but we are excluding those who live the most transient lifestyles, and we know that Gypsy and Traveller parents have low life expectancies and various other social problems. That would be true of their children too.

There is an issue about being exhaustive regarding who is missing. Let us be clear which of the nation’s children we are not including, and which of those we might have some concerns about. Is there any way we can either graft them into the statistics or have some auxiliary reporting, so we do not forget them? The risk is that we end up with the old target problem we talked about earlier: most of the nation’s children are in the figures, so those are the ones we focus on; there is no point helping Gypsy children because they are not in the figures. I do not suppose that any Government would be callous enough to do that explicitly, but if they are under legal pressure to hit a target that includes one set of children and omits another, we can see how that could happen.

Amendment 26 is about the fact that all these statistics are based on sample surveys. We are grossing up from perhaps 25,000 households in a survey to a population of some 25 million households. If a certain sort of household is less likely to respond to a survey than another, something has to be done. Who is less likely to respond? People in urban areas are less likely to respond, as are members of ethnic minorities, people who do not have English as a first language, the very elderly and so on. Clearly, we need to correct for those biases. To be fair to the Government, they do try to do that. They re-weight the survey so that if, for example, there is one very elderly person from an ethnic minority living in urban London, they get a bigger weight when the statistics are created.

However, I am not convinced that the grossing process—the re-weighting process—picks up adequately the under-reporting of children who are most likely to be in poverty. The serious point is that the current poverty numbers might be too low because some of the things not adequately weighted for are correlated with higher risk of child poverty. For example, I believe there is a regional weighting but not an inner-London weighting. I am open to correction but that is my understanding. If one thinks child poverty is higher in inner London but the weighting is only for London as a whole—with Twickenham gaining as much weight as Southwark, for example—one does not get the right figure for child poverty. I wonder what analysis the Minister has undertaken of the risk of not responding to a survey, and the correlation between that risk and the risk of child poverty.

I have one further question before I draw my remarks to a close. The “Households Below Average Income” statistical appendix describes the way in which the sample is weighted to achieve the population total. It says:

“In order to reconcile control variables at different levels”— these are the true totals—

“and estimate their joint population”— an example would be old people living in the north for whom English is a second language; in other words, a combination of such factors—

“software...provided by the French National Statistics Institute has been used. This software works by iterating towards an optimal solution that, given the particular control totals, minimises the range...of the grossing factors chosen.”

I am sure that that is true, but does it worry the Minister that in a sense, one ends up forcing the sample one has to match a whole grid of national population characteristics through a quite complex statistical forcing mechanism that tries to get the numbers right, but which may not represent what is going on in the world? In other words, one can have solutions that give very big weights to a particular household in a sample because it gets the numbers right—it gets the totals and the cross-tabs right—but one ends up with some odd results by giving hefty weights to particular sample members and low weights to others. It makes me uneasy that something we are going to spend tens of millions of pounds of public money on and debate at great length is based on an obscure statistical process that I am not sure has ever been demonstrated not to distort the data a bit too much. It gets the totals right and we are all happy about that, but do we know how robust the process is? That is my concern. I think the Minister has a higher mathematics degree, so he knows exactly what I am talking about and I hope he will run the Committee through the process.

All that probably sounds a bit technical but if we do not get these things right, they will have very big, real implications. Yes, it is real people we are talking about, but if we get the numbers a few hundred thousand out one way or another, the policy implications could be very serious. We need to have the information to hand to appraise the statistics we are being given, and that is the purpose of the amendments.

It is not the first time this afternoon that the background of the hon. Member for Northavon as an academic in the social sciences has come to the fore to the benefit of the Committee. The hon. Gentleman, with the three amendments, is seeking clarity and transparency so that we know what we are talking about, so that these things are not hidden away in footnotes in HBAI tables, which are perhaps not readily available and understood. The amendments call for

“an analysis of the sensitivity...to the methodological assumptions” and request information about how poverty is to be assessed and how under-reporting in response to surveys is to be dealt with. If the hon. Gentleman is looking to press the amendments to a vote, I would be minded to ask my colleagues to support him because I think clarity and transparency are to be welcomed, so that we know exactly the basis on which we are proceeding.

There is one group that the hon. Gentleman has not mentioned. He has mentioned children in local authority care, as well as Gypsies and Travellers, on whom we had some discussion, but I am not entirely clear about the position of the children of asylum seekers, or whether they are included. In the northern part of Bedfordshire, not in my constituency, we have the Yarl’s Wood detention centre. Will that be surveyed, and will all this be part of it? We need to be absolutely clear about which children we are and are not talking about. We might have a legitimate argument about whether some children should not be included for a particular reason, but I would be grateful if the Minister gave us some clarity about what the targets are based on.

I want to ask a little about qualifying households, net household incomes and deductions. I share others’ concerns about what happens to Gypsy and Traveller children, the children of asylum seekers and children whose immigration status is unclear. There is a clear distinction between vulnerability and poverty. The two might overlap, but this is about poverty. Children in the care system who have issues of vulnerability will come under the scope of different legislation and might not necessarily be the same as children who live in poverty. I would therefore like us to focus on children who live in the community rather than children who live in institutions. Somewhere such as Yarl’s Wood might not be quite the place to look for such children, because it is a residential, institutional setting, where they will be for only a certain amount of time. I am more concerned about children whose asylum claim has been turned down, for whom there can be some to-ing and fro-ing. As we all know, such cases can go on for years, at the end of which the children’s status might still be unresolved. If their parent cannot get benefits and cannot work, their poverty is acute and their housing is often pretty atrocious.

The same applies to Traveller children who might be centred around one address but who might travel around quite a lot and therefore do not come under the scope of any particular survey. The hon. Member for Northavon has made this point. I would not support the amendment because such children probably could not be included under any survey, partly because they are not static for long enough—especially Traveller children, whose families may still be quite mobile. Also, such children might not be very willing to come forward, especially if they are without immigration status, and it can be very hard to pin them down. I suspect that virtually 100 per cent. of children in those groups will be living in poverty, and I should like to know what other measures my right hon. Friend the Minister is taking to deal with poverty in those groups.

Finally, on net incomes, if one looks at poverty figures after housing costs are taken out, they are much more acute because of problems with housing costs. What is the Minister’s thinking about ensuring that housing costs do not push families into complete poverty?

The hon. Member for Northavon has made some important points through his amendments, and I am sure that he is right that certain groups will not be included in any household surveys. I am particularly concerned about the two groups that my hon. Friend the Member for Northampton, North has just mentioned—Gypsy and Traveller children and asylum-seeking children. I know that it is extremely difficult to get information about Gypsy and Traveller households, but it is very important that they are taken account of in such surveys and calculations. As the chair of the all-party parliamentary group on Gypsy and Traveller law reform, I have had extensive meetings with the census people about getting Gypsies and Travellers registered on the census. There is a huge effort going on to try and get those people registered on the census, by working in liaison with Gypsies and Travellers, trying to identify caravans, and through word of mouth in different areas. I hope that we will get a certain number of Gypsy and Traveller families on to the census, but it is extremely unlikely that they will ever show up in regular household  surveys. Gypsy and Traveller children are some of the most deprived children in the land. There is not much financial information about them, but we know that they lose out in lists of material deprivation, particularly if they are moving around and not on local authority sites.

When the Minister responds, will he state what particular efforts will be made to include Gypsy and Traveller children in the Bill? We know that the income received by asylum-seeking families is below what is needed to live in the way that the rest of the country lives, and we know that those children are particularly vulnerable. I would be interested to hear what measures will be taken to help that group of children.

Clause 6 enables a number of technical terms underpinning the child poverty targets in the previous clauses to be defined in regulations. I believe that a draft of those regulations was provided to members of the Committee last week, and they are subject to the affirmative procedure. I will follow the hon. Member for Northavon and speak to each amendment in turn.

Amendment 24 would add a discretionary regulation-making power to clause 6, requiring analysis of the sensitivity of child poverty statistics under the four targets in the Bill, and for the methodological assumptions underlying those statistics to be set out in the regulations. There are already provisions in the Bill that ensure that the underlying methodology used to calculate those statistics is robust and in line with accepted statistical practice. Clause 14(2) requires that the statistics used to measure progress on the four targets are national statistics. That means that they must adhere to the code of practice produced by the Statistics Board. That code states that producers will endeavour to ensure that methods used to produce national statistics are objectively chosen and based on sound statistical methods. It also states that quality measures must be regularly published to enable users to assess whether the statistics are fit for purpose.

The “Households below average income” series, where our child poverty figures are currently published, already includes information documenting the underlying assumptions. It presents statistics for a range of child poverty thresholds, both before and after housing costs have been deducted, and the figures on both are published. It provides a comparison with the modified OECD and McClements equivalence scale that was mentioned, which is used to adjust incomes for household size and composition. It also gives the confidence intervals for the number of children in relatively low-income households, which provides an indication of the measure’s precision. It might be helpful for me to make it clear that we expect all this information to be included in future editions of the publication.

Amendment 25 enables the Secretary of State to make regulations setting out how poverty should be measured for children who are not covered by the surveys, and I have listened with interest to the points raised by a number of hon. Members on that matter. The amendment would define ways in which the poverty of those children could be assessed, but it would not  explicitly require poverty levels to be monitored or reduced. I assume that the broad purpose of the amendment is to ensure that children are adequately covered by provisions in the Bill, and that concern is shared across the Committee.

The targets that we are setting will be effective only if progress towards them is measurable. They apply to children in qualifying households, which are those households included in surveys. Of course, it would be great if we could include every child, but practically speaking, my hon. Friends the Members for Northampton, North and for Cardiff, North are right. There is no way around the fact that some households will not turn up in surveys for the reasons indicated.

We have been careful to frame the Bill so that it does not exclude those children. Clause 8(2) sets out that the UK strategy needs to do two things: comply with

“the duty to ensure that the targets are met”; but do so also

“for the purpose of ensuring as far as possible that children in the United Kingdom do not experience socio-economic disadvantage.”

That obligation on the strategy will ensure that the children we have talked about, who may well be missed out from the survey evidence, will nevertheless be beneficiaries of the Bill, because all children are covered by that provision, regardless of whether they are picked up by the surveys.

Will the Minister write a note to all members of the Committee—as has helpfully been done on two occasions so far this week, which I appreciate—in respect of non-resident parents? With parents who are separated, children are perhaps spending a considerable amount of shared care time with, typically, their father, who may have a very different income. My suspicion is that the surveys are primarily, if not overwhelmingly, picking up the households in which the parent with care lives. If there is a significant degree of shared care, the surveys will not be capturing what is happening to those children for the rest of the time. The issue is probably a technical one, but a note to the Committee would be helpful.

I am happy to do that. The survey works through the random selection of postcodes, I think, so there will be no bias towards the parent with care rather than the other parent, but I am happy to provide such a note.

The main groups of children that we are talking about are Gypsy, Roma or Traveller children, children in care, children in bed-and-breakfast accommodation and asylum-seeking children. As I said, if the household does not have an address or postcode, I do not think that the surveys will pick such children up.

For asylum-seeking children, the great majority will be in households with an address or postcode, so I would expect the great majority to be included, but there is some uncertainty about precisely how many children we are talking about. A Barnardo’s report published last year estimated that since 2002 around 40,000 children in asylum-seeking families had arrived or been born in the UK and that there could be more than 100,000 children in total in the UK asylum system. We are certainly talking about quite a large group of children there, but the great majority will be in private  households and will be picked up by the survey. Children in Yarl’s Wood, to take the example given by the hon. Member for South-West Bedfordshire, would not be picked up by the survey. For children not living in private households—children in Yarl’s Wood, for example, but children in care homes as well—there is not really a way to allocate household income, even if they were covered by the survey. Again, I do not think that we shall find a way of including them in the target, but they will be covered by clause 8(2)(b).

Development of the child poverty strategy will involve identification of those groups of children most at risk of being in poverty, including harder-to-reach and disadvantaged groups, and measures to help them. As we discussed in the evidence sessions last week, the well-being of many children not covered by the targets is monitored by Government and local government in other ways. For example, there are outcome and well- being measures for looked-after children who live in accommodation not covered by the surveys; there are minimum statutory standards for looked-after children in residential care; and there is a Government target to reduce the number of households living in temporary accommodation, including bed-and-breakfast accommodation. Targets are in place for a number of such groups.

Amendment 26 is designed to ensure that the weighting procedures for the survey statistics take into account under-representation in the survey of particular groups. I sympathise with the intention behind the amendment, but I hope that I can persuade the hon. Member for Northavon that it is not necessary. The main survey used to calculate child poverty rates, the family resources survey, is subject to weighting procedures, which help to ensure that the survey data accurately reflect the make-up of the population. The data are weighted on a range of variables, including housing tenure and family type. The weighting regime for the survey was reviewed comprehensively in 2005 and the results are on the Department for Work and Pensions website. In addition, the methodology used to create the survey statistics has to adhere to national statistics standards for integrity, quality and freedom from political influence. That means that regular independent quality reviews of the statistics will be carried out.

The family resources survey weighting procedures do not account directly for the possible under-representation of all at-risk groups. We correct indirectly for some of the under-representation of at-risk groups by taking account of characteristics such as household tenure and family type, which are statistically related to other characteristics of at-risk groups. It is not possible to control for all at-risk characteristics. First, it is best practice to try to limit the number of control factors that underlie the weighting procedure, because too many control totals can produce misleading results. Secondly, if few households from a given group are captured by the survey because there are relatively few such households in the country, weighting procedures will be unreliable. Thirdly, we do not know enough about the populations of some groups, such as Gypsy, Traveller and Roma people, to use weighting procedures reliably, because there is no robust total to be controlled to.

I hope that I have assured the Committee that sufficient, accurate weighting procedures are in place and that they are regularly reviewed. Amendment 26 is therefore unnecessary.

To respond to my hon. Friend the Member for Northampton, North, the use of income before, rather than after, the deduction of housing costs is the subject of our next debate. The main argument for using income before housing costs is to enable straightforward comparisons with other European countries, because that is how they measure it. We had an interesting discussion this morning on how we stand compared with other EU countries. We would not be able to make that comparison if we did not use income before housing costs data. However, I emphasise that we will continue to publish both measures so that everybody can see exactly what the position is.

I welcome that response. I accept that there is shed loads of sensitivity stuff relating to the 60 per cent. of median income information at the back of the HBAI statistics. I am less confident that we know how robust the newer statistics are, such as those on persistent poverty and material deprivation. A huge amount of effort goes into the statistics that we have had for a long time. I seek reassurance that we will believe the newer statistics to be more robust when they are published more regularly. That is currently less pronounced.

I take the Minister’s point about children not in the surveys. It is welcome that the strategies are not just about the children covered by the targets. That is reassuring, provided that it has teeth. If a hypothetical Government cut benefits for asylum seekers and the welfare of children in those households diminished, it would come under the section that he referred to. If the child poverty commission has no teeth, life will go on.

If that happened, it would be picked up by the survey because the great majority of those children are in households and will be covered like other children. It is only children in Yarl’s Wood, for example, who would not be picked up.

That is fair. It is helpful that the Bill is not solely concerned with the target.

Finally, on under-representation, I take the Minister’s point that we do not have control totals for many of the groups. The point about inner London is probably quite important. Governments notoriously cannot count how many people live in such areas and I know that it is not straightforward to do so. Given the correlation between child poverty and living in inner London, the fact that the weighting is for Greater London, not for inner and outer London, means that the headline child poverty numbers are probably too low, albeit perhaps not by much—I do not know by how much. Will the Minister look at that again?

I appreciate the seriousness with which the Minister addressed those concerns, so I beg to ask leave to withdraw the amendment.

Amendment, by leave, withdrawn.

I beg to move amendment 28, in clause 6, page 3, line 14, leave out subsection (2).

An audible cheer went up, Mr. Key, when we reached this substantive and important amendment. Under clause 6, the definition of “income” may include and exclude all sorts of things, but the one thing that cannot be deducted is the cost of housing. That is explicitly prohibited by the Bill. Amendment 28 would remove that prohibition and allow the child poverty commission, when drawing up the measure of poverty to target, to consider taking housing costs off. That is all it would do. It would not even oblige the commission to do so; it would just let it think about doing so.

What difference would that make? According to the latest figures, on the before-housing costs measure, 2.9 million children are living in poverty, and on the after-housing costs measure, 4 million are. The difference is huge, and the amendment goes to the nub of the issue. There are arguments for and against, and they have, conveniently, been published at the back of the document on households on lower incomes. The argument for not taking off housing costs is that if someone spends more on housing because they buy a plusher, nicer, more luxurious place, that is their choice and the cost should not be deducted because it makes their life better. If housing costs are just a proxy for quality, we should not knock them off; however, they are a real measure of need that cannot be avoided and are different in different parts of the country, so they should be knocked off.

The Minister generously told us what his answer would be when he said during the previous debate that the key argument for using the before-housing costs measure is that that is what everyone else does, so we can make international comparisons. However, no one said that we should not publish before-housing costs measures; no one said that we should not be able to compare internationally by continuing to generate before-housing costs numbers. Surely the Bill should be about the best interests of British children and should be the right target for the welfare of children in this country. The ability to make comparisons with other countries should be a secondary consideration, but we can do that anyway. Using an after-housing costs measure does not preclude international comparisons because we will calculate the numbers anyway—all that is involved is two sets of figures, one of which has housing costs knocked off. There is not even any extra effort involved.

The question is: which is the best measure of real living standards? There is a strong London dimension, as the Minister will appreciate, because average housing costs in the capital are substantially greater than elsewhere. Most people would not think that that is because of quality. People who pay high rents for overcrowded rented accommodation in central London are not having a whale of a time; they just have to find the money. Deducting the cost of housing seems to be particularly appropriate in areas where child poverty might be concentrated—for example, urban areas.

A further reason why the before-housing costs measure is not the right one to use is that, perversely, it includes housing benefit. If I lived in the same house, year after year, and rents, whether council, housing association or  private, rose faster than inflation, I would be better off on the measure that the Bill proposes to use because my housing benefit would go up. How can that be right?

I hesitate to mention that a third measure is knocking around, which I have just remembered. The Johnson-Webb measure was proposed in an economist’s academic paper a while ago. It suggested that an alternative is to exclude both housing benefit and housing costs. If including housing benefit in income is a problem because it goes up when rent goes up, which makes no sense, and knocking off housing costs is not perfect because it might reflect quality, one could use a statistic that includes neither, and perhaps bring in housing in another way, perhaps in material deprivation.

There are variations, but the key point is that both sets of figures—before-housing costs and after-housing costs—will be published. That has been the case for years, and should continue. We are not precluding putting that information in the public domain, but what should future Governments be targeting?

The hon. Gentleman makes an extremely powerful case. May I take him back to his point about international comparisons? Is it true that because of the shortage of houses and problems with housing in this country—I will not rehash this morning’s debate—we have a particular problem with the cost of housing for those on low incomes?

Certainly, the British housing market is very different from everywhere else in western Europe. Housing impacts on people’s living standards in this country in a very particular way. It is entirely legitimate to look across Europe on a harmonised definition—I have no problem with that—but if we want to know what is happening to people’s real living standards, it does not seem right simply to ignore their housing. I have said all that I need to say on that point, and I hope that the Committee will support the amendment.

The Minister gave us an early glimpse of what he was going to say in his response to the previous group of amendments. I recognise that we need to have a measure of poverty that we can use to draw comparisons with other EU countries. Indeed, we had that debate this morning, which was necessary. The hon. Member for Northavon and my hon. Friend the Member for Beverley and Holderness are right to point out that the after-housing costs measure is probably what families focus on. They want to know how much they have to live on to cover food, clothing, heat and so on once they have paid for their accommodation. I accept what the hon. Member for Northavon says in that respect. We know that both sets of figures will be published. They will be available and child poverty lobby groups will draw attention to both sets, and particularly to the after-housing costs set.

I will listen with interest to what the Minister says. The hon. Member for Northavon makes a valid point when he says that his amendment is permissive rather than prescriptive. It would not require anything to be done; it would merely remove a prohibition from clause 6. On that basis, I think that the case for his amendment is persuasive, but I will listen to hear if the Minister has any especially good arguments up his sleeve for why we should leave the clause as it is.

Unfortunately, my briefing does not cover the Webb-Johnson methodology so, sadly, it will be neglected in my response.

The Committee has heard what the amendment would do. We have used a before-housing costs measure of child poverty since the consultation on measuring child poverty in 2003 but, as I have already emphasised, the households below average income data series continues to present child poverty estimates using both before and after-housing costs measures of poverty, and our intention is for that to continue.

I have mentioned one of my reasons to the Committee, but there is another. The most telling consideration is that income is measured before housing costs to allow comparisons with other European countries, which measure poverty in that way. There is a dynamic in the Bill that we aim to be among the best in Europe. The target that we have set for relative poverty would place us as the best in Europe at the moment. In setting the targets—clause 6 is about interpretation of terms used in relation to targets—we should use the before-housing cost measure to enable those comparisons to be made and for us to have a sense of whether or not we are achieving a position of being the best in Europe. The hon. Member for Northavon said that we could still look at that data. However, I turn that round. We will still be able to see the after-housing cost data and make whatever use of them we wish, but in terms of the target, it is right that we compile the figures in a way that can be compared with others.

As we all know from this morning’s debate, measures of housing quality are included in the combined low income and material deprivation measure. If a child is experiencing poor housing, that will be reflected in their material deprivation score. Families who cannot afford things because they have high housing costs will be picked up in the material deprivation measure as well. The measure reflects the impact of high housing costs and a family’s ability to afford decent housing. The hon. Member for Northavon is right about the impact of high housing costs on child poverty—my hon. Friend the Member for Regent's Park and Kensington, North frequently makes that point. Through the low income and material deprivation measures, we are able to track that impact because paying out a lot of money for housing curtails the ability to afford other things.

Finally, as the hon. Member for Northavon mentioned, if one deducts housing costs entirely, those who choose to pay for better quality accommodation could have their relative standard of living understated. In preparing a UK child poverty strategy, we need to consider any housing measures required to tackle child poverty, as set out in clause 8. It is right that the targets are defined as being before housing costs, for the reasons that I have set out, but people will continue to monitor what is happening on an after-housing costs basis and the data will continue to be provided to make that possible.

I am not sure that I am convinced by that. The Minister seems to be saying, “Well, don’t worry about knocking off housing costs, because we’ll pick it up some other way. We’ll pick it up through deprivation, so if people have high housing costs, the children won’t be able to go on holiday or have a new pair of shoes.” That is a pretty indirect way of picking it up, and you would not be able to disentangle the  housing impact from another impact. The family might not be able to go on holiday because of other costs, so that would not identify the impact of housing uniquely. I take the Minister’s point that the after-housing costs numbers will still be published, but targets matter—as his motto would be. If one targets income before housing costs, there is not the same policy impetus to tackle the housing costs of low-income households, as the hon. Member for South-West Bedfordshire pointed out. If income after housing costs is targeted, tackling such housing costs becomes a much more explicit policy objective because the Government will be held to account for it. What we target matters.

I am also concerned that the Minister wants to pre-empt the child poverty commission—he has great faith in it and is going to stuff it with experts—from even thinking about housing. I cannot see why he would want to stop the commission from at least looking at whether, on balance, taking off housing costs gives a better guide.

I apologise for interrupting the hon. Gentleman’s flow. Were his amendment to be accepted, what would be his intention for the targets in clauses 2 to 5? What would be the practical effect of the amendment? He talked about it having a permissive character—by allowing the child poverty commission to do different things—but what does he intend to change in the targets in clauses 2 to 5.

If we are going to create a child poverty commission, the commission should be asked. Clause 6 is about the Secretary of State making regulations, presumably on the basis of expert advice. Therefore, the child poverty commission, when deciding all the details, should be asked whether deducting housing costs is the right thing to do. It would then either reaffirm the exact targets as they stand just after housing costs, or come up with different thresholds on the basis that a slightly different thing is being measured. I have a view on what the right answer is, but that is just my opinion. There is a strong argument that what is being precluded is worth looking at, so we should not pre-empt the commission and stop it from looking. I am not saying that it should look at everything, but housing costs are so fundamental and central to people’s real living standards that the commission should be able to assess whether they should be in the target.

May I press the hon. Gentleman a little on his point about constraining the commission? The clause specifies what can be in the targets. The child poverty commission might want to look at a range of matters, but it is right that the set target should be a before-housing costs target to make comparisons with other countries.

We do not need a child poverty Bill with a legally binding target to make comparisons with other countries—we can do that anyway. I do not see that anything I am proposing prevents that from being done. The question is: what should we target? Clause 6(1)(c) states that regulations may provide for what is to be regarded as income, and all I am saying is that the child poverty commission should be able to advise the Secretary  of State that, as the Bill stands, it cannot take off housing costs, but that it would be a better measure of living standards if it did.

The current position is that the after-housing costs numbers are published. The Minister’s main argument seems to rest on our targets having to correspond with the international comparison numbers. Surely, if it is better for this country to use after-housing costs for our target in statute, that is what we should have. We would still publish the before-housing costs numbers, which would allow the international comparisons to be made. We would have all the information we needed, and a target that best suited the country’s needs. I do not think that the Minister has really addressed that central point.

I could not have put it better. On that basis, I commend the amendment to the Committee.

Question put, That the amendment be made.

The Committee divided: Ayes 7, Noes 9.

Question accordingly negatived.

I beg to move amendment 27, in clause 6, page 3, line 18, at end add:

‘and the costs associated with disability.’.

I should alert you, Mr. Key, to the fact that my hon. Friend the Member for Edinburgh, West will also seek to catch your eye on the amendment. You have probably heard enough of me, but the treatment of disability is an important issue. As we discussed in the evidence sessions, people receiving certain disability benefits to reflect additional costs may have a perverse effect on the statistics as they are currently constructed. The living standard of those people is measured by their income, which includes the benefit they get because they have extra costs, but nowhere are those costs taken into account. Those people, therefore, appear higher up the income and living standards scales than they truly are. Their promotion up the ranking is artificial, because all the benefit does is make good the costs that non-disabled households do not face.

We simply suggest that account be taken of the costs associated with disability, and in a particular way. There are two obvious ways of taking account of disability benefits. One is to exclude them. I am grateful that the Under-Secretary of State for Work and Pensions, in  following up a question I asked in a previous Committee sitting, has supplied us with a briefing note on the impact of excluding disability benefits from the HBAI measure. In parentheses, the note suggests that five topics are covered, but marginal deduction rates do not appear anywhere in it, so I hope we can have the missing section at some point.

The analysis shows that the aggregate number of those in poverty is pretty much the same whether or not disability living allowance is excluded. The totals are not affected. I am grateful for that useful information. However, we were getting at the composition, not the totals. The commentary on the numbers precisely states that if DLA is taken out, the median income is reduced, and therefore a set of children goes out of the figures. Presumably, they are children in households where nobody is disabled. However, there is presumably another set of children who will come into the figures—children in households where there is a disabled person. Although the total may be the same, the composition may be different. I do not think we yet have information about the composition.

If one is trying to measure relative living standards, excluding DLA is closer to the truth. One argument for excluding DLA is that we would get the right children. The total may not be different, but we would include more children in households with a disabled person than we do on the current definition, and the policy response may therefore be different. That all sounds like nerdy statistical stuff, but the point is that if our poverty figures include more children in households where there is someone with a disability, we will do more about disability—it follows. We need to get the right people in the figures, and not just the right totals. Therefore, there is a case for excluding DLA.

The ministerial comment on the Government’s figures stated that if we excluded DLA and then had a take-up campaign—so that more people received DLA—it would not show in the figures. That would be a bit odd, as clearly the Government have done something about poverty and it would not be fair if that did not show in the figures. I can see the logic, but our amendment proposes to tackle the issue in a slightly different way.

We have linked amendment 27 to the process of equivalisation. At the moment, we do not quite say that in a household two people can live as cheaply as one, but rather that a second mouth does not take twice as much as the first one, so there is a scaling. We take the total household income and scale according to the number of mouths to feed, the ages and all the rest of it to get an equivalent income—to scale households with different costs. We are arguing that we should have the same process for disability. Hence, rather than exclude a piece of income that someone is getting, which is one way of doing it, when we scale different households to reflect their different costs, we recognise that households with a disabled person have additional costs. They should have a higher equivalence scale, so that we compare like with like.

I recognise that that raises complications. The equivalence scale that the Government are using in their proposed figures is incredibly crude—the OECD scale is gobsmackingly crude compared with the one the Government used formerly, which was much more refined and based on real living standards. I accept that it may be difficult to graft the costs of disability on to the  OECD scale, which is horrifying in its simplicity. But we have to do something. Otherwise, the wrong children will be classified as in poverty and we will have the wrong policy response. Either we knock the income out or we take account of the costs. It is not good enough simply to say that we cannot do anything and we will just put up with figures that do not include children who are in poverty when they should be included, and vice-versa. I shall be grateful for the Minister’s response on how we can address the proper measurement of living standards of children in households with a disabled adult.

Members might remember that this was a point on which I questioned the Under-Secretary of State for Work and Pensions during our evidence sessions last week. I think the hon. Member for Northavon has made his points well. Disability living allowance is given to offset the increased costs of disability, which are many and varied, depending on the disability. Certain types of product may need to be bought to cope with the disability, and there may be increased heating costs, extra transport costs, as public transport is not suitable, or specialised child care costs. The issue will become larger as individual budgets start to loom on the horizon.

I, too, looked at the note we had from Ministers yesterday evening. I will be grateful if the Financial Secretary can explain something that appears curious. We were told that 23 per cent. of children were in households with below 60 per cent. of median income, with DLA and attendance allowance included as income. We were then given the figure for children in households with below 60 per cent. of median income, but excluding those allowances as income. The figure actually fell by only 1 per cent. to 22 per cent., which is slightly counter-intuitive, unless I am missing something. It tends to suggest that DLA is being taken up by slightly wealthier families, and that we have a problem with take-up at the lower end of the income scale. Those figures certainly are not what I expected, and I am left rather puzzled from a policy perspective, as to what that tells us and what we should do about it.

It is really a question of being honest. I hate to use that word, and I do not imply anything negative by it, but we have to reflect the reality of people’s lives. If income is given to cope with specific extra costs, one cannot really count it as ordinary income in the way that people who are not disabled would count additional income. That should be a given across the board; it is not a party political point, but a statement of the reality of the situation. I look forward to hearing what the Financial Secretary has to say.

My hon. Friend the Member for Northavon does not often leave his colleagues with much to say on such issues, so I feel like an apprentice with the master by his side.

One group of people who have been very keen on the Bill are those with a disabled child or a disabled parent in the family, because they realise that the move towards the targets will have a great impact on them. As has been mentioned, it is vital that increased costs for heating, diet, transport or child care are recognised as additional costs on the family, and that additional income from DLA or other sources should not be seen  as disposable income that the family will be left with. When we consider clause 6(c), which concerns

“what is to be regarded as the income of a household”, and when we consider what deductions are to be made, it makes perfect sense to consider the costs and the deductions that have to be made to put a family on a level playing field with all the other families who do not have a disabled person in the house. My hon. Friend’s amendment therefore makes perfect sense.

Equivalising incomes to take account of the costs of disability is tricky, and we discussed the issue in last week’s evidence sessions. I am not sure whether the hon. Member for Northavon was proposing a methodology for doing that—if so, I should be interested to know a bit more—but it is certainly a difficult issue and there is no consensus on how it should be done.

Last week, we were asked whether it is possible to determine the impact on child poverty levels of not counting DLA as income. My hon. Friend the Under-Secretary then quickly produced the letter that has been mentioned. As we have heard, the figures show that there is almost no difference in the number and percentage of children who are counted as being in relative poverty using income figures that include, and exclude, DLA and attendance allowance. The hon. Member for Northavon explained why that was, but let me express it slightly differently.

Excluding DLA and attendance allowance from household income would reduce median household income. Attendance allowance and DLA are more likely to be received by families without children. There are, as the hon. Member for Edinburgh, West has rightly pointed out, families with children who receive DLA for those children. On the whole, however, it is more likely to be received by older people who are under state pension age and who do not have children, or whose children have grown up, whereas attendance allowance is more likely to be received by pensioner households. The effect of taking those components of income away would be to improve the relative position of households with children in the income distribution, because fewer households with children would have had their incomes reduced, and to worsen the position of pensioners, as a higher number of households’ incomes would have been reduced. That is another way of looking at the issue.

It is also worth noting that the households below average income series, where our child poverty statistics are published, is carefully reviewed. As national statistics, they have to be. The most recent review, in 2004, which is on the Department for Work and Pensions research website, addressed the issue of the extra costs of disability and how disability benefits should be treated. The review recognised that disability benefit recipients’ position in the income distribution figures is quite high. That might be a poor representation of their standard of living, but the review concluded that removing disability benefits from income would not be the solution, as it would not reflect any changes over time in the receipt of those benefits, and would then overstate the living standards of non-recipients who had extra costs.

There clearly are additional costs associated with disability but, as the hon. Member for South-West Bedfordshire rightly pointed out, they vary significantly  in level and nature between individuals. There is no general agreement on how to measure them, at least not yet, so there is no generally agreed method to equivalise income to take account of those costs. My hon. Friend the Under-Secretary of State for Work and Pensions made the point in the evidence sessions last week that this does not mean that the extra costs of disability are not picked up by the measures in the Bill.

Once again the combined low income and material deprivation indicator comes to our aid. It allows a fuller assessment of the living assessment of those households facing particular difficulties due to high living costs such as the cost of disability because they will then have less disposable income available to meet other needs. That indicator will capture families who have an income that is higher than that captured by the relative low income target, but who still have a lower standard of living because they have additional costs related to disability or other additional costs to bear.

In developing the child poverty strategy, we will consider the evidence on which groups are likely to be most vulnerable to poverty and what measures are needed to meet their needs. Those groups are likely to include children in families with a disabled family member. The variable nature of the costs associated with disability make it difficult to equivalise income to take into account the costs of disability in the way proposed by the amendment. The hon. Gentleman has once again raised an important point. I am grateful to him for doing so but I hope that he will accept that his amendment does not give us a solution to the challenge here.

If the amendment prompted the Department to do more work on the costs and impact of disability to assess the adequacy and coverage of disability benefit levels, that of itself would make it worthwhile. I accept that there is not a straightforward way of doing this and perhaps the work required to do it might delay the whole process, which we do not want. I notice the Minister again relies on the argument that the material deprivation indicator would pick this up. It is going to pick up disability costs and housing costs so it will take quite a fine statistician to work out what is going on in this catch-all. We need to pick these things up a bit more directly. I beg to ask leave to withdraw the amendment.

Amendment, by leave, withdrawn.

Clause 6 ordered to stand part of the Bill.