HoC 85mm(Green).tif

Work and Pensions Committee

Oral evidence: Children in poverty: measurement and targets, HC 1138

Wednesday 3 March 2021

Ordered by the House of Commons to be published on 3 March 2021.

Watch the meeting

Members present: Stephen Timms (Chair); Debbie Abrahams; Siobhan Baillie; Steve McCabe; Nigel Mills; Selaine Saxby; Dr Ben Spencer; Chris Stephens; Sir Desmond Swayne.

Education Committee Members present: Dr Caroline Johnson.

Questions 1-44


I: Dr Keetie Roelen, Research Fellow and Co-director, Centre for Social Protection, Institute of Development Studies; and Dr José Manuel Roche, Research Associate, Department of International Development, University of Oxford.

II: Dr Kitty Stewart, Associate Professor of Social Policy and Associate Director, Centre for Analysis of Social Exclusion, London School of Economics and Political Science; Edward Davies, Policy Director, Centre for Social Justice; and Helen Barnard, Director, Joseph Rowntree Foundation.

Written evidence from witnesses:

Joseph Rowntree Foundation [CPM0019]


Examination of witnesses

Witnesses: Dr Keetie Roelen and Dr José Manuel Roche.

Q1                Chair: A warm welcome, everybody, to this meeting of the Work and Pensions Committee, for our inquiry into children in poverty. I particularly welcome the witnesses who join us this morning. I would like to ask both witnesses on the first panel briefly to say who they are, then we will start with some questions. I first call on Dr Roelen to say hello.

Dr Roelen: Good morning, everyone. Thank you, Chair. I am Keetie Roelen, research fellow at the Institute of Development Studies in Brighton. I am also co-director there of the Centre for Social Protection.

Dr Roche: Good morning, Chair. I am José Manuel Roche. I am a research associate at the Oxford University department of international development—specifically at the Oxford poverty and human development initiative, which is an economic research and policy centre within the department.

Q2                Chair: Thank you for joining us. May I ask both of you, first of all, to tell us how child poverty is defined in international work—for example, in assessing progress towards the sustainable development goals? Also, having signed up, as the UK has, to the sustainable development goals, what are our obligations in relation to defining, measuring and reducing child poverty? Keetie, could you start off on this one for us?

Dr Roelen: Yes, of course. First of all, in terms of the definitions of child poverty in international terms, there is no one single definition that is used across the world, but there is agreement in terms of the commonalities across definitions, and that is that income is important and also that there are other deprivations that might not necessarily be directly linked to income that are important to look at. That is exactly what is reflected in the sustainable development goals, and particularly in sustainable development goal 1, which says that we should end extreme poverty for all by 2030. Within SDG 1 there are two targets. The first one speaks to international poverty measures and particularly the income measure of $1.25 a day. The second indicator speaks to poverty “in all its forms everywhere”. That is based on national definitions. It is in that target that the multidimensional form of poverty—indicators other than income—are recognised. This is a considerable shift from the millennium development goals that we had previously, where there was more of a narrow focus on income poverty.

If we look at definitions of child poverty across the world, we see that these two elements are largely recognised everywhere. So, child poverty is about living in families with lack of adequate income, but it is also about being deprived in other areas that are often closely related to lack of income, such as being unable to go to school, or having healthcare if you are ill, or living in adequate housing. But it is not always directly related to income, which is why we need to look at those other types of deprivations more directly.

Q3                Chair: Thank you. José, would you want to add anything on this?

Dr Roche: Yes, I could add a little bit. As Keetie said, the first sustainable development goal of the framework is ending poverty in all its dimensions. I think there are two things it would be useful to add to what Keetie just mentioned.

As well as having two sets of targets and goals—one about monetary or poverty income and another that is in all its dimensions or all its forms, which refers to health, education and many other dimensions—the framework has two levels. There is the level of indicators that are for global comparison. For example, the World Bank produces poverty levels or poverty statistics to compare across countries, and there are different types of levels. If you are a poor country, a middle-income country or a rich country, you use different minimal thresholds.

Then there are other measures to compare multidimensional poverty globally as well. For example, the UNDP—the United Nations Development Programme—in collaboration with the centre I am associated with, produced a multidimensional poverty index that allows comparison for a set of countries such as a middle rank of countries. What is very important is that, as well as the global goals, countries are expected to define their own goals—levels of goals or targets—and their own metrics. Countries then voluntarily prepare reports to the United Nations showing whatever targets that they have defined nationally and any other types of indicators that they are using nationally. For example, almost every country in the world would have measures of income poverty, with national targets and measures. Most counties would also have some level of multidimensional poverty. For example, up until now, in the OECD, 64 countries have reported on multidimensional poverty—poverty in all its dimensions—and 28 counties have a specific indicator that puts together dimensions in a metric that they can then communicate and use for policy. That gives a little bit of an overview of how these have been implemented.

Q4                Chair: So is the UK obliged to submit a report along these lines?

Dr Roche: It will. Countries report voluntarily to the UN. Every year, a set of countries propose that they want to submit their report. So yes, the UK is expected to report on these targets. The UK has signed, like every other country, the SDGs. As you know, the previous goals—the MDGs—only applied to developing countries. The new set of goals are no longer just about poverty; they are also about issues of environment and development more broadly, and they apply to every country. They apply to the UK and yes, the UK is expected to report.

Q5                Chair: When is that report supposed to be produced?

Dr Roche: The voluntary report happens every year. A set of countries report every year, so the UK would have to decide exactly when they want to report. The set of goals is to 2030, so it is expected that by 2030 every country would have a final report of what they have achieved.

Q6                Chair: That’s interesting. I have two more points on this. What is the reason for having shifted away from the narrower focus on income that was in the millennium development goals? Secondly, if you were looking around the world for one or two countries that have done a really great job on measuring child poverty, where would you point to the best practice that we might see? That is for either of you.

Dr Roelen: The shift towards more complementary measurements of poverty—not replacing an income measure but adding a multidimensional measure of poverty to an income measure—follows a lot of analysis across the world looking at poverty using this income measure, and also looking at what children’s lives look like in other areas and domains, and seeing that, although lack of income is strongly correlated or associated with other deprivations, it doesn’t always explain what is going on in children’s lives. As I said earlier, that could be education or health, or even things like housing or access to water and sanitation, where income is usually thought of as being quite a strong determinant of having access to these things, but it turns out that there is a quite a mismatch when using these indicators.

Looking at material deprivations or deprivations in other, sometimes psychosocial domains, is sometimes seen as a more direct measure of children’s lives and families’ lives, and moves away from a narrow reliance on just the monetary measure. That was quite a heavy debate at the time when we had the millennium development goals.

As José mentions, part of that conversation was a recognition that this is not only important in lower and middle-income countries, but in countries across the world. In this shift towards the sustainable development goals, we now have a broader view of poverty, in terms of what we should look at and where in the world we should look at these things.

In terms of best practices, there has been a lot of effort in low and middle-income countries in trying to find measures of multidimensional child poverty that are both at a national level and context-specific, as well as relevant for the international comparisons that we just heard about. I think actually a lot of higher-income countries—rich countries—can learn quite a lot from these efforts in low and middle-income countries, because there has been a lot of experimentation with measures. For example, in Vietnam the Government have been developing and working with measures of multidimensional child poverty added on to income measures since 2008.

When we talk about using income measures and multidimensional measures alongside each other for the population as a whole, Mexico is a really good example. They use both measures on a regular basis to track what is happening and to work towards targets.

Dr Roche: To add to what Keetie has just mentioned, I think it is very important to notice that, although there is a novelty in the SDGs, they also take what was already done under the MDGs—the goals that were set up in the 2000s. At that time already, the old set of goals recognised that poverty also manifests in education and in health. Goals were set for all these different areas of poverty—there was one on income, but there were all those other areas. What the SDGs—the new set of goals—do is bring to the first goal, when it refers to poverty, an explicit mention of “poverty in all its forms”, so it makes it a little more clear. Countries have agreed that they will have specific measures and a specific way to monitor multidimensional poverty.

On your second question on where to look for good practices, there are plenty of good practices. In terms of countries, there are three that the UK could look at. Mexico has done fantastic work. Its Parliament established an independent entity to be in charge of all monitoring of poverty and its policy impact, and it was given a mandate from Parliament to design a set of measures to monitor poverty—both poverty of income and multidimensional poverty. There is a whole interrelation between that entity, Government and a strategic plan to reduce poverty.

Colombia is another example, but they did it slightly differently—from within Government. We can talk about it later, but President Santos, the President of Colombia at the time—from, I think, 2010 to 2014—first made the Government develop a multidimensional measure, but he also established a roundtable with Departments and parts of the Government to co-ordinate the monitoring and policy impacts around poverty. Those are two examples.

In richer countries, New Zealand is a good example as well. It does not put it together into an index, but it has a whole set of institutional organisations to monitor and implement this.

It may be useful to mention that there is a network called the Multidimensional Poverty Peer Network that brings together 61 countries and 19 organisations to focus on multidimensional poverty. Some of those 61 countries will have a synthetic or aggregated index of multidimensional poverty; others will choose to have it separately. But that is a peer network where Governments are challenging and supporting each other in the development of this type of methodology, which is also quite new.

I mentioned that the MDGs had a multidimensional approach, but more sophisticated measures of multidimensional poverty are relatively recent. While measures of income poverty were developed in the ’70s and ’80s, multidimensional poverty just started in the ’90s, and it was developed much more around 2000. That also explains why they are at different stages of development.

Chair: Very interesting. Thank you very much. Steve McCabe.

Q7                Steve McCabe: Good morning. I want to understand a little bit more about the measures that are used, particularly in relation to child poverty. As I understand it, quite a number of people in the Global Coalition report say that using household measures quite often means that we almost miss the number and the state of children living in extreme poverty, so I wonder how important it is to understand the situation of individual children when we are trying to measure poverty.

Dr Roelen: Trying to understand the situation of individual children, instead of children as part of families, is very important to find out the situation and vulnerabilities of individual children. As one can imagine, younger children or, often, girl children are more vulnerable than older children or boy children; or, depending on context, other individual characteristics may play a role. Current measurement of poverty—this is the world over, I would say—is quite limited in picking up on these issues and individual differences. Income poverty is especially ill-equipped to pick up on these differences, because income poverty is a household measure in the way it is developed. We pool all the resources of a family or household together, and then disaggregate it by individual to understand individual poverty, but in essence, it is all based on this collection of the household’s resources. To use that to understand individual children’s situations is very difficult.

Here, the more multidimensional measures offer promise, because you can look at indicators that are not income-related and indicators that are very specific to individual children: whether one specific child goes to school or not, what kind of food they eat, or whether they are malnourished or not. There is definitely potential there, and certainly a lot of progress has also been made in trying to develop these more individually focused measures of deprivation. Still, by and large, there is a lot of progress to be made. When it comes to data, we are quite constrained in how we really focus in on the individual. Oftentimes, for example with food, we will still look at what kind of food might be available within the family or household, or the main respondent to a survey will be asked how many meals are being eaten on average by members of the family, so it is very difficult to get the information about individual children.

At the same time, though, other indicators around education and health can be collected, and there are more efforts to do that. I should also point out that of course, for some other dimensions or indicators, it is absolutely logical to look at it from a family or household perspective: if we look at housing conditions or living standards, for example. But yes, it is vital to try and focus on the individual as much as possible, and to work on the way in which we are collecting data and how we build that into our monitoring system to make that possible to the best extent.

Dr Roche: As Keetie is saying, and as you are rightly asking—[Inaudible]—to solve the problem as well, rather than doing it individually. The reason is that child deprivation changes by age group: for example, babies do not go to school, and you cannot necessarily measure learning at a very early age. The child deprivation may be linked to the household level: for example, if a child lives in a household that does not have an employed adult, then of course that household has a greater risk. You have these different layers and I think it is worth explaining some of these as well. Perhaps the advice there, based on what other Governments are doing, is to think first about the policy uptake and how it is going to be used and communicated. Your aim with some of these measures is often to have something that is relatively easy to communicate, because you want to communicate with the public, but you also want the measures to drive policy to allow the Government to co-ordinate between Departments or with local government. You need to think a little bit about how you are going to design the measure and then you may come to a compromise in the way that you define it. The most common approaches in government have three different avenues. One of them is to include child indicators in a national measure. You would have data on schooling or learning in the national measures. You may have data on nutrition and food security related to children. Then you have, of course, indicators that refer to the whole household, for example housing inadequacy or services.

A normal approach—a frequent approach—is to disaggregate these metrics: to disaggregate income poverty measures and reports on children. We know from analysis that there are more children than adults who are poor, so children are more often at risk of being in poverty. Equally, with a multidimensional measure, you can disaggregate children and you might do intra-household analysis. That is more complicated because intra-household analysis that leads to policy may not be something that you communicate as much to the public, but you have the overall measure and disaggregation by age and analysis for policy.  The third, and last approach, that is often used, is to build a child MPI, multidimensional poverty index, directly linked to national metrics. That is what the Government of Mexico has done, in conjunction with UNICEF. Coneval, the institution that they set up, and UNICEF have worked together to have a metric, specifically for children, that is then linked to a national metric of poverty that the country has. They are able to explain both of them, report it and also, of course, to work around policy with both metrics.

Q8                Steve McCabe: Can I ask about that last point? I thought the situation with the global multidimensional poverty index was that it was probably slightly better for comparing countries than looking at the situation that individual children might experience in a county. Are you saying that, where you link it to a better national measure as well, you get a much fuller, or more complete, picture? Is that the point you are stressing?

Dr Roche: That is precisely right. It is very important to think that we have a metric that allows us to compare countries. There is the question of what countries you want to compare. If you are going to compare rich countries, you need to have the metrics that are relevant for comparing rich countries. There is work there, for example done by UNICEF, on child poverty across rich countries, but that was not for comparison. When you are going to have metrics for your own country, then you are expected to have your own metrics—your own design. It is what countries do, because you would have your own set of priorities and national consensus. You would build that consensus and define what the—we say dimensions—forms of poverty that matter in the policy discussion that you are having in the country are. What Mexico has done is that the Parliament defined priority areas and then this other technical entity that they established proposed the metrics, so they work around the methodological issues to then implement that framework, as do the parliamentarians. In that discussion, they have a national and a child poverty Bill within it. Different countries have different strategies to address that, but that is a technical discussion. In the case of Mexico, the Parliament defined what they wanted and created the policy set-up. In Colombia, it was the Government—the presidency—who set up the mandate. Different countries choose different paths, but it is a national metric.

Steve McCabe: Thank you very much. That is very helpful.

Chair: Dr Ben Spencer.

Q9                Dr Spencer: Thank you for that very interesting discussion. Building on that a bit, as I understand it, what you are saying is that, in terms of multidimensional measures of poverty, you have a debate that defines the key policy areas that we want to influence and think are important, then metrics are designed to quantify that. That can return either a single score or multiple scores on each of these different dimensions. During that process, if you are trying to return a single score, how do you go about weighting that, and how do you think that through, bearing in mind that you are going through different dimensions? What is your experience of the overlap or co-linearity between the different dimensions when you are looking at poverty? Although we talk a lot about housing and education, how much do they overlap with each other when you break it down into different metrics? How do you resolve those issues?

Dr Roelen: Picking up on your very first point about policy areas and what José just said about identifying the dimensions that you want to focus on, there is a question about whether you want the dimensions defined on the basis of the policy areas that you want to focus on, or whether you take a more open-ended approach, whereby you look at dimensions of child poverty that you think are important in children’s lives, then have that information guide you as to how you should formulate policies that try to improve children’s lives. There is a bit of a chicken-and-egg story here. If you have already defined your dimensions on areas where you want to intervene policy-wise, you might be missing out on other areas that are extremely important in the lives of children and families that you wouldn’t be able to pick up on. I just wanted to make that point before moving on to your two other questions about the weighting and the overlap between different dimensions.

Weighting is a very tricky discussion, and it can very quickly become very technical. There are lots of statistical techniques that one might use to figure out in the data which dimensions and which kinds of deprivations could weigh more heavily than others. By and large, when looking at multidimensional measures of poverty, as well as child poverty and how they are implemented for policy, they tend to use equal weighting of dimensions—key areas that are important in people’s lives. That is not to say that dimensions are not weighted, but there is a decision that all dimensions are equally important. In part, that is because it is difficult to make judgments about which one is more or less important than the other. Is access to health more important than access to education, for example?

Some would also argue that, on a rights-based approach, you have to value all these things equally, so equal weighting from that perspective makes the most sense. From a very practical perspective, it is also the most intuitive way to convey findings of these different indicators. It is easier to present them individually if we have rates of deprivation in relation to housing or access to food, for example, and also when they are combined in the summary statistics that we have been mentioning. Equal weighting is then more intuitive.

Q10            Dr Spencer: Do you mind if I come back on that? That is very helpful. It strikes me that if you have an equal weighting approach, that will lead policy makers, perhaps by preference, to tackle areas that are more easily resolved than more difficult ones, because then you see a bigger change in the metric for a certain intervention. I guess my question is whether there is a danger that that leads into a policy approach where you hit the low-hanging fruit, as opposed to really trying to tackle the entrenched, very difficult problems in policy.

Dr Roelen: Yes, that is absolutely right. It is particularly a concern when you only use summary statistics, whereby you aggregate all the information across different dimensions into just one number. It is very easy to then hide the fact that one might have targeted the low-hanging fruits, and it is unclear that some of the other areas still suffer from a lot of deprivation. A common approach is therefore to use the summary statistic, where everything is lumped together, alongside a presentation of findings that break the summary statistic down into its various components—the different dimensions—so that you can come to a more nuanced understanding of where progress has been made, acknowledging that these might have been the easier areas to reach people or to make change, and indicating where a lot of progress still needs to happen.

That also speaks to your point about the overlaps between these different dimensions. The overlap of deprivation between dimensions that are non-income related, such as education, health, water and sanitation and so on, is very dependent on context, but when we look at analyses that have looked at the overlap between income and multiple dimensions of poverty or material deprivations across contexts, from our own work, we see that it is almost split up in four groups, if you will. If you adjust rates so that they are equal across the different groups that are identified as income or multidimensionally poor, we see that there is about a quarter of the population that can be identified as being poor according to both indicators—income and material deprivation—and then there is about another quarter of the population who are only identified as poor using either income or these multidimensional measures. This holds in analyses of countries in the EU, as well as in lower and middle-income countries. So, the lack of overlap between income measures and multidimensional measures is quite persistent.

Dr Roche: Maybe I can add to that. The weight—why the weight? We like to say that the weight is the values—the normative value or the importance that you attribute to a particular element. For example, you may say that learning is very important, so being deprived in learning should have a really high weight—should have a really high value—whereas other elements, perhaps, are slightly less important, in relation to certain types of assets in the house. But health, for example—being free from hospitalisation, for example—is very important, so you give it more value.

There is an element there where it is possible to have this discussion at the policy level. It doesn’t have to be only technical. There is a discussion around building a consensus on what matters for the particular country and for the particular citizen. There are different ways to attribute weight. Some mechanisms consult the public on what they attribute more importance to in their lives or in relation to child poverty. As you rightly say, that links to a policy trade-off. We are constantly, in terms of policy, having to make a trade-off between putting resources in one place or another. Those trade-offs are linked, partly, also by those values. So that discussion should happen through the metric.

What is very important there is that you need to have some criteria. The criteria need to be transparent and easy to communicate, because you want to have that debate—you want to have the debate about the potential trade-off and the importance of the different areas in children’s wellbeing. Then you want the weight and the indicators, and the whole structure of the index, to represent either a political or a social consensus in the country. You want to try to build it that way.

For practical reasons, often we follow a level of equal weight, but it is not quite equal weight. For example, in the MPI, you have three dimensions. You have health, education and living standards. Each of those dimensions has equal weight, but within the health dimension you have two indicators, which are child mortality and nutrition. For living standards, you have six indicators. The equal weight of living standards is divided among six indicators, where the equal weight of education is divided between two. As a result, you have a much higher weight for child mortality than you would have, for example, for a certain type of services. That, of course, makes sense from a consensus perspective. You can build an indicator in a certain way that is relatively transparent, that is easy to communicate and that builds on the political consensus, and then, like Keetie was saying, the metric can disaggregate.

Let me add one thing here. The choice of the dimension is very important, and it can be very political, and that’s good, because you want to build a consensus. Keetie mentioned what you should focus on ideally. Government focus in on wellbeing. What are the elements of wellbeing of the children that we care about? We care about their nutrition, their health, their living standards, their education and learning, and so on. So you include areas that refer to wellbeing.

In order to improve their wellbeing, the policy may be more complex. So you would have the wellbeing indicator there that tells you if the children are improving their wellbeing—or “illbeing” if you have left children with illbeing—and then you would have the weight. The last thing that is useful to say there is that then you have the aggregation of all this. There are different, multiple way to aggregate.

Once again, when it comes to policy, you want to think of the policy use of it. You want something that is easy to communicate, that is intuitive and that the public would understand. What is best, both with child poverty and normal poverty, is to use counting approaches. By counting approaches, I mean that, instead of just getting an index that is quite abstract, you can actually say how many people or children are experiencing multiple deprivation at the same time. You can also say what is the severity of that type of deprivation, and you can say what type of deprivation they experience, or what packages of deprivation they experience.

You can use metrics to do that. The method that we use—the Alkire-Foster method—is the one that 88 countries use. That is the one used for the UNDP and for the global metric. It is a metric that allows doing this method of counting. UNICEF use the same metric within the MODA, which is another index that they have, and have the same method, the same counting—the same arithmetic.

Again, that is a technical discussion. You could set up the parameters. You can say, “We want something that is easy to communicate, that can lead to policy co-ordination, that is transparent and that the public can understand.” With those parameters, those who are going to work the technical element of it then need to find the metric that best meets the parameters that you are interested about.

Q11            Dr Spencer: Thank you. I have another quick question before I hand over. In terms of developing a metric for national use, it is not the case that you say, “Let’s build this metric and make policy to respond to it.” The strong argument that you both made, as I understand it, is that the development of a metric goes hand in hand with the development of the policy on how to respond to poverty, because how you choose what is inside the metric is driven by what your priorities are and by that debate and consensus in terms of how you define wellbeing and areas you want to change.  The two go hand in hand, working together going forwards to alleviate and inculcate change.

Dr Roche: Allow me. Maybe I can explain the case again. The areas are coming from a legal framework. In the case of Mexico, they come from the constitution of Mexico, which established the basic fundamental rights that individuals should enjoy in relation to minimum access to education and minimum learning, for example, and access to health and being free from illness, and so on. That legal framework is the one that then inspires the choice of the indicators that are used to operationalise—to put into practice—the mandate that comes from the legal framework. Then, of course, Parliament set up a whole institutional layout.

Let’s give another example. In the case of Colombia, the Government set up an institutional layout. What is very interesting is that the index and the institutional set-up that they established facilitated co-ordination between the Government Departments and local government. They had a national development plan for the areas that they wanted to improve in the country, and they had two mechanisms. They had a roundtable, that the President would chair—in the UK Parliament, it would probably be different, but here it is the President—and that convenes Ministers and agencies working on reducing poverty. Because the metric brings together elements that relate to income, living standards, education and health, you get a conversation happening between the different Departments about the type of policies and the type of co-ordination that they need to establish in order to address these dimensions. Education is linked to health; if you are better educated, you will have better health, because you know how to use the NHS and look after yourself. There is a relation between these dimensions.

The other thing that they had, as well as the roundtable, was the dashboards. They have a dashboard of indicators. It is like when you are driving a car and you have a dashboard that tells you what is going on. If you have a plane, it is more complicated; you have a bigger dashboard. The way it works is that you have one metric that is the summary—the final summary and the way that you communicate. Then you have a set of indicators that comes down to that one metric. That is the whole dashboard. Of course, if you do not understand the full dashboard, you may be flying the plane based only on a couple of indicators. You can still get to the end, but if you understand the whole dashboard, you are probably flying the plane much better.

Dr Spencer: Thank you. That is really helpful.

Q12            Debbie Abrahams: Good morning, everyone. This is a very interesting discussion, but I have to express a little concern, as a former academic, when we are suggesting that we might introduce politically devised measures. I want to clarify the evidence that should underpin the development of aggregated measures, and to ask whether you have witnessed any unintended consequences of measures that may have been politically devised rather than evidence-driven.

I am thinking of our own local circumstances, where we know, for example, that, in terms of the increase in infant mortality between 2014 and 2017, there was found to be a causal association for every 1% increase in child poverty—as an economic measure, there were five additional deaths per 100,000 live births. If we were looking at how that might be politically used in terms of, as I say, an aggregated measure, it raises a few concerns, I have to say. What are your views on that, and where have you witnessed unintended consequences of politically devised measures?

Dr Roelen: If I can come in here, on the point about politically motivated measures and the conversation that preceded this question, I want to be very clear that, from my perspective, any measure of child poverty, or poverty, should be based on what we think is important for children’s lives and for families’ and individuals’ lives. Poverty should be measured in terms of what we think people should have access to and what constitutes a basic standard of living. If we can devise a measure that links to areas that are important in terms of policy making, that is really helpful in terms of informing the policies. But I am not advocating that the measure should be driven on the basis of policy priorities. 

Q13            Debbie Abrahams: You used the phrase, “what we think” in terms of what this is driven by. What we think is value-driven; it is not evidence-driven, or not always. What we think drives poverty is different from what the evidence says drives poverty. Economic poverty is driving the increase in infant mortality by five additional deaths. That is evidence. This is where I think we get into some difficulty.  

Dr Roelen: Yes, that is a very important point. Let me make that distinction even more clearly. We need to distinguish between what we think constitutes poverty and so the outcome that we are looking at—that is the standard of living that I was referring to, and the various dimensions of it. That is an economic situation, as well as some of the material or other deprivations that we have spoken about.

Then there are the drivers of poverty, or causes of poverty, which we need to understand in order to then think about how we can adequately improve the situation of children and families. What kinds of policies need to speak to these causes and drivers of poverty?

In some measures—I would also include the UK in there, and how child poverty is measured at the moment—these two things are confused. When we look at worklessness, for example, this is a factor that might be driving child poverty, but not necessarily. If we look at child poverty as a set of deprivations, what the situation is like for families in terms of work—

Q14            Debbie Abrahams: I am sorry, but I need to check you on that because, again, the evidence does not support that. What you are saying about worklessness—we know six out of 10 families in work are in poverty. We know that three quarters of the 4 million children in poverty are from working households. Perpetuating these sorts of stereotypes and unfortunate inaccuracies is really a dangerous and slippery slope.

Dr Roelen: I agree 100% with you, and I’m sorry if I am not making myself clear, but that is exactly why worklessness should not be a measure of child poverty. But at the moment, it is used as an indicator of child poverty, and I think that is wrong. It is not an indicator of child poverty; it is about a potential reason for why children might be in poverty. But as you say, in many cases that is not true—not just in the UK, but the world over. That is why in most other countries in the world, an issue such as worklessness or the educational attainment of parents is not used as an indicator of poverty or child poverty, because it might be a reason, but it is certainly not always the reason. There is a very complex set of reasons why children or families might be in deprivation or poverty, and it is very important that we distinguish between these outcomes and use that to monitor the situation over time in relation to all the potential different factors and how these might be combined in relation to policies in place in order to get to the right situation: reducing poverty with the right set of measures in relation to those factors.

Dr Roche: Can I come in there?

Chair: Briefly, if you would.

Dr Roche: Your point is a very strong and good point about how it should be evidence-based and technical, and there is a distinction between political and policy, or consensus and technical. That needs to be quite well established, and I have to disagree with Keetie. Employment sometimes is in multi-dimensional metrics. It is in the case of Colombia; they have employment and issues of employment.

On the way it is seen—I agree with Keetie on this—they see employment not because it is a cause of poverty, but because it is also a manifestation of poverty. You want to be healthy, you want to be educated and you want to have good employment—that is what people want. When you do not have employment, you suffer deprivation—you want to work but are unable to work.

That is the political level. The political level is to get an agreement on the elements of wellbeing that we want people either to experience, at least at a very minimum level, or to avoid falling below. There is a political discussion there, and it is value driven. You need to say what the elements of wellbeing are that matter, and that value can be driven by the Parliament or by society—by what people in the UK think matters to their lives. Those are values.

The technical discussion comes just after. I think you are absolutely right. There is the issue of what is the best way to measure it. What are the best indicators to measure it—indicators that cannot be changed with a very simple policy intervention? You need to have a comprehensive intervention to improve the situation, so you tend to choose certain types of indicators that are easy to communicate and based on evidence.

Finally, when you discuss the costs of poverty, the index or indicator should then help you, based on evidence, to understand what it is costing you. I think that that is what Keetie was explaining: you have the metric of what wellbeing is, and then you have the analysis on evidence of what drives that wellbeing. The evidence says that employment is, of course, very important for wellbeing, and good employment, which is obviously paid, at the very minimum.

Chair: Thank you very much. Selaine Saxby, you have the last question for this panel.

Q15            Selaine Saxby: Thank you, Chair. Some of this you have already touched on, but back in 2016, the UK changed its approach to how we measure child poverty, from looking much more at income drivers to looking at other measures. What is your opinion of the changes that were made?

Dr Roelen: Some of this we just spoke about. I think the measures have not changed for the better. In fact, I would much prefer the measures that were in the Child Poverty Act 2010. Earlier on, we spoke about best practices, and José mentioned New Zealand. I would also like to point out New Zealand, where they have a collection of indicators to look at child poverty and have targets attached, which are very much like in the 2010 Act, looking at lack of income before and after housing costs, at a set of material deprivations for children, based on indicators that we have referred to before, and at a measure of persistent poverty—whether families and children stay in poverty over a longer period.

Of course, that is very much related to what was in place before. It is also partly in line with what happens in other European countries and their at-risk-of-poverty and social inclusion measures, which are, in some part, also a combination of looking at lack of income and material deprivation. Like I said before, I think that is important, because it is vital to distinguish what is valued as outcomes for people—the situation that we want people and children to live in—from some of the factors that drive those outcomes, to be able to put policies in place that speak to those factors and drivers.

Dr Roche: Briefly, I am no expert on the UK, so I will not go into too much detail, but from what I have seen and heard, part of the problem has been, first, moving from income to multidimensional, then ignoring income. Of course, income is still a complement. The second point is about making a distinction between what we value, what matters and what the policy, the political discussion and the consensus are, and then the technical element. It is probably useful to distinguish between those two things and to have a very open, frank discussion on what we want to see in terms of the wellbeing of children in the UK. We of course want them not experiencing income or living standard deprivation. We want them to be educated and learning and developing. We want them to be healthy and avoiding certain types of illness. Then there will be other things: we want their housing to be at a certain level and so on. But it is for Parliament to propose what this should be, and to discuss it with citizens.

The second set of things would be about how you measure it and what the determinants are. What things would be involved? How would we get there? What are the costs? But I would leave the costs of poverty to the analysis that you will do and to the policies that you implement. What you want is a metric that tells you about the outcome that you want to achieve.

Selaine Saxby: Thank you.

Q16            Chair: You have told us about some interesting examples of dashboards, metrics and things that are in use around the world in looking at child poverty. Could you point us to where, after this session, we can see those? We would like to get a feel for what is being done in Mexico. I don’t know whether what is happening in Mexico is presented in English at all, but if it is, it would be interesting to see that. You have talked also about New Zealand and one or two other examples of the approaches that you have described to us. I think the Committee would find it interesting to have a look at those.

Dr Roche: You can check the CONEVAL website—CONEVAL in Mexico. That is in both English and Spanish. It is not only the dimensional aspect; they have a whole framework to do the evaluation of the social poverty in the country, and policies as well. That, for me, is one of the best practices. What is really good about it is that it is driven by Parliament, but this is also an independent technical entity, with a very clear mandate coming from Parliament. I would say yes, you look not only at the metrics and the dashboard, but at the institutional set-up that has been established. New Zealand is another good example. There is Colombia as well. We can share some of those. You could probably also go to the website of the Multidimensional Poverty Peer Network. There are different publications there. They are not academic; they are actually policy publications, because it’s Governments coming together. There are different types of policy publication there on how countries are implementing them.

Q17            Chair: Thank you. Keetie, are there any other examples that you want to point us to?

Dr Roelen: No specific examples beyond the ones that we have mentioned. I would point out that the Global Coalition to End Child Poverty have done a lot of work bringing together different experiences from around the world in measuring child poverty, but also on how that measurement sits within those political and policy landscapes. They also do an annual report that looks at how countries report against the sustainable development goals in their annual reporting, as José mentioned at the beginning of this session. I can also point you in their direction.

Chair: That is really helpful. Thank you both very much indeed for giving us a very useful international perspective as we start our inquiry. We look forward to those additional pointers that you can draw our attention to. That concludes our session with the first panel.


Examination of witnesses

Witnesses: Dr Kitty Stewart, Edward Davies and Helen Barnard.

Q18            Chair: We warmly welcome the three witnesses joining us. Thank you for being with us. Can I ask each of you, as I did with the previous panel, very briefly to introduce yourselves, so that the Committee knows who you are, starting with Helen Barnard?

Helen Barnard: I am Helen Barnard, director of the Joseph Rowntree Foundation.

Edward Davies: I am Edward Davies, policy director at the Centre for Social Justice.

Dr Stewart: I am Kitty Stewart, associate professor of social policy at the London School of Economics and associate director of the LSE’s Centre for Analysis of Social Exclusion.

Q19            Chair: Thank you all very much for joining us. I am going to ask the first question. I was the Minister who took the 2010 Child Poverty Act through Parliament. It was repealed in 2016, and the child poverty targets were taken off the statute book. I would be interested to hear what each of you thinks, with the benefit of hindsight, the advantages and disadvantages were of having statutory child poverty targets in place. Let’s do it in the same order, shall we, and start with Helen?

Helen Barnard: I think the advantages are that if you define an ambition, you define what you consider to be success, and you have regular points at which it is reviewed. That can create a good drive towards action. There are probably two downsides. One is that there can be a danger that you motivate Governments to take shorter-term actions, because they will show up faster against the target, and it can be harder to create room for the much longer-term interventions, which also need to happen. For me, though, there is a slight difference between the benefits of having a target and the benefits of legislating a target. I think Government should have a target, should publish it, and should be held to account for it.

I am quite sceptical about the added value of putting a target into legislation. Wwhat you should legislate for is publishing the measure of how you will know which way it is going; that is what we need. I am not convinced that actually legislating targets in and of itself drives action. It is down to whether the Government of the day has the political will to do that, not whether it is in law.

Edward Davies: I agree with some of that—not all of it. I do think that what gets measured tends to get done, so there is something quite important in targets. The one that really stands out for me—I used to work in health—is the four-hour A&E target. It really drove waiting times down. The other benefit of it is that when we are not hitting that target, it gives the media and the Opposition something to hold the Government to account on—“You said you’d do this; you have not. Why aren’t you doing it?” So there is a real driver there. What we notice at the moment, as there isn’t really a target here, is there is not anything to push on, because there isn’t a strategy at the moment. There is nothing on we can say “Well, you said you’d do that, and you haven’t.” It is quite important to have these things.

It is good for Government, too, because if someone comes to Government and says, “You are not doing this thing,” they can say, “Ah, but we are. We are doing this,” so they should want this. They should want to be held to account, because then they can talk about what they are doing.

Dr Stewart: I think the targets, even before they were legislated, focused policy attention on this problem. They also acted as a signal that Government took this seriously, and that may have had an impact on public opinion and what the public thought was important.

The main disadvantages are that targets can potentially create distorting effects—you focus on what the target is, and not other aspects. One example of that is when the target is a “before housing costs” measure. We saw much more improvement on a “before housing costs” poverty measure for children than on an “after housing costs” measure. In some senses, that shows the effectiveness of the target, but if the target had been a different one—after housing costs—we might have seen slightly different action. That is not a reason to abandon targets, but to be careful about what they are, and to have a broad set of targets.

I also echo a little bit what Helen said: the really important thing in that Child Poverty Act was all the structure that came around the targets. It is no good just having a target. Also important were the annual reporting mechanism, the requirement for Government to have a national child poverty strategy, and local government. That did drive a kind of long-term focus, as well as short-term focus, because there was a need to point out each year not just what had happened this year and why, but “How are we going to get to this 2020 goal?”.

Q20            Chair: Do you share the scepticism that Helen expressed about having the target in law?

Dr Stewart: I think it is interesting having it in law, because it ties a future Government. There is something slightly undemocratic, potentially, about tying it into law and having a future Government’s hands tied. That said, given that I think this is a really important goal that we should all be signing up to right across political parties, I think it was effective to have it in law, and I was sorry to see the targets go.

Q21            Sir Desmond Swayne: If we are to go beyond income in seeking to measure poverty, what other things should we consider, and how should we weight them?

Helen Barnard: Can I say one thing about what Kitty said about having it in law? I actually think it is empirically clear that the attempt to tie a future Government’s hands did not work. In the years leading up to the abandonment of the target, there is no evidence that the Government felt bound by it, because they did not have a strategy and they did things that made child poverty rise. I agree it is undemocratic, and I also think that it clearly just does not work. The Government of the day, if they are elected democratically, can choose to ignore it, and we kind of watched that.

To answer the question, with poverty measurement, I think that you need to do three things. You need to measure the thing itself; we need to know whether people are in poverty, and most people would define poverty, roughly, as your resources being well below what you need to meet your needs. You need to know how many people are in poverty; you need to know how deep that poverty is; and you need to know how long they are in poverty for—that is a kind of “measuring the thing itself”. Then what you need as well, equally importantly, is a framework that lets you measure the things that we know are the drivers of poverty, the things that we know make it harder to get out of poverty, and the things that we know make your life worse when you are in it—a set of indicators around skills, education, health and family stability.

You have to measure those things, because we want a Government to be held to account for not just how many people are in poverty today, but what they are doing on the drivers. What it does not make any sense to do is take all those things, put them in one bucket, and make them spit out a single number, because they are conceptually different things. We need to measure them separately, because that is the only way you can hold to account what is being done on each of them. If you combine them all, you actually obscure what is happening and make it harder to know, harder to hold to account—and harder to get the credit. If you are a Government doing amazing things on skills, you should get the credit for that. You need a measurement framework that shows it.

Edward Davies: At the CSJ, we consistently talk about five things that we see that would be the other dimension. If we are going to have a multidimensional approach, the five things we would always pick out are worklessness, educational failure, debt, addiction, and family instability. The reality is that all these interplay and can cause poverty, and if we are not tackling those, as you say, future poverty will come. Again, to measure what gets done, the Government have to respond to those five things.

Six out of 10 people in poverty are in work, but four out of 10 are not, and the situation there may be about education, so it is important to measure education as well. We have 5 million adults in the UK who are functionally illiterate and innumerate, so measuring education will be really important to ensure that those people can get on. Likewise, debt is a really big one, in that it does not really matter what your income is if your debts completely outweigh it, so making sure we are measuring and conscious of that is really important; otherwise the income stuff does not matter so much. Those are the five things that we consistently see in our work that come out time and time again, and which I would focus a dashboard on.

Dr Stewart: We heard some really interesting discussion in the earlier session about multidimensional poverty measures and the way they are used around the world. I am very clear that it is really important that in the UK, we continue to do what we have done for more than 100 years. We have a very rich tradition of poverty discussion and measurement, and poverty is understood to be a lack of material resources. I think we should focus poverty measurement on that—on a lack of material resources—and I agree with Helen that that would include headcount measures of how many children we think live in poverty, as well as persistence measures and measures of poverty gap. Then, of course, as part of a strategy to tackle that, we want to think about all of the things that are potential drivers and how we can tackle those, and some of the things that Ed mentions would certainly be in consideration there.

The question was also about other dimensions. We know—the evidence is very strong—that income poverty really has an impact on children’s lives and their development, but of course, it is not the only thing that we need to address. As well as poverty measurement, we should take an approach that thinks about all the other dimensions that children need, and in CASE we have done work on what those dimensions might be. Some of them are fairly obvious: health, education, physical security and so on. Within that, I would advocate the kind of dashboard approach suggested earlier, whereby we think about each of those dimensions separately, measure them separately and publish data on them. We used to have the opportunity for all indicators, which did a sort of job in that way. However, it is really important that we keep poverty measurement separate, and use the language of poverty in the way that we have in the past.

Chair: We have a couple of additional points to raise quickly. First of all, we give a warm welcome to Caroline Johnson, who is on the Education Committee.

Q22            Dr Johnson: Thank you, Chair. I wanted to pick up on points made about the targets, and specifically on how you ensure that target-setting does not have unintended consequences.

I will give an example that Ed used: the four-hour wait in accident and emergency. First, it was poorly understood, because it was actually the time that it took for a full assessment and to leave the A&E department, and not just the time to see a doctor, as was widely reported in the media.

Secondly, it actually increased the amount of work in the hospital, because if someone was getting to three hours and 50 minutes, they would just be moved on—for example, in my case to the children’s ward—where they would then need to queue, potentially for hours, for a children’s doctor to make an assessment, which wasn’t really necessary. All they were doing was waiting, and they could have waited in A&E for another half an hour until nature took its course and a urine specimen, or whatever was being waited for, was provided. It distorted things: there was an increased use of healthcare; increased use of cleaning resources; and increased use of administration. It did not really help, but there was no way of measuring that. When you are setting a target, how do you make sure that the target doesn’t have significant unintended consequences?

Edward Davies: Yes, I think you raise exactly the right point. The point I was making about the four-hour target is not that it was the right target, but that it really drove change. I think that is probably the important thing about targets; what gets measured gets done. They drive change. I think you would have to admit that more people are being seen more quickly now.

The consequence in that example was that there were serious clinical distortions. There were people gaming it; there were fines that came in that caused problems. On your point, yes, absolutely, there are consequences. However, when it comes to poverty measurement, my point would be that if we are not even measuring the things that matter, there will be no response.

The second question is: what might be the unintended consequence of this? For example, in the discussion about relative income, we talk about the unintended consequences—what happens if you just use that measure without other things? There will be unintended consequences. However, thinking those through is secondary. The first question has to be, “What are the things that matter? Let’s set up targets for them.” And then we will look at the targets that cause the least number of unintended consequences.

Helen Barnard: I agree with that. The other important thing is that you need to have a measure that, when it moves, you understand why it is moving. That is one of the things about having a framework of measures, rather than trying to combine everything into a single measure. You need to be able to give an everyday account that an ordinary person would understand about whether the measure moving in a certain direction is a good thing or a bad thing, why that has happened, and how it relates to the levers that you have.

The debt point is really interesting. In our work with the Social Metrics Commission, we moved away from using just income to talking about available resources, because income is one of the resources you have. However, if you have liquid savings, you can use those day to day for living costs, and if you have debt deductions and your other cost reductions, that money isn’t available for day-to-day living costs, so conceptually it makes sense to add in debt to the core measure. It doesn’t make sense to add in some of the other things, because when the measure moved, you would have no idea why it was moving until you decomposed it back to the composite parts.

Having that intellectual underpinning, so that you can give an account of, “If you do x, it will move in this way”, is how a measure will drive good policy. And if you can’t do that, there is no way for it to drive good policy.

Dr Stewart: I echo that, and what Ed said as well. There can be unintended consequences, and we need to think through what they are. The narrower and more complex a measure is, the more likely it is that those things will arise.

When we look at the measure that we had for our target for relative income, we don’t see in practice many distorting effects. As I said, I think it focused us on before housing costs, not after housing costs—and that is a problem, because we should also be looking at housing costs. It focused us on headcount, and meant that some children were missed. Children who had no recourse to public funds, for example, are not really covered at all, because if you are below the headcount, what happens to you? That isn’t picked up, so that is something else.

I guess there were two other critiques. One is that, because it was 60%, there was this focus on poverty plus £1—just moving people across a very arbitrary line. It is very difficult for Government to do that: to practically identify those people and move them across the line, so that may be harder than the example that you used of the waiting lists and waiting hours.

In practice, we don’t see that that is what happened. If we were looking back and saying, “Oh look, they moved people just above the 60% line, but the 50% line and the 70% line got worse,” I would be more worried about having that target. In fact, in this case it is helpful, because it drives a broader strategy that can’t focus very specifically on a small group of people, and it did drive broader change.

The other thing it could do is focus our attention on income and not on other things, such as investment in education, health and so on. Again, in practice, when we looked back that is not what happened. When the targets were taken seriously, we also had broader investment in lots of other services for children and lots of other aspects of children’s lives. Since then, we have seen the opposite. We have had disinvestment both in income support and in other aspects of things that children need. It could in principle happen, so it is another reason to have a broader set of indicators as well, so we are also checking what else is happening for children.

Q23            Debbie Abrahams: I think I am right in saying that the Institute for Fiscal Studies said that, even with the £20 universal credit uplift, the UK has the lowest level of out-of-work support of all OECD countries. Do you think we should be recognising social security adequacy as one of the drivers of child poverty to monitor?

Edward Davies: There is definitely a question here that we haven’t answered for a long time in this country. Helen used the phrase “resources below need” earlier. When you see people going to food banks, I think we have got into talking about food poverty or hygiene poverty, but it is poverty and it is because they don’t have enough money to buy things. We need to make sure that our level of support is such that it is set against the cost of living—that is the reality—so that the safety net is a genuine safety net.

I suppose this is sometimes where I get into a slight differentiation between whether that is a poverty strategy or not. I think that should be a given. The core work of the DWP is to ensure that our safety net is a genuine safety net. I would almost differentiate that from the further things within a poverty strategy, which are slightly different by their nature and don’t necessarily speak to income, whereas the core of the DWP’s work is to ensure that people have enough income to survive when they hit hard times. There are two parts to it there.

Dr Stewart: I think social security support is a very significant driver of child poverty, when we look at the evidence certainly. It is not a measure of child poverty; we need to be measuring child poverty, and I guess social security would come into it as part of the child poverty strategy. How are we going to achieve this? How are we going to get this down? What would our strategy be for this year and the next five to 10 years? That is why it is important to have not just targets but a strategy at all levels of Government to think it through.

Helen Barnard: When we look at the last few years, we can see there was a rising tide of child poverty overall and there were increasing numbers of people pulled into more extreme forms of poverty. Our destitution measures show that between 2017 and 2019 the number of people who had experienced destitution went up by more than 50% in two years.

When we looked at what is pulling people into poverty, there were three main things. There is the fact that too many people are going into jobs that are not only low paid, but insecure and unpredictable, and you don’t get any training and you can’t progress. Too many people were getting stuck in expensive housing—private rented homes, which are overcrowded, often damp and not particularly good quality, but also expensive. The third thing was that social security was weakened. The whole gamut of cuts and freezes meant that the support people were getting from social security was going down and down, and that undercuts some of the benefits of really great policies, like the national living wage.

In a sense, this speaks to what we were talking about: the wider framework. When you look at the overall child poverty measure, whether you use the relative income or the better Social Metrics Commission, you can see the impact of the lack of either a short or a long-term strategy, because you can see the cuts to social security immediately plunge people into hardship.

The failure to have a strategy on housing that increases low-cost rented homes and a strategy around good-quality work, also directly pulls people into poverty and puts a lot more pressure on the social security system, because it is having to pick up the pieces from having markets that aren’t functioning properly. You can see all of that coming together in the few years leading up to covid—and, of course, it was children and families who were paying the price for that whole collection of failures.

Q24            Chris Stephens: Good morning to the panel. Following on from the excellent questions from my colleagues, I have a few questions on using income and debt as a measure. May I start with Dr Stewart first? Dr Stewart, the Government’s 2012 consultation proposed moving away from using income as a measure of poverty, but your own research showed overwhelming support for keeping income at the heart of poverty. What do you believe might be the policy consequences or the social consequences if income was dropped or downgraded as a measure of poverty?

Dr Stewart: You say the consultation showed support for moving away from income—that was the Government’s interpretation of their consultation, but they never published the consultation results in full. With a colleague at LSE, we had a freedom of information request and we managed to get 251 of the 257 responses. We looked at them very carefully: in fact, they show overwhelming support for keeping income at the very heart of poverty measurements.

Some 88% of those responses were very clear that income should remain absolutely at the heart of poverty measurement, even if some people wanted additional dimensions measured as well. Some 1%—two responses out of 251—were in favour of removing income from poverty measurement. One of those wanted expenditure measures used instead of income and one out of 251 wanted to remove it.

There is a different piece of research to be done about the role of consultations in Government policy, because obviously that is not where things went. Those responses came from a wide range of people. They were from academics who studied poverty measurements, from local authorities, from frontline charities working with children in lots of different ways and from NGOs, as well as from individuals who were interested and had their own experience of poverty; there were several of those.

Obviously, I was disappointed to see what happened in the Welfare Reform and Work Act 2016. It was it was a big, major mistake to remove income. It is hard to say how much effect that in itself had. What has really driven the increases in poverty over the last few years is the actual changes taken in the Welfare Reform and Work Act and in other policy changes, which have damaged the ability of social security to protect people when times are tough, as Helen indicated. That has been the real driver. I think not having them has made it more difficult to hold the Government to account and has meant that the Government does not need to come back each year and explain why this is happening and what it is going to do to turn things round.

Q25            Chris Stephens: Thanks, Dr Stewart. I agree with what you said about learning from consultation exercises. How important are relative and absolute income as measures? What can we learn from comparing data from each of these measures?

Dr Stewart: They are both really important. Poverty is a relative concept in essence—poverty is about not having enough resources to be a part of the society in which you live—so it has to be measured in relative terms. One example we have seen over the last year is that children who do not have access to a laptop or an electronic device and broadband have been in a very difficult position to participate in their learning; twenty years ago, no one had heard of broadband. So things change and what children need changes. I should say that that view came out very strongly in consultation—it is widely shared.

So I think that the basic headline relative measure that we use—60% of median equivalised income—is pretty good at tracking over time the way in which living standards change and social norms change. In the short term, it can have problems; it can give us misleading views when the economy is changing very rapidly—in a recession, you can see relative poverty fall simply because median income has fallen—so it is really important to track other measures as well.

Indeed, the Child Poverty Act had the relative measure and what we call an absolute measure—it is not actually absolute, but it is a fixed income anchored measure where we see how the share of people living below 60% of a level of income held constant in real terms. We also have a measure that includes material deprivation, which also picks up some of those aspects as well. We really need to be seeing all three of those moving in the right direction to be making progress.

I know that there has been some sort of shift to saying, “Maybe we should go back towards having the fixed, anchored measure as being our main one. As long as that is improving, that shows that people’s living standards are rising.” But I would echo that, I guess, if we had anchored our measure in the 1980s, we would have no poverty at all in the UK today—every child lives above 60% of the median income in 1985, say. But that does not really tell us anything very helpful and serious about our society today. As I said, there are plenty of things that children had not heard of then that they really need today in order to participate. So it is important to have relative income as the main one and certainly the long-term one, and also to have anchored and material deprivation measures to keep track year on year, just to ensure that we are not getting some misleading results from the relative measure.

Q26            Chris Stephens: Thank you; that is really helpful. Helen, you will be aware that the Social Metrics Commission recommended a measure that accounts for both income and unescapable costs. Your own organisation has developed a minimum income standard which considers what is needed for a minimum standard of living. What role should approaches like that play in measuring child poverty?

Helen Barnard: I think that sometimes when we talk about measurement, we do not relate the measure to the concepts that it is supposed to be operationalising, and that can lead to confusion. This speaks as well to relative versus absolute, which is obviously not absolute; it is just relative from 10 years ago.

So if you think about what is the purpose of that line in the measure, essentially it is a proxy—a benchmark—for social norms. If the concept of poverty is that people at the bottom should not fall too far behind what is normal in our society, because that leads to hardship—you can see that, as Ed said, in foodbanks, people skipping meals and so on—you have to think about what is the best way that we can track social norms, because they change over time. You cannot have it the same.

We have seen through the minimum income standards work that what those groups of the public defined as normal and decent—not luxurious, but not problematic—standards of living have changed over the decade of that research. They definitely change, but they do not change every year. You do not want your benchmark line changing every year, because that does not represent what really happens in society, which is that people’s expectations and our way of life go either up or down over time.

With the Social Metrics Commission, we spent a long time talking about how we do that. How do we have a line that will change over time as society changes but will not fluctuate every year in a way that does not relate to reality? Where we eventually land is that we will say, “Okay, we will use that median total resources line. What do people in the middle of society have available?” We will say, “As that changes over time, we will use a rolling three-year average.” Each year, you take the average of this year and the previous ones.

What that means is that if you have a big recession, living standards across society fall. That feeds into expectations and social norms, but it does not do it straight away. The line slowly changes with society, and you measure, each year, whether people are falling too far below that, but you know it is because that is the proxy for social norms. You can check on it with stuff like minimum income standards, but minimum income standards cannot substitute for that line, because that is not what the research is actually doing.

Q27            Chris Stephens: Thanks for that, Helen. That probably brings me to Edward. Edward, I have a question for you, particularly in relation to debt. Is there a relationship between debt and child poverty, and is it linked to income? 

Edward Davies: Yes, absolutely. On the question of debt, the reality is that if your outgoings supersede your incomings, you are going to be in problems. The way that we see it most in our work at the moment around universal credit is that, actually, we do not have lots of people coming to us saying that the amount in universal credit is utterly problematic. It is for some people, but it is not the big thing.

The big thing is when you put debt on top. At the moment, it is 30%; I think it will go down to 25% as the maximum you can have clawed back. But if you are already on a tight income, that will put you into real trouble. What is sustainable is very much brought into reality by the size of your debt. Absolutely it does matter, and we should be measuring it and looking at it closely.

To come back on a couple of the other things that were mentioned about relative income and minimum income standards, I probably take a slightly different view on that. I think that the relative income stuff is more problematic—I am less enamoured by it. There are three major reasons for that. The first one was touched on: we have probably just had one of the best years ever for child poverty by that metric. In terms of pushing Government to do the right things, quite clearly big recessions lead to reductions in inequality and perverse outcomes. It has problems with it that are really quite large. In and of itself, that isn’t great.

The second thing is that it becomes the be-all and end-all. On this point about having a multidimensional dashboard, the reality is that it seems to have become the default. We do not talk about other things; we just talk about this. That is really problematic, because we are not talking about all the other factors that may be involved.

That leads to the third point, which is that I just don’t think it is very ambitious. A poverty metric that is really ambitious for people—yes, money matters, absolutely. But actually, I do not want them to be dependent on the amount that their universal credit is. I want them to stand on their own two feet. I want them to get a job. I want them to have healthy relationships. I want them to not be in debt. I want all these things. A strategy will come from a poverty metric that has these things at its heart, so it really distorts what Government does when it is the main focus. Having it as part of a wider thing become really important.

Q28            Chris Stephens: Thanks, Edward. Maybe I could ask Helen and Kitty about their view. Is debt a pathway to poverty—a cause—or is it an indicator of poverty?

Helen Barnard: With most of these things, it is both. That is true when you look at the framework of indicators that—Ed is right—you absolutely have to track. That is why the Social Metrics Commission has poverty now, debt persistence and lived experience indicators, and you want to track all those. The main reason why people on low incomes get into debt is that they haven’t got enough money to cover their outgoings. When you look at people right at the bottom, the debt they are in tends not to be consumer debt. It tends to be debt to public authorities. The debt you see at the bottom is people paying back universal credit advances, council tax debt—massively—and it is utilities.

It is that kind of debt that people at the bottom get into. It is because they have not got enough money to pay for their outgoings. It is definitely a consequence of poverty. It is also very rare to have somebody who is doing completely fine, is thriving and has lots of money, and debt itself is the thing that pulls them into poverty. There might be a few but it is not a major part of poverty.

However, once you are in that situation, it makes it much harder to get out of it again. You have built up these debts, rent arrears and so on, because you have not got enough money to pay your rent. Then, if you do get a bit of a pay rise or whatever, you are still paying back the debt.

We did have a go with the Social Metrics Commission: could you divide things into causes, consequences and things that make your life worse? To be honest, I gave up because almost everything falls into each of those three buckets. We should just accept that debt is one of those that you put into the core poverty measure because, if you are paying back debt, it reduces your resources available for other things. That is the one that you really should be wrapping up with it. The others, you track along the side, because they are consequences and they make it worse and harder to get out.

Q29            Chris Stephens: Thanks, Helen. Dr Stewart. Anything to add?

Dr Stewart: I agree with what Helen has said. I am slightly worried about complicating our pure income measure—not so much the debt, but other aspects of what the Social Metrics Commission does can make it harder rather than easier to see what is really going on.

Certainly, the public debt is already taken out of our income measure, where people are paying back their universal credit. Maybe there is a case for taking that out, taking out the private debt as well.

Can I briefly come back on what Ed said? As I said, there can be a short-term issue with the relative measure, but that is not a good reason to throw the baby out with the bathwater, so to speak. It is better to have other indicators as well, so that we can see. I do not think it is true that it takes over and that we do not look at the other measures.

Just to underline again, we know that the evidence is very strong about the causal effects of income poverty itself on children’s outcomes. Having that central is very important and it is very ambitious. We are seeing child poverty levels rising as things currently stand.

Q30            Chris Stephens: Thanks for that. I am fascinated that you have all mentioned universal credit deductions as an issue. I found that interesting. I think that is something we can come back to, Chair, at a later point.

I want to ask one quick topical question of each of you. The suggestion is that today’s Budget is going to maintain the £20 uplift for six months. How would child poverty be affected if the uplift was not made permanent?

Helen Barnard: At the point the uplift disappears, whether it is in April or six months, it will affect about 6 million households. We think it will pull around 400,000 people into poverty, with a good chunk of children in those. It is £1,000 drop in annual income overnight; whether that happens in April or six months, it is going to be pretty devastating for people.

Q31            Chris Stephens: Does any other panel member have a view?

Dr Stewart: Very briefly, I absolutely agree. We also need to remember that the universal credit uplift is the same amount—£20 per household. It goes less far for a family with two, three or four children, than for a family without children. What we have not seen at all over the last year is a focus on child poverty and on what families with children need, so an increase in child benefit or in the child element of universal credit would be really helpful, as well as the £20 uplift.

Edward Davies: I would add that, yes, I would keep the uplift, and I think there is a really important point in there about how welfare has been whittled away in recent years, and keeping it would be a good thing. This is another example of how we have focused a lot on the financial and the financial loss. The Ofsted report that came out after the first lockdown looking at what had happened to children and children’s welfare during the lockdown said that even controlling for deprivation, the thing that mattered more was the support structures that they had at home. That stood true even for children in care; if they had good support structures, they could weather the storm over the last year quite well.

I will come back to the point that, yes, this does matter—the money does matter—but this is why we have to have the broader dashboard looking at this other stuff and the support structures around children: when crisis hits—it will not be a pandemic next time; it will be something else and it might be just that one family—having those strong support structures matters as much as deprivation. It is not me saying that; it is Oftsted, the education people, so just to be clear, there is some really good evidence on that out there.

Q32            Selaine Saxby: I want to talk about worklessness. I know that we have already touched on some of that this morning, but given the changes in the labour market and low-paid employment—we are all aware that a large number of people in poverty are working, either part time or, in many cases, full time—how would you incorporate worklessness into a measure of poverty and child poverty?

Helen Barnard: I see worklessness and indicators of the quality of work as things that you would have in the lived experience indicators that you would be tracking alongside the core poverty measure. If you are out of work, you have a much higher risk of being in poverty and in deep poverty. As you said, the big growth in child poverty in recent years is among working families, and we are seeing more and more people—particularly parents—pulled into very poor-quality work that does not take them out of poverty, and actually often brings a whole load of mental health and other problems along with it.

There is also the unpredictability and the insecurity of work. When I talk to people who are workers in poverty, they talk about the pay and so on, but they also talk about that insecurity, and during the pandemic we have really seen how that impacts people. I would say that there is a cluster of labour-market indicators that you need to track, including worklessness, in-work poverty, the quality of jobs that people can get, and things such as progression. Being in a low-paid job may be okay for a little bit if you are just starting out; it is not okay if 10 years later you are still there and trying to raise a family, still in a very low-paid insecure job with no chance of getting anything else.

Edward Davies: The importance of measuring worklessness will be found in the responses to it as well, once you have put that on the table as a fundamental “This is something we are measuring”, whether we decide to measure long-term worklessness or insecure work. The Government then has to come up with an answer. When they look at those problems, they will come up with answers, and I suspect that a lot of answers will be around things such as childcare as well. A lot of people are struggling to have their kids looked after if they ned to go to work—there is the complexity of people’s lives in there. It comes back to actually just measuring it, and making sure that that is one of things on the dashboard, because then the Government have to come up with answers if it is not going well.

Dr Stewart: I maybe see it slightly differently. I certainly do not see it as a measure of child poverty; I do not think it should be in the child poverty measure. I see it as being part of the strategy to tackle child poverty. That is obviously a crucial thing to think about in terms of how to reduce child poverty, and it involves thinking about what the barriers are to people’s employment, which may be things such as childcare, as Ed said,  but it also involves thinking about why we have so much in-work poverty and why work is not proving to be a solution. Why are we seeing child poverty rising, particularly in households with three or more children, and why is poverty rising in households where two parents are working full time? How can that be the case, and where are we going wrong?

It would also focus on thinking about the risks of poverty for households where no one is in work, because the vast majority of those households, certainly on the eve of the pandemic, were households with a lone parent and a young child, or where a family member has a disability. Given the fact that worklessness is so associated with poverty, there is also a political decision about how we support households that, for one reason or another, are not able to access paid employment. Those are all things that one needs to think about as part of a strategy in order to reduce poverty, understood as a lack of material resources.

Q33            Dr Johnson: I wanted to ask about children’s education and, in particular, what you see as the links between poverty and educational attainment, particularly as we look towards catching up children. We know that some children from deprived backgrounds have struggled more during the pandemic for a variety of reasons. What do you think the links are, and how should we measure them? Perhaps we can start with Helen, since she is nodding.

Helen Barnard: The single biggest predictor of attainment is your family circumstances, which include poverty and deprivation. The ways in which that affects children’s lives come together around lots of different things. If you are a child and you are living in a home that is overcrowded and damp, and is not a secure place for you to be, and you have parents who are struggling with their own mental health, partly because of the stress of poverty, hardship, debt and having to skip meals, children pick up on these things. Most parents do their absolute best to protect children from that, but actually children are smart and they see that happening. There was someone just this morning taking about his memories as a child of having to be very, very quiet because his mum was working out what they could not afford anymore, so what out of their essentials they were going to be knocking out. He remembers the feeling incredibly strongly of watching his mum do that.

Then you have what you can participate in in the school day. As you go through school, there are all sorts of ways that, if you do not have a comfortable income, you find yourself shut out of things. That can be about activities, bringing things in from home, school trips, sports—it is all these things. It is made very clear to children in poverty in their school experience that they are not coming from an equal start to everybody else.

Then you also have the fact that disproportionately children from poorer backgrounds end up in less good schools—not in London, actually, which is a bit different, but outside London. Any system in which there is an element of parental choice and of schools choosing kids will always, as far as I can see from the analysis, disadvantage kids from a poorer background. Even things like whether you will go to a school that is slightly further away because you think it is better academically, if you do not have the money to spare for that, you are probably not going to make that choice. Also, the schools choose, and quite often they try to choose kids who are going to do good things for their league tables.

All these different factors come together in children’s lives and mean that poverty massively impacts on attainment. Actually, there is now a pretty good evidence base about what you can do to reduce that, and part of it is reducing poverty. I always think that you pour money into schools trying to fix this, but it is like they are swimming upstream if you are letting child poverty increase. There is lots of evidence that shows that if you reduce poverty outside in people’s homes, and you have interventions in schools—the Education Endowment Foundation has a wonderful toolkit that tells you the best value for money for those—you can make an incredible difference to children’s outcomes, but it does take a lot of focus on that. It is not something that happens naturally within the system.

Edward Davies: To add to that, the home environment is everything, really. We pile so much on to our schools at the moment, but kids spend 84% of their lives outside school. An awful lot of what is going on is happening at home. Poverty is part of that, absolutely. I note that the Education Committee is currently looking at why white working-class boys do so badly. One of the major themes coming out is that, by and large, they don’t have dads. Looking at the breakdown by ethnicity and income of the different patterns that you see emerging across society, if you are born in the top 20% as a white boy in the UK there is an 84% chance that you have married stable parents, and a further 10% have co-habiting stable parents. If you are born in the bottom 20%, it is 19%. The fundamental difference there in terms of your life chances—basically, if you are born in the top 20% you will probably get a dad, and if you are born in the bottom 20% you probably won’t—brings up the whole question of really tricky, sensitive, personal stuff. Actually, it really matters that we talk about how family structure and stability impact the ability of kids to go to school and learn.

NHS figures show that it has a huge impact on mental health problems. About half of children—it is by far the biggest factor—attending CAMH services cite family relationship problems as the reason they are there. As I mentioned earlier, Ofsted cites the stability of homes as more important than deprivation now, so as difficult as it is we have to go there in our metrics, and we have to understand that—in the previous panel, we heard about it internationally—we need to respond to what is going on in our own country. We are a total international outlier on this. Our rates are worse than anywhere else in the world. There are one or two that come close to us. Addressing this will be really important as part of an education strategy, a poverty strategy and a work strategy.

Dr Stewart: Part of the reason why low-income children do less well in school is because of their low income. I think that is really clear. We did a systematic review of the evidence on the impact of income on children’s outcomes a few years ago at CASE, looking only at studies that use quite rigorous causal methods. The evidence, particularly for cognitive attainment and school achievement, was really strong that money itself makes a difference to children’s outcomes. I think Helen highlighted the reasons quite well. It is partly to do with what money buys in terms of books, resources, healthy food, better housing, warm housing, and so on. It is also to do with what it means for the emotional environment in the household, and the levels of stress and anxiety among parents and what that means in terms of their ability to focus on children’s needs, listen to them, help with homework and so on.

There are huge assumptions here, but the figure that we came up with was that if you got rid of the income gap between a child on free school meals and a child on a median income you would get rid of about half of the attainment gap. It would not be the whole story. There are other things going on, but it would be a big part of the story, and I think that what is really important to underline is that—I do not know Ed’s work on fathers and so on—it is much harder for the Government to give a child a father than it is to make sure that they have an adequate income. Although it is not the whole story, we really need to remember that it is a big part of the story, and it is one that Government have much easier policy levers than some of the much more difficult things for us to get at.

Q34            Dr Johnson: I am on the Education Committee and am a guest on this one today, because of the topic that the Work and Pensions Committee is doing. The free school meals and university entrance evidence that we were given is quite interesting, because it shows that, yes, children on free school meals do less well at getting into university than children not on free school meals, but for some reason white boys who are not on free school meals do less well than children of other ethnicities who are not on free school meals, suggesting that it is not only income. Do you think that it would be misleading to measure educational attainment as a marker of poverty? Do you think it is a separate marker in and of itself?

Helen Barnard: Yes. What we want to understand is how many people are poor, how deep is the poverty, how long are they in poverty and, for those people who are already in poverty, what that means for all their other outcomes and, therefore, their futures. You need to do both—you wouldn’t just chuck them in together.

On family structure and stability, it is also worth saying that poverty, family structure and so on are not completely separate. They are not happening in a vacuum. Relate research says that the leading cause of arguments and stress between couples is money and debt problems. It is a contributing factor to family breakdown in and of itself. Once that happens, the fact that we have incredibly high lone parent poverty in this country is a consequence of policy choices. It is a consequence of the fact that our labour market is terribly designed for allowing lone parents to get and keep a good job. They get stuck in the bottom end of the labour market.

Our social security system includes design choices that lead to very high levels of poverty. We can make different policy choices and, elsewhere in the world, there are different policy choices that mean that lone parents are not disproportionately poor in the way that they are in this country.  

Chair: Thank you. Steve McCabe?

Q35            Steve McCabe: I think Edward started to hint about this earlier. I want to ask about family stability. The Government said in its consultation paper that it considers a measure of family stability would give a more realistic indication of what it means to grow up in child poverty. Do you think family stability should be a measure? Edward, should I start with you?

Edward Davies: Absolutely, yes. It is one of the five big things that we talk about. It is one of those things, as Helen said, that interacts with all the others and to try to treat it separately is probably not the right thing to do. It needs to be part of the metric. The reason we don’t is because it is really hard. Exactly as Kitty said, how do you give a boy a dad? But there are things you can do. If you put it into the metric, it encourages the Government to think about those things you can do to give family stability.

Q36            Steve McCabe: I don’t understand, and I noticed quite a few of the Child Poverty Action Group and the Social Mobility Commission didn’t understand either, what, exactly, family stability means.

Edward Davies: Basically, when a child is born, about 90% to 95% of kids are born with two parents. Then, in the first five years of that life, and if you have ever had kids you will know why, it gets really hard, so you tend to find that a lot of instability comes in those years. That is not inevitable, and money is a part of this. As has been said, you need to make sure that people have enough money to survive, but actually, a lot of other factors can bring stability: help with parent support; relationship support; family hubs, which we have talked about before; and using birth registration as a point to intervene in people’s lives. There are all sorts of small things you can do to help. I have mentioned our fatherhood work. Fathers often do disengage in the first year of a child’s life. There are some small nudge things that would cost nothing to encourage dads to engage with their kids in the first year of life—just making sure that they are invited and free to go to perinatal appointments, and things like that. There is a whole range of nudge things you can do to make those first five years easier on parents to ensure the child has a stable home environment for the first years and onwards for the rest of their life, because they are so important. If we measure it, again, the Government has to ask what we can we do to make this look better, and the outcomes will be better.

Q37            Steve McCabe: Do you fear, at all, that there is a danger that making that a factor—I do not dispute what you are telling us—is a convenient way of stigmatising lone parent families and disguising issues about benefit payments?

Edward Davies: No, I absolutely don’t. It is really difficult, and I totally accept that this is sensitive subject matter, because when we talk about family breakdown we are talking about my family, your family and everyone else’s families here. The fear of upsetting people has led us to not have a frank conversation about the effects this has, particularly on children. I come back to the point that we are talking about child welfare and child poverty metrics, not about adult freedoms and choices. Particularly in a setting like this, we need to be really clear about the effect it has on children when they experience instability in the home. We need to accept that it is going to be awkward and difficult, but if want to make the right policy and measure the right things so that children have the best future, those metrics will be a really important part of it.

Q38            Steve McCabe: Thank you. Helen, what is your view on this?

Helen Barnard: There are two factors I would want brought in to how we measure poverty. First, the basic thing is that for the core who are in poverty, we need to know what kinds of family those people live in, because we need to see which groups are disproportionately stuck in poverty. So, you put that into the demographics. I also agree that in the lived experience indicators—the things you are tracking that are not poverty itself, but you know are really related to it—you need a cluster of relationship support system measures. For me, that is less about family structure actually, but more about family stability and things like social isolation. When you look at the research on child outcomes, the thing that often makes the difference is not the event of a couple staying together or being apart, but what was the quality of relationships leading up to that and what is the quality of relationships after it, and how that affects the economics.

Q39            Steve McCabe: I understand what you are saying, and fair point, but is it easy to measure that sort of thing?

Helen Barnard: It is not easy, but it is entirely possible, and there is quite a big source of research literature that gives you some indicators that you can use for that purpose. Research suggests that if you take a family, all other things being equal—I know as a parent that having two of you around is a lot easier than being on your own with a child; fundamentally, it is—a child can thrive in any family situation if there are good relationships between the child and the parent; if there are good relationships between the two parents; if there are two parents around, whether they are together or not; if they have wider support structures; and if they are economically not going without things. You can achieve all that.

When you look at child outcomes, if, after a family breaks up, the family bounce back economically and don’t get stuck in poverty, positive parent mental health is supported and the involvement of both parents, whatever that is, is supported, children do okay; they do fine. The trouble is that for too many families that is not what happens. The parents break up; the resident parent and child are plunged into poverty; they do not recover because the parent can’t work full time and the social security isn’t helping. You then get strained relationships, and they don’t have a wider support system. You need to measure all those things, because we need to bring them together to give those children the right start in life without, in my view, making a judgment on what their family structure should be. They need good, positive adult relationships around them, and then they will do fine.

Q40            Steve McCabe: Thank you. Kitty?

Dr Stewart: I would echo that. The evidence is very weak that growing up in a lone-parent household has a negative impact on children’s outcomes. When you control for income in particular, that seems to be the mechanism. When you compare the literature showing the impact of lone parenthood with the literature showing the impact of income, the evidence is far, far stronger that it is poverty and low income that make a difference. To equate those things is a mistake and is not true to what the literature tells us. That is not to say that it is not really important to do all the things that have been suggested to provide support and help families to stay together. There are other reasons why that is important, but the idea that lone parenthood is what is damaging children is not borne out by the literature.

Q41            Steve McCabe: I acknowledge the point that Helen and Edward have made about stability being the issue rather than structure, but I was struck by the fact that the Government cited as part of their argument for stability the fact that after a marital split the income of women with children falls on average by 12% but the separating father’s income increases by 31%. Isn’t that a comment on the child maintenance service, and the welfare system, rather than anything to do with a measure of stability?

Helen Barnard: Yes, it is. And it is also a comment on how hard it is to balance caring and work, because of the design of the labour market.

Q42            Steve McCabe: The reason I ask is that, would I be right to conclude that while you all agree that stability is a factor that should be considered, we have to be very careful as to talking about structure and failing to talk about weaknesses in the welfare system? Is that a reasonable conclusion to draw?

Edward Davies: I would probably go a little further than the other guys and say that if you are talking about stability, there are more stable structures than others. That is when this gets really awkward, because there just are. Statistically, there are reasons why, when people get married, they tend to stay married, and cohabiters tend to break up more often.

The thing is, as soon as we start talking about stability, I think there is a fear of going into that territory—that marriage makes a difference—but actually, part of the problem that there is not great data around it is because we do not do a lot of research around it, because we find it difficult to talk about in the UK. However, there are a couple of really good bits of research happening in the US, where they look at issues like the level of equitable commitment being really important. The reality is that in a lot of cases, blokes, I am afraid, tend to walk out because they are not as committed as women, and that deciding, not sliding—having a stated, “This is what we are committing to”—tends to be really important. We call it marriage, because that is just what it is called, but those two factors of equally committing and equal expectation do appear to have an effect on the stability of that relationship, so you get in a position where, if stability matters, you cannot pretend that the structure does not, because it does.

Helen Barnard: There is some evidence in the UK. The Institute for Fiscal Studies did a very good study looking specifically at the impact of marriage, and what they basically found was that the issue is actually commitment, as Ed has said. The issue is, “Are both people in the couple committed to one another, and is it a longterm relationship?” Marriage per se does not really make much difference once you control for the fact that it tends to be richer people who get married in the first place. It is the commitment of the couple, and poorer couples tend not to get married because they are poorer, and because of other factors. The marriage itself does not seem to have a causal impact. Yes, you can say, “Life is harder when you are a lone parent,” in lots of ways, but it does not have to determine that you are going to be in poverty. That is a choice we as a society can make, to say that is not going to be the case.

Q43            Steve McCabe: Thank you. Kitty, did you want to make one last point?

Dr Stewart: I would perhaps go further in the opposite direction: this would not be on my list of top things that I think we need to focus on, in terms of ensuring that children have everything that they need, partly because I think the evidence is not strong enough. I also worry—as I think you hinted at the beginning—that it is stigmatising, and it sort of shifts the sense of responsibility for the fact that some children are not doing as well as they could on to parents who are already experiencing high levels of stress, and coping with difficult circumstances and low incomes. I am worried about putting it up there as a key driver, given that the evidence is weak and given these potential stigmatising effects.

Steve McCabe: Thank you.

Chair: There is one final area we would like to ask you about. Siobhan Baillie has a question.

Q44            Siobhan Baillie: I will be quick. Can you talk to me about the evidence for the link between disability, ill health and child poverty, and within your answer, could you also let us know whether you think disability should be taken into account in measures of child poverty? If so, how can that be done?

Helen Barnard: Yes, disability is massively linked to poverty. Overall, if you take everyone who is in poverty, around half those people are either disabled themselves or live with someone who is disabled, so disability is an enormous portion of poverty.

In terms of how that manifests and how it can be linked, there are a few different things. One is that being disabled carries extra costs of lots of different kinds. It is more expensive to be disabled than it is to be not disabled. Obviously, within the benefits system, we have benefits that are designed to deal with those extra costs—so, in terms of how we measure, you would ideally want to account for the fact that you are having to pay more out to cover those extra costs. Technically, it is quite hard to do that yet, so for the Social Metrics Commission, as a kind of halfway house, we take off the money you get for extra costs, because that is not raising your standard of living; it is just helping you to account for some extra bits. So, disability brings costs with it.

The other way it impacts is through the labour market. We know that disabled people have much lower employment rates. There has been a bit of progress on that, but not much. Many disabled people want to work, but jobs are not designed for them, employers are often reluctant to employ them, and you need a lot more support than we have.

Of course, the other thing that has happened is that there have been various reductions in the support that disabled people get through the social security system. When we looked at this last year with JRF, we found that poverty among disabled people who are on disability benefits has got worse and has also deepened. When you talk to disabled people, their lives have got massively harder in recent years because there is less and less money, but also because the way the system is designed is incredibly stressful for people who are already, by definition, coping with health conditions.

So, there is a big piece of work to do to open up employment for disabled people and overcome the stigma and discrimination that they see; open up better-quality employment, because disabled people are paid less for a given level of skills than non-disabled people; and have a social security system that gives people that support but also treats them in a dignified way and does not increase their distress and health problems due to the nature of the system they are being forced to engage with. That is too often what you hear at the moment from disabled people.

Edward Davies: I would probably agree with most of that. The bit I would draw particular attention to is the employment stuff. I mentioned at the beginning that there are the five things that we focus on at the CSJ, and there are lots of other things that come into it beyond the five—disability, mental health, health and housing are all part of it—but where we see disability most tends to be in the employment area. A huge number of disabled people want to work but cannot, for various reasons. Actually, closing that gap in employment should be a major focus. Again, going back to what we measure, if we were to measure worklessness, I would want disability to be a huge part of the response to that, because at the moment the gap in employment between disabled people and everyone else is way too big and needs really focused attention.

In fairness, I think the Government is reviewing this at the moment, so I would be interested to see if anything comes out of that. It is certainly something I would push at the moment, quite hard.

Chair: The Committee is looking at it as well.

Dr Stewart: In terms of measurement, I would simply measure poverty separately below the headline measures that look at poverty in households with a disabled adult, with a disabled child and with both, and use those to inform a strategy. I agree on the different elements highlighted in terms of extra costs, employment and social security support.

I disagree with Helen on the Social Metrics Commission’s approach of removing the extra costs disability benefits from their measure of income. I know why that is done. As Helen said, the best thing to do would be to adjust for those costs—essentially to equivalise for the needs of households in the way that we do for children—but it is very difficult, almost impossible, to imagine how we would do that effectively. We cannot at the moment. The extra costs benefits are taken out of their income measure because those are there just to help with the extra costs, which kind of makes sense.

Certainly, if you are comparing a household with a disabled person and a household without a disabled person, it makes sense to do that because otherwise you underestimate the real levels of poverty in households in receipt of disability costs benefit. But, of course, if what we are doing is tracking over time, that is really problematic, because you do not see the effect.

One of the ways in which the Government can support households with extra costs is through those benefits. On the Social Metrics Commission, for example, you see a very flat line for households with a child with a disability during a period in which those benefits actually increased in real value. So, for those households, there will have been a real material change, and we lose that. That is a specific example, but it is also an example of why we need to be wary of messing around too much with the income measures, because it can make things a bit less transparent in terms of what is happening.

Chair: That concludes our session. Thank you all very much indeed for the evidence that you have given us in a very helpful set of answers. There is lots for us to think about as we set out on this inquiry over the next few weeks. Thank you all very much for being willing to speak to us.

I must make the point, as we normally do, that if other thoughts occur to you after the meeting that you think we ought to be aware of, please do send us an email and draw our attention to anything that we perhaps missed this morning. That concludes our meeting. Thank you everybody for being with us.