In a slightly longer post than usual, Professor John Veit-Wilson has kindly given us permission to publish the first part of his response to the government consultation on child poverty measures.
Guest post by Professor John Veit-Wilson, Newcastle University
Introduction: What is the nature of the consultation?
This consultation is not about how to measure child poverty but about how to describe it better. Mr Duncan Smith has stated clearly in the House of Commons that the decision on ‘measurement’ has already been taken — “Income will be part of it, but not the dominant part” [Hansard 21.1.13 col 130]. Similarly, an informed report stated that —
the key aim of the new measure was to give a richer picture of the experience of poverty. This will enable the Government to look at different combinations of problems experienced by low income children aiding understanding of problems and facilitating better responses.
The status quo will not be an option: even if there is disagreement on whether a new measure is a good thing, our understanding is that it is going to happen and Commission will want to advice [sic] the Government on how best to do it as well as consideration of wider issues. [SMCPC staff members quoted in minutes of meeting with academics 18.1.13.]
This consultation is therefore simply about how best to present that fact to the British public. It is an exercise in market-testing the public acceptability of the predetermined message and not an enquiry into different types of measurement.
The consultation document does nevertheless present it as being about measuring child poverty. In their Forewords, the Secretary of State and the Minister of State make it clear —
- that they simply seek public agreement that child poverty can best be measured in terms of descriptions of the characteristics of families in poverty;
- that they do not believe that low incomes are a sufficient description of what it means to be poor;
- that they want to count the number of children in poverty in terms of characteristics which reflect some of the outcomes, the consequences for children of the lack of resources at the household and community levels;
- that they believe that the personal characteristics of the poor children’s parent or parents are among the criteria which distinguish poor children from children who are not poor.
So either the DWP and DfE are genuinely confused between measurement and description, or they are being deliberately misleading. This is a serious matter.
Two examples to illustrate the confusion.
Personal characteristics distributed across the wider population cannot be used to distinguish children in poverty from those who are not. This is obvious if we consider a couple of common examples, drunk driving and flu. Identifying and counting all the bad drivers is not the same as measuring and counting drivers with excess blood alcohol. Similarly, flu victims suffer many forms of malaise, but counting all the people with headaches, pains and temperatures will not identify those infected with flu virus. If a government cares about excess alcohol consumption or preventing flu, it needs to identify the causal agent not the consequent condition and behaviour.
A simple way of testing whether or not the symptom is a criterion for measurement of the phenomenon in question or even a description of it is to ask, if you abolished the symptom would you have abolished the phenomenon? Even if no one appeared to be driving badly, it would not prove that they did not have excess alcohol in their bloodstream and be liable to accidents. Similarly, taking analgesics against headaches will not cure flu but simply mask the symptoms.
The Policy Exchange, which published the original proposals for descriptors of child poverty on which this consultation document is based [in 2002 and 2011], has recently suggested adding such descriptors as children who have been in the care system or whose parents have a criminal conviction . It really is remarkable that intelligent people can confuse the phenomenon of poverty with so many stereotypes of social problems which are found right across the social spectrum, some of which may be exacerbated by poverty but have no necessary connection with it. They may not even be the outcomes of poverty or deprivation which the Policy Exchange authors assert. This is muddle or mendacity.
Muddle or mendacity?
Are the DWP and DfE (and Policy Exchange before them) confused between measurement and description, or are they are being deliberately misleading? Either explanation suggests that the capacity for serious enquiry for policy making has broken down under this government. It looks instead as if the government is starting from a position of deliberate and insouciant ignorance about a well investigated and documented subject. There is a large existing body of both UK, European and international material on measuring poverty, built up over many decades, which has been disregarded by those who drafted this document, if they were even aware of it in the first place.
If this weren’t simply a marketing exercise, the government would have consulted on how to apply what is already known to improving the current statutory poverty measure. For instance, the DWP carried out a similar consultation on measuring child poverty in 2002 and the government could have studied the evidence submitted then. Many of the responses to the current consultation will probably repeat what has been publicly said many times before, though perhaps this time it is to readers to whom it is largely unfamiliar and new.
Since the government is already using public opinion to defend and justify its confusion between measurement and description (as in the DWP polling mentioned below), and since it is already promoting its decision as a better way to measure poverty and not merely to describe some of its characteristics, no response would be complete without an explanation of the basic facts and methods to those who have to read the responses and who may have to summarise their contents for politicians, officials and press officers.
I’m therefore responding in simple terms (as far as possible in a complicated subject) and without the usual scholarly paraphernalia. All that evidence already exists somewhere in the DWP or is readily available elsewhere to anyone who genuinely seeks it. Other experts will certainly refer to it, so the government can’t claim any lack of information or expert advice about what is known on the topic of measuring poverty in the UK.
Confusion between describing symptoms and diagnosing causes.
Of course many respondents will agree with Mr Duncan Smith and Policy Exchange that the conditions and characteristics which exercise them and which the document mentions may be found among people in poverty. The DWP report of the telephone polling on Public Views on Child Poverty [January 2013] shows respondents agreed that they associate certain conditions with ‘growing up in poverty’, just as they would agree that bad driving is associated with excess alcohol and headaches with flu.
But however strongly the public agrees that it associates certain deprived conditions with poverty, that does not make the conditions those things which identify poor people and distinguish them from people who are not poor. A moment’s thought shows that addiction and family breakdown are sadly shared experiences right across society from the poorest to the very richest people. It has been said that superfluous wealth may be a cause of both addiction and family breakdown, as so many rich people seem to suffer from both of these conditions.
Conditions such as addiction or family breakdown are known to be caused by a variety of factors, but while too little or too much income may exacerbate them, they cannot be used as indicators of poverty — or if they could, then equally as indicators of superfluous wealth. Nor can the strength of public agreement be taken as a defence or support of the government’s confused or misleading position, just as the world is not flat even if 90% of the public agree that this is how it looks. Child poverty is far too serious a problem to be treated so ignorantly.
It really is very important that the government does not pretend that public support for what it calls ‘dimensions’ is taken as any kind of scientific endorsement of their use as part of a poverty measurement tool.
The telephone poll, like the consultation, shows that the DWP approach only makes sense if understood as a market testing exercise. If people respond to market research surveys that they believe green food is more nutritious than brown food (whether or not it actually is), the manufacturer will dye the food green in order to sell more of it. Endorsing the government’s dubious political marketing practices is not what social scientists’ and research methodologists’ roles are about. But since the consultation document claims to want to improve measurement, the next section outlines what the government would have to do if it actually wanted to be serious about it.
Measuring poverty — defining or describing?
To measure the quantity of something you need to be able to identify and count it. Identification means ‘what this is’ which is ‘not that’. Take a simple example. You can’t count men in a crowd unless you can distinguish them from women. You can describe the clothes which men might wear, but women might wear them, too. You will then have counted all those wearing trousers for example, but not have distinguished the men from the women. If you use skirts as your criterion, you have the problem of counting kilts. Or you could use height or any other characteristics which may be shared by men and others. But in each case you run up against the same problem — you are not using the single criterion which separates men from women so that you can count men.
The problem with counting children in poverty is just the same. You can describe all the characteristics which children in poverty have, such as are listed in the document with an invitation to supply more, but unless you find one which children not in poverty do not have, you will not have found anything which helps you to count children in poverty. Thus when the document asks for better ways of counting children in poverty, what it is in fact doing is asking for better descriptions of the lives of children who already experience poverty. That is what the public opinion poll supported. It is not what the consultation ostensibly claims to be about.
If you want to use descriptions like those in the document as the means for distinguishing poverty from non-poverty (as is implied by the document), the result will be that millions more children in the UK will be counted as being poor (as even the latest Policy Exchange document emphasised). That may or may not be true in other senses, but it is hardly what the government wants to show. For example, if you want to describe men as being those who wear trousers and you count all the trouser-wearers, you will have vastly more ‘men’ in the population (but miss the kilt-wearers) and fewer women than is the case when more reliable methods of distinguishing men from women are used.
So if you count children in poverty by such matters as broken families or parents who do not have paid work or suffer addictions (and many other similar aspects mentioned by the document) you will count children right across the whole population, even those in the Royal Family. This will not tell you what you want to know.
I hope this elementary point is clear, because it affects the entire document and possibly also responses to it. There is a solution, and it is this — the one thing that has always been known and tested, which distinguishes children in poverty from children with similar problems and characteristics but who are not in poverty, is that their parents lack sufficient household resources, chiefly income, to enable the family to cope with the problems and achieve a normal socially-included life according to society’s usual standards.
Lacking power over resources to achieve a minimally adequate level of living and that factor alone is what separates poor people from people who are not poor.
That is why members of the Royal Family are not poor even though they suffer many common human problems like those mentioned in the document as identifying poor children. That is why Mrs Rausing, the multi-millionaire whose addiction killed her, was not poor.
The challenge for any government who wanted a better measure of child poverty is to find a better way of identifying who are the families with inadequate incomes than the present method of counting families whose incomes fall seriously below the median household income, whether or not the incomes are adequate or inadequate. If the consultation were to focus on that issue instead of misrepresenting other factors, it would be a remarkable achievement, since no UK government has ever faced up to that challenge.
What really is meant by ‘child poverty’?
Poverty cannot be defined in any way one likes. Its technical definition in research should not stray too far from the everyday meaning of the word.
Children normally live in families with both or either of their natural or adoptive parents. To speak about ‘child poverty’ is therefore to address family poverty, since conservatives and liberals agree it is no business of government to interfere in internal family matters unless children are damaged by abuse or neglect. If children live in poverty, especially with consequences which damage them, government is responsible for removing the cause of the poverty to protect the children. That is an elementary political principle in all non-totalitarian states, conservative as well as social-democrat, except those where government has abandoned any pretence of protecting all citizens against dangers (as is the case in some individualistic liberal states).
It is a fundamental principle of our free society, one to which the Secretary of State has often alluded, that people should be free to make their own choices and to act responsibly, a principle which depends crucially on their having power over the necessary resources with which to do so. To lack the resources demonstrably needed with which to make such choices is to vitiate the implementation of the principle itself. For example, everyone supports the idea of freedom of choice in the market, but if one lacks enough cash with which to enter the market and exercise one’s choices, one is not free. Nor is one free if the market offers no choices at the price one can afford on the income one has. Children whose parents do not have the freedom of choice which adequate resources gives them to provide what children need, are not free. Governments which believe market methods are best are therefore logically responsible for ensuring that all families have sufficient resources to enter the market.
Similarly, parents without paid employment cannot exercise their responsibility to escape poverty through paid work if they lack willing employers to offer it to them. Employers will not take on staff if they cannot make a profit from their labour or if they are not themselves paid to do so. Parental willpower does not bring willing employers into being. As William Beveridge (a lifelong expert on unemployment questions) wrote in his celebrated Report,
… the only satisfactory test of unemployment is an offer of work.
Thus to emphasise and exemplify the point about personal characteristics which exercises Mr Duncan Smith, the responsibility to ensure that all parents seeking paid work can find it lies on their potential employers and behind them the government’s control over the economic system. No amount of enquiry into the personal characteristics of the parents in ‘workless’ households (Mr Duncan Smith’s first example of dimensions of child poverty) will help the government to understand why children are poor or reduce their poverty.
This example also shows that the crucial element is the payment for the work, not the fact of work itself. If the wage or salary is too low, the family remains in poverty, as shown by the fact that most children currently counted in poverty in the UK live in working households.
If the government wants to use parental worklessness as a criterion of poverty, it must focus on the availability of paid work and the adequacy of incomes in or out of work, not on the willpower of those seeking adequately paid work which is not currently available to the parents however great their willpower. This point is such an elementary one, but one which is frequently denied by political rhetoric like Mr Duncan Smith’s. We must therefore disentangle the many issues similarly raised by the consultation document.
Mr Duncan Smith’s tenuous grasp of the difference between rhetorical assertion and economic and social reality is demonstrated by his frequent repetition of the myth about ‘three generations of families and even communities where no one has ever worked’. He of course means employment, but the most strenuous efforts to find any such people or communities have been completely fruitless because the most intensive research effort has totally failed to find any of them. It is fair to assume that, like zombies, they almost certainly do not exist, since even two workless generations were ‘a very rare phenomenon’.
While the DWP may think political rhetoric is an acceptable basis for market research, it bears no relationship to real social science research such as is needed to address the urgent problems of child poverty.
What is the common understanding of poverty itself?
Sadly, the consultation document contains confusions between poverty’s causes, conditions, correlates and consequences as if they were all relevant to measuring the poverty of children. To document the consequences of poverty in a reliable manner may be commendable, but to suggest that they offer a better measure of the extent of child poverty in the UK suggests extremely muddled thinking about what poverty is, as understood not only in the UK but generally throughout the civilised world, and how best to measure it.
The government is of course free to adopt whatever concepts, definitions and measures of poverty it chooses. But its idiosyncratic approach undermines the public credibility of its claims. The more it diverges from what people understand as the essence of poverty, the less the public will believe the government understands the real problems or wants to do anything real about them.
This would matter if the government wants to be taken seriously about its claim to develop a more methodologically robust measure. Of course, if the aim of the marketing consultation is merely to sell the already determined policy instead of devising a better measure, then the credibility of the measure is not an issue for the government even if it is for everyone else.
The following paragraphs try to disentangle and demystify some of the confusions displayed.
Few people would disagree with the World Bank’s definition of poverty —
‘Poverty’ can be said to exist in a given society when one or more persons do not attain a level of material well-being deemed to constitute a reasonable minimum by the standards of that society.
This is however a static definition: poverty as the condition in which people do not attain the minimum socially approved level of living. A more dynamic definition raising the issue of causes is that given by the leading figure in 20th century poverty research, the late Professor Peter Townsend —
People are relatively deprived if they cannot obtain, at all or sufficiently, the conditions of life – that is, the diets, amenities, standards and services – which allow them to play the roles, participate in the relationships and follow the customary behaviour which is expected of them by virtue of their membership of society. If they lack or are denied resources to obtain access to these conditions of life and so fulfil membership of society they may be said to be in poverty.
Politicians and others often misunderstand or misrepresent the meaning of ‘relative poverty’ or ‘relative deprivation’. ‘Relative’ does not only mean one mere statistical position relative to another. Its ordinary everyday meaning is being deprived relative to the common standards of the society at the time at which the deprivation is suffered. The relativity is to minimum social standards, and not just to numbers as in the mistaken assumption.
The United Nations Committee on Economic, Social and Cultural Rights affirmed in 2001 that poverty is —
… a human condition characterized by the sustained or chronic deprivation of the resources, capabilities, choices, security and power necessary for the enjoyment of an adequate standard of living and other civil, cultural, economic, political and social rights.
A succinct formulation of these definitions of poverty is —
… an enforced lack of socially perceived necessities.
Most considered definitions of poverty focus on it as the lack of resources and not on the consequences such as poor conditions or coping behaviours, and some emphasise the dynamic of the enforced lack or deprivation, to make it clear that this is indeed a matter of lack of power and not of choice.
The consultation document takes it for granted that the consequences of children’s relative deprivation is the essence of the child poverty to be measured, but that confuses the exclusion from customary lifestyles and perceived necessities with the condition and its cause, the sustained, chronic and enforced lack of resources with which to buy inclusion.
This raises the questions what are the resources and why are they lacking. They may be an individual family’s flows of disposable cash income or stocks of assets such as wealth, housing and so on. Or they may be collective resources such as Mr Duncan Smith implies when he refers to poor housing, troubled areas or failing schools. But individual purchasing power cannot bring a supply of well-maintained and spacious housing into existence, cannot see that the environmental and socially needed facilities of an area are supplied, or improve the quality of schools. These are matters demanding collective action, and usually also demanding levels of collective resources which can only be mobilised by governments at local or national levels if they themselves have not only the statutory powers and resources to do so but the political willpower.
It is essential that politicians such as Mr Duncan Smith who believe that markets are the right and proper response in such situations must try to understand that questioning that view is not simply ideological, but it is logical and functional. Private (profit-seeking) markets are peculiarly bad at responding to individual need for the simple reason that they can only respond to effective demand (in the economic sense). But the demand which people in poverty put forward to relieve their needs is not ‘effective’ in the economic sense because it is not backed by enough money to enable the markets to be profitable. As a moralist, Mr Duncan Smith should be troubled by the failure of his preferred market panacea to cure the problems which trouble him.
The more the government focuses on such collective aspects of deprivation in its proposed measure, the more that the findings will reflect on the government’s failure to make provision to overcome the deprivations itself, and the more it emphasises that the experience of poverty is not a reflection of parents’ capacities alone. If the government does not wish to offer its critics a rod to beat it with, this would seem politically foolish, quite apart from being inappropriate to a poverty measure. (Of course, as I said before, if the aim of the consultation is merely to sell the already determined policy instead of devising a better measure, then the illogicality of the measure is not an issue for the government even if it is for everyone else.)
Whose minimum standards should be applied?
But the prior question which frames the issue is, what are the socially-approved minimum standards and to whom are they applied and by whom? Are they the personal standards of the politicians in power at some point, or are they those of the population in whose society this poverty exists? This is no mere rhetorical question but goes to the heart of what it means to live in a free society. When governments decide ‘what is best’ for the population we call them totalitarian, but free democratic societies let the people themselves discuss and decide what are the reasonable minimum standards for everyone in society.
Of course the process of arriving at such decisions may be a complicated one in large and complex societies such as that of the UK, but it remains a fundamental principle that in a free society it is the people and not the government which decides on such important matters as this one. Only totalitarian governments insist that the government standards shall take precedence over the standards which the population would apply to itself. Both of the parties in the UK’s current government assert their deep-seated commitments to freedom and democracy, allied to a rejection of totalitarian dictation and a distaste for interference in people’s private lives. These essential basic principles must therefore be reflected in the measures which the government uses for such an important question as the poverty which damages children and their futures in the UK.
To measure child poverty therefore also demands clarity about the minimum standards which this UK society (not any current government politicians) applies to define and identify poverty and the individual and collective resources which can overcome it. Given the importance which this government places on markets rather than public service methods of dealing with problems, and given that this society is a largely marketised and consumer-oriented one, Mr Laws the Minister of State is right to say in his Foreword that —
… the lack of a decent income is and always will be at the heart of what it means to be poor.
But the other factors which he and Mr Duncan Smith mention are either consequences of a lack of income or are widely experienced. A moment’s thought will show that every one of the bullet point factors which Mr Duncan Smith mentions in his Foreword can be managed, made tolerable or even overcome if one has enough money.
This is a crucially important point for the credible measurement of poverty. Conditions and characteristics such as worklessness, problem debt, poor housing or troubled areas, unstable family environments, failing schools, occupational and social skills, and even poor health, are not characteristics which distinguish poor people from the non-poor population. In this context it can’t be repeated too often that even members of the Royal Family have over time displayed characteristics such as worklessness and lack of skills (in the sense of employability), problem debt, unstable family relationships such as divorce and single parenthood, and poor health. But in each and every case, and also including housing and schooling, having enough money has enabled that particular family and most others like them right across society’s income spectrum to escape from or deal with the problems.
This absolutely key point is not just a matter of personal opinion but has been tested in robust and reputable international research. An extensive European study of the characteristics of people suffering poverty and social exclusions found —
… the defining characteristic of the poor is that they have a material standard of living that is socially regarded as unacceptable; the poor do not share any other characteristic or combination of characteristics that distinguishes them from the non-poor. The poor are not necessarily excluded in the sense of having low status or being restricted in their social contacts. They cannot be identified on the basis of behaviour, or any other observable characteristic only.
What does distinguish those people who are poor from those who are not poor is that poor people do not have enough ‘decent income’, as Mr Laws describes it, to buy their way out of the problems we may all experience. If we were to use the characteristics Mr Duncan Smith describes in order to count people in poverty, the obvious question would be why have you chosen the consequences when you could far more simply and justly have chosen their cause?
My response to the consultation’s numbered questions is therefore coloured by this core understanding — it is not the consequences of not having a decent income which is the key to identifying and measuring poverty, it is the lack of household disposable cash resources at a level which allows parents and children —
… to play the roles, participate in the relationships and follow the customary behaviour which is expected of them by virtue of their membership of society.
But if the government’s implicit political aim is to draw attention away from adequate incomes to mere behavioural consequences, it is surprising to choose to do so in ways which are likely to suggest that the numbers in poverty are even greater than they are by ‘decent income’ measures, and take in large sections of the non-poor population who may resent being labelled in a way which the government itself uses to stigmatise people.
What’s wrong with using the existing HBAI measure to count people in poverty?
The internationally-used HBAI income inequality currently used by the UK to measure and count households at risk of poverty has been criticised because it measures income inequality and not income adequacy. Income inequalities demand to be measured in their own right, but they are not a substitute for measures of income adequacy.
Household incomes above the median may be insufficient to achieve the socially-defined minimum decency level of living, as was reported to be the case by research in Eastern Europe after the collapse of the Soviet system (even majorities of the population may be in poverty by their own standards). But the only credible alternatives to the HBAI measure which make any logical or socially-recognisable sense are those based on robust and reliable social science evidence of UK society’s minimum standards and the incomes needed on average to achieve them. So whatever the reasons for the government wanting a better measure of child poverty, the place to start is with the household disposable income measure, not with consequences.
The next section describes the principles of any good, serviceable poverty measure, as set out by the US Government’s intensive study in 1995. Any good measure must obviously distinguish between those who do and those who do not experience the condition at issue, avoiding all confusions between causes and consequences like those discussed above. The point is not that the detailed US proposals should be adopted by the UK — they are inappropriate in many respects — but that the UK government should pay careful attention to the principles which have been thoroughly investigated and clearly enunciated. They deserve attention in the UK.
Principles of any good serviceable poverty measure.
The US Congress’s Joint Economic Committee commissioned the US National Research Council to carry out a study —
to address concepts, measurement methods, and information needs for a poverty measure, but not necessarily to specify a new poverty ‘line’.
It recommended the retention of “the basic notion of poverty as material deprivation’”. That was the US approach to poverty, worth noting in a country at least as preoccupied with the behaviour of people in poverty as is the UK government!
What’s also crucially important is its emphasis that it was not concerned with the different question of the level of means-tested income maintenance benefits such as social assistance. A lot of people muddle these two distinct aspects. Public opinion about the minimum standards of decent living which should apply to the whole population may not coincide or be comparable with views about the adequacy or inadequacy of the social security or social assistance systems. They must not be confused with each other.
The US study concluded —
Our recommended changes are based on the best scientific evidence available, our best judgement, and three additional criteria.
First, a poverty measure should be acceptable and understandable to the public.
Second, a poverty measure should be statistically defensible. In this regard, the concepts underlying the thresholds and the definition of resources should be consistent.
Third, a poverty measure should be feasible to implement with data that are available or that can fairly readily be obtained.
If the UK government really were concerned with counting children in poverty, it would not want to adopt any methods or criteria which are less rigorous than those recommended by the US government’s official expert panel — the best scientific evidence, minimum standards which are acceptable to and understood by the population as a whole, a statistically defensible, robust and reliable measuring tool based on consistent concepts and definitions, and one that is administratively feasible based on available data.
This is the first part of a response to the government consultation on ‘better’ measures for child poverty from Professor John Veit-Wilson, Newcastle University.