Last week the government announced its child poverty strategy – but at the same time revealed that, after a year of consultation and consideration, it has still not been able to reach a conclusion on how to measure success. Prof Ruth Lupton explains why the government should stick with the measures it has got.

Thanks to the Child Poverty Act of 2010, eradicating child poverty is a legislative commitment in the UK and one to which all the main political parties are committed. But what is child poverty, and how should we measure success?

Traditionally, the measure of child poverty has been the proportion of children living in households with less than 60 per cent of the median income. This measure is not perfect.  For one thing, it is a narrow, one-dimensional indicator which captures only some of the experience of growing up poor.

It is also less meaningful in times of economic shocks. When the median income falls suddenly, as in the recent recession, fewer people may be deemed ‘poor’, even though their own circumstances may not have changed.

Relative measures make sense conceptually (what influences a child’s life chances is what they have relative to others in society), and they work well in times of stability. But they can make less sense in times of economic turbulence. That’s exactly why the current set of measures also includes changes in poverty measured against a fixed, real standard – a measure which has been particularly important in the years since the crisis started.

The Coalition government’s response to this was to propose a new multi-dimensional measure, incorporating other elements of child impoverishment such as living in bad housing, being in a family suffering from unmanageable debt, and not having access to a good school.

It has reiterated last week that the existing measure is ‘discredited’ and needs to be replaced. Yet a year after its consultation on the proposal, no agreement has yet been reached on an alternative.

It isn’t for me to speculate on the internal wrangling that might be blocking agreement– what Alan Milburn has described as “beyond Whitehall farce”.    The substantive problem is of greater interest. In fact the existing measure is not discredited. On the contrary, experts in the field almost unanimously responded to the government’s consultation by rejecting the new proposals and advocating sticking with the existing measures.

They did this for three main reasons. One is that, although a wider concept of child poverty has wide agreement, a reliable multi-dimensional measure is hard to operationalise. The difficulties of collecting accurate data and assigning meaningful weights to issues as diverse as household overcrowding, bad diet and neighbourhood quality would make producing a reliable single number, capable of capturing change over time, very difficult.

This doesn’t mean indicators across multiple dimensions shouldn’t be tracked, but a dashboard approach would be more meaningful that a multi-dimensional index. Second, there is the problem of combining apples with pears. Some of the things the government wants to look at are risk factors for income poverty, some are measures of income, and others are measures of the consequences of low income.

For these reasons, we are better off with a robust if limited measure than one that is meaningless though well-intentioned in its breadth.

Third, income – the basis of the current measure – is the most important aspect of child poverty to measure. Since the consultation last year, a major independent review of the relationships between income and children’s outcomes has concluded unequivocally that money matters for children’s outcomes, independent of other factors.

In fact, raising the household incomes of those currently eligible for free school meals to the average level would go as far as to halve the attainment gap between FSM children and others. Other factors such as parental addiction, neglect, depression and so on may increase the risk of income poverty, and have effects on their own, but the most fundamental problem is that children growing up in households with low relative incomes will find it harder to thrive in absolute terms and relative to their peers. This is the essence of child poverty.

Changes in the incomes of households with children must remain at the heart of any measure and not be rendered less visible by a mashing together of different concerns into a single number.

Perhaps most importantly, we must also be very careful not to use new measures to shift the blame for poverty onto the poor themselves. Last week’s government statement identified the root causes of poverty as “entrenched worklessness, family breakdown, problem debt, drug and alcohol dependency”.

In other words, people’s personal failings – to hold a job, manage their money and relationships and keep off drink and drugs – are what we should be tackling, and measuring.

This account gives insufficient weight to what should be more pressing problems for government: jobs which don’t pay enough for people to feed their children or heat their home; the cycle of low and no pay that leaves families grinding out a depressing breadline existence, putting relationships under strain; the increasing living standards of the rich and the availability of credit at exploitative interest rates.

Government should not be looking for measures of how the poor manage themselves but of how poverty is being prevented by strong and fair labour markets and a strong and fair welfare state that protects the incomes of families so that their children have the best chances in life.

Arguments about child poverty measures are not just about the technicalities of measurement. They are also about what child poverty is and whose responsibility it is.

I, for one, am pleased that the government has not yet adopted the measures it proposed last year, and hope that it will continue with a robust, income-based measure, despite its imperfections.

• This post originally appeared on LSE’s British Politics and Policy Blog.