Lindsay Mitchell: CPAG research flawed

Recent research by the Child Poverty Action Group (CPAG) is flawed.

CPAG's analysis of Child Youth and Family child abuse data claimed, "The data suggests there is no correlation between benefit receipt and child maltreatment". This despite earlier Auckland University research finding, "Of all children having a finding of maltreatment by age 5, 83 percent are seen on a benefit before age two".

I asked the Ministry of Social Development for the data supplied to CPAG and was given the number of substantiated cases of child abuse and the 0-17 year-old population for each CYF site office. These show that CPAG's calculations are incorrect. For instance, their report states, '...the proportion of 0-17 year olds who were victims of abuse in Papakura was not 4.0% but 0.40 of 1%.' (p9)"

In fact the proportion was 4 percent (608 distinct cases in an estimated 0-17 population of 14,413). The flawed methodology was repeated for every CYF site office recorded.

In an attempt to ascertain correlation between child abuse and benefit dependence, the report went on to estimate the number of beneficiaries that lived in each CYF site area.

At this point CPAG counted all working age beneficiaries whereas the relevant population to use would be those beneficiaries with children in their care - a minority of all beneficiaries.

The rate of benefit dependency was also incorrectly calculated.  Data at Figure 5 (p11) is labelled "% income-tested beneficiaries estimated in total population". Data at Table 6 (p24) purports to be "rate of benefit receipt in working-age population". Yet the two sets are data are identical.

Using the example of Papakura again, CPAG's estimate for the rate of benefit dependency was 10%. In fact the number of working-age beneficiaries was 6,096; the 18-64 population was approximately 31,302. That results in a benefit dependency rate of 19.5 percent.

I have written to CPAG about these errors. They conceded that their report needs amending and said, 'An amended version of the report will be available on our website as soon as practicable.' Over two weeks later the faulty version is still on-line.

CPAG research is publicly promoted to influence social policy. It's therefore hugely important that policy-makers can trust their work. If CPAG is publishing faulty research that trust would be misplaced.