MSD's annual tome of poverty statistics was released last week.
The NZ Initiative provides a good summary here. Eric Crampton wryly notes:
"Spin machines revved up quickly, trying to find the least indefensible ways of using the data to support whatever position they had before they’d opened the report."
The data features children aged 0-17 in primarily the Auckland region. Foodbank use is at least once in the last 12 months. AHC 50 relates to the percentage of children in households below fifty percent of the median income After Housing Costs and MH 9+ relates to the percentage of children experiencing 9 or more items of material hardship eg going without fresh fruit or doctor visits.
Click to view
Unsurprisingly in every area the relationship is similar. But it is certainly not uniform. The shapes of each area cluster differ.
The ratio of AHC50 to foodbank-use shortens moving left to right. In Albert etc. 14 percent are at AHC 50 and 3 percent used foodbanks. In Southern wards the numbers are respectively 25 and 19 percent. The gap is much narrower.
Without knowing how many times beyond once a foodbank was used, the dependency appears disproportionately greater moving left to right.
This might reflect a greater density of foodbanks in poorer areas.
But I also notice the marginal relationship between AHC50 (financial hardship) and MH9+ (material hardship) is not constant. It too tends to close but Western and Howick etc. have the biggest margins. This says to me that low incomes in those areas go further. Perhaps caregivers budget better. Perhaps they carry less debt. There could be a myriad of reasons.
Theories about how to measure poverty have been multiple and varied through the ages. The study of poverty has historic roots. But some have argued it might be better identified and understood by measuring not what goes into a house but what goes out. Expenditure counts as much as income.
This particular report arises from income and says, "The use of expenditure is not generally accepted as an alternative, in part because it also is not a good proxy for reporting consumption possibilities, but also because the quality of the expenditure data is often patchy."
Fair enough but not enough is known about expenditure (bar housing) to properly understand what might make a positive difference. And that is the point of these reports.
(The report does not, by the way, capture recent data or people living in emergency housing.)
Lindsay Mitchell is a welfare commentator who blogs HERE.
Click to view
Unsurprisingly in every area the relationship is similar. But it is certainly not uniform. The shapes of each area cluster differ.
The ratio of AHC50 to foodbank-use shortens moving left to right. In Albert etc. 14 percent are at AHC 50 and 3 percent used foodbanks. In Southern wards the numbers are respectively 25 and 19 percent. The gap is much narrower.
Without knowing how many times beyond once a foodbank was used, the dependency appears disproportionately greater moving left to right.
This might reflect a greater density of foodbanks in poorer areas.
But I also notice the marginal relationship between AHC50 (financial hardship) and MH9+ (material hardship) is not constant. It too tends to close but Western and Howick etc. have the biggest margins. This says to me that low incomes in those areas go further. Perhaps caregivers budget better. Perhaps they carry less debt. There could be a myriad of reasons.
Theories about how to measure poverty have been multiple and varied through the ages. The study of poverty has historic roots. But some have argued it might be better identified and understood by measuring not what goes into a house but what goes out. Expenditure counts as much as income.
This particular report arises from income and says, "The use of expenditure is not generally accepted as an alternative, in part because it also is not a good proxy for reporting consumption possibilities, but also because the quality of the expenditure data is often patchy."
Fair enough but not enough is known about expenditure (bar housing) to properly understand what might make a positive difference. And that is the point of these reports.
(The report does not, by the way, capture recent data or people living in emergency housing.)
Lindsay Mitchell is a welfare commentator who blogs HERE.
1 comment:
The definition of poverty is so debatable, and foodbank availability, and the measure of foodbank use so inconsistent that any correlation between the two is fraught. Some define poverty as not having 2 hot meals a day. I grew up in a typical middle family (as were near all) in the 1950s. One hot meal a day, maybe two on weekends. Sometimes hot milk on the Weetbix for breakfast if that counts. My father was allergic to beef so it was unrelenting mutton. Except for the weekend roast, I skipped. Was I poor? Walked a mile to and from school including home for (cold) lunch. No cigarettes, alcohol, or cordial except latter at Christmas. Takeawys (fish and chips) maybe 10 tines a year. All the same as near all schoolmates. We owned an old small car, with operation very seriously inferior to then current models. As foodbanks proliferate and the goods include more and more attractive items, and as knowledge of availability spreads, so does demand accordingly. Should persons above the average BMI qualify for classification as "poor"? Unlikely in most of the rest of the world. Has anyone done a comprehensive study of the detailed spending of the "poor"? Many eventually "wealthy" persons are and have been more frugal than the so called poor. A friend used to do home meat deliveries. They put a lot of mince in the fridges of the "rich" and steaks in the state house fridges. My main sympathy today is with the mothers obliged to work full time. It seems to have been a very retrograde step in the advancement of society.
Post a Comment