Not Making Census

Hey there, time traveller!
This article was published 15/9/2013 (1411 days ago), so information in it may no longer be current.

Most data nerds look forward to census day, and I’m no different. Last Wednesday, I was pretty geeked up about the last release from the 2011 national household survey, the juicy stuff on income. I knew the data would probably be dodgy, but I held out hope.

I spent most of the morning trying to find a story in all the numbers, clicking through page after page of StatsCan data, downloading tables, getting more and more frustrated as my boss kept asking me questions I couldn’t answer. What’s the richest town in Manitoba? In the North End getting poorer? Where do the one-percent live?

Those are questions I could have answered in the last go-round in 2006. As you drill down to the tract level – the only level we care about as a local paper – the data becomes so spotty as to be nearly useless.

Here’s an example – a nifty chart showing what proportion of each community lives below the poverty line. Problem is, more than half the Manitoba census subdivisions have no data because not enough people responded. Most of the communities with no data are the ones whose poverty rates we’d most like to know about - First Nations reserves. And, is it really possible that 43 per cent of Ethelbert-ians live in poverty? That sounds nuts.

I spoke with Dr. Pat Martens, who runs the Manitoba Centre for Health Policy and whose staff use sophisticated data crunching to tie illness with income, right down to the neighbourhood level. She called the new NHS "heartbreaking," especially because it’s often poor people who fail to answer voluntary surveys and so get overlooked in what faint numbers exist. She mourns the loss of what was the gold standard of census data, one researchers and health policy experts relied on heavily.

In most cases, we cannot measure the most important thing: Change. In fact, Stats Can posted a big disclaimer warning journalists against comparing the NHS data to the 2006 census. That means we can’t say whether there’s really gentrification happening in West Broadway or how much richer Steinbach has become or where millions in anti-poverty programs are really working.

Another example: We don’t exactly know how many people live downtown, and it took me a day of backing-and-forthing with Stats Can staff to cobble together a reasonable figure. One batch of numbers, the straight population figures from the mandatory census, suggest there are 700 more people living downtown. That sounds very low, and it made downtown boosters like the BIZ’s Stefano Grande, cringe. But, it’s the only independent count that exists.

The funny thing is, another batch of numbers from the NHS, numbers I didn’t use, painted an even more dire picture of downtown revitalization. Those are the "population in private household" figures, which suggests there are even fewer people living downtown, particularly in the West Exchange, despite dozens of new condo units. And, we have no idea how many people live in private households south of Portage. That data doesn’t exist this year. It’s suppressed due to low response rates.

Without good data, we’re just living in the dark. We won’t know what neighbourhoods really need housing funds. We won’t know if downtown condo subsidies are working. We won’t know whether Vita needs a hospital. We won’t know where new immigrants are settling or what schools might need ESL programs. We won’t know whether government, with its vast, multi-billion-dollar reach, is actually working.