A few months ago the McNeil government’s reputation took a hit with news that while child poverty was dropping across Canada it was on the rise in Nova Scotia.The source of the disclosure was Stats Canada’s Canadian Income Survey (CIS) for 2017. As reported here, the release was given wide publicity by the federal Liberals as proof that their anti-poverty measures are working.

But sadly, although the Canadian rate of child poverty purportedly dropped from 11 to nine per cent between 2016 and 2017, it jumped from 14 per cent to 17.1 per cent in Nova Scotia. The February survey also showed that Nova Scotian families had the lowest median income in the country, slipping from third lowest in 2013.

As I wrote at the time, the income and child poverty picture provided through the CIS would be followed later by more solid data based on tax filings. Because of sample size, Statistics Canada rated the provincial child poverty numbers from the CIS as either “acceptable” or “use with caution,” instead of the more reliable “good” “very good” or “excellent.” As advertised, the second, more reliable set of income data, the T1 Family File (T1FF), compiled from tax returns for 2017, came out in JulyWhile still pretty dismal, the stats for Nova Scotia from the T1FF are less bad than the findings from the CIS. On the negative side:

  • Nova Scotia still has the lowest median family income of any province – $6,080 or 11.6 per cent below the national average;
  • The number of low income Nova Scotians increased by almost 4,000 from 2015 to 2017.

On a positive note:

  • Unlike the CIS, tax data show the rate of child poverty dropped slightly in Nova Scotia in 2017;
  • While the CIS showed Nova Scotia with the highest rate of child poverty, T1FF data put Nova Scotia only third worst in 2017, better than Manitoba and Saskatchewan.

The table below compares percentage of children in low income by province according to the Canadian Income Survey with those derived from taxation data and reported in the Census Family Low Income Measure (CFLIM-AT).[1]

 CIS-MBM 2017CFLIM-AT 2017CFLIM-AT 2016CFLIM-AT 2017 change from previous year
Canada9.0 %18.7%19.6%-0.9%
Nova Scotia17.1%24.4%24.8%-0.4%
Newfoundland-Labrador 10.6%20.4%20.0%+0.4%
Prince Edward I. 9.1%18.9%20.3%-1.4%
New Brunswick13.1%22.0%22.8%-0.8%
Quebec 8.4%15.3%16.3%-1.0%
Ontario 9.6%18.9%19.5%-0.6%
Manitoba 9.5%28.1%29.0%-0.9%
Saskatchewan10.3%26.4%26.7%-0.3%
Alberta 5.0%16.7%18.1%-1.4%
British Columbia 9.8%19.3%20.3%-1.0%

As a comparison of the data in the first two columns shows, the percentage of those 17 and under in low income is significantly higher in all provinces with the CFLIM-AT than with the CIS. There are two reasons for this. First, the percentages in the first column are derived using the Market Basket Measure (MBM) – which takes into account regional differences in median incomes and living costs, a factor which tends to bring down rates in poorer provinces. The effect on poverty rates of using the MBM – which the Trudeau government wants to establish as Canada’s official poverty line – was discussed here.

The second factor is that CIS measures the total income of everyone in a household while the CFLIM addresses only Census families comprised, according to Stats Canada of “couples (married or common-law, including same-sex couples) living in the same dwelling with or without children, and single parents (male or female) living with one or more children.” So household incomes are higher than census family incomes.

Admittedly, it’s all a bit arcane, but ignoring the higher rates and focussing only on the relative positioning, the latest income data takes some of the political heat off the provincial government, at least as far as child poverty’s concerned. Column four of the Table shows, using the CFLIM-AT metric, that the percentage of children in low income dropped by 0.9 per cent nationally and in every province except Newfoundland and Labrador. Nova Scotia had the second smallest reduction from 2016 to 2017, 0.4 per cent. But that’s better for the government’s reputation than the increase shown with the CIS data earlier this year.

Federal claims challenged

The latest income data from the T1FF has the opposite political impact on the federal Liberals. They would no doubt like to campaign on the anti-poverty achievements they were crowing about earlier this year with the release of the Canadian Income Survey. CIS data and the Market Basket Measure showed that as of 2017 the Liberals – thanks in part to the Canada Child Benefit – had already reached their five-year target of a 20 per cent reduction from 2015 poverty levels. The T1FF-based Census Family Low Income Measure shows nothing of the sort.

According to that measure there were 5,959,300 Canadians in low income in 2015 and 5,922,090 in 2017 – a reduction of less than one per cent. For children 17 and under the picture was considerably better – a drop of almost nine per cent, from 1,502,710 to 1,368,700. But that improvement was partially offset by an increase in the rate among seniors, the continuation of a trend reflected in CFLIM-AT data going back to 2000, as the Table below reveals.

Age group2000200520102014201520162017

All ages
17.7%17.7%17.7%17.7%17.3%16.8%16.8%
0-1724.2%23.6%22.2%21.8%20.9%19.6%18.7%
18-6417.2%16.9%17.3%17.4%17.2%16.9%16.9%
65 and over 7.8%11.2%12.6%13.4%13.2%13.0%14.1%
Source: Table 11-10-0018-01

There is an apples-and-oranges aspect here – the Liberal government’s claim was based on using the Market Basket Measure, while the number of low income tax filers is calculated using the after tax low income measure (LIM-AT). Presumably the Liberals will defend their record by promoting the superiority of the MBM and household income as the best way to measure poverty. But what is striking about the T1FF data in the table is the continuity in the overall poverty rate from 2000 to 2014. It stays pretty much the same, with modest improvements in child poverty being offset by sharply increased rates among over-65s – incidentally the age group most likely to vote.

Since 2014, coincident with increases in federal child benefits, the rate among children has come down modestly, pushing the overall rate below 17 per cent for the first time. On the campaign trail the Liberals could legitimately claim some credit for that. But any rhetoric about achieving a 20 per cent reduction in poverty three years ahead of schedule must be taken with a huge grain of salt.

[1] Individuals are defined as having low income if their adjusted after-tax income falls below 50% of the total population’s median adjusted after-tax income.