Census sensibilities III

On March 13th, when the 2006 census figures were released, Tom Marshall engaged in the following bit of self-congratulatory massaging of the digits:

The Honourable Tom Marshall, Minister of Finance and President of Treasury Board, said the census counts released today by Statistics Canada are not final population estimates, but merely a step in a larger process used to determine final official population estimates.

The minister said census counts are headcounts based on 2006 census forms actually received by Statistics Canada. "Some people are missed during the census process. Separate estimates are made to ensure that everyone who should be counted as part of our population to determine equalization entitlements, demands on health care and education services, are included in the official population estimates" said Minister Marshall. "Census counts provide raw data that must go through several levels of analysis and be used in conjunction with other information before a final population estimate can be reached."

[...]

"While these counts show that our population has declined 1.5 per cent, this is a far cry from the 7.1 per cent drop in the census covering the years 1996-2001. Government takes issues surrounding outmigration and challenges facing our rural communities very seriously."

[Emphasis added.]

Fair enough, on both counts (ha! pun!). The census figures are subject to post-censal revision. And the population decline between 2001 and 2006 is much less than that from 1996 to 2001.

But a closer look at the intercensal figures provides pretty compelling evidence that this "decline in the decline" is probably little more than a landing on a steep downward staircase.

First, consider StatsCan's quarterly interprovincial migration figures. This chart shows net migration (in-migration minus out-migration) for Newfoundland and Labrador from 1990 to 2006. At the far left, during the recession of the early 1990s, there were episodes of net in-migration. (Just as there were during the recession of the early 1980s; there was less reason for people to leave, and in some cases, reason to pack it all and go home. This will happen in the next North American recession, too.)

(Clickificate to enlargimafy.)

Things went downhill rapidly after the closure of the cod fishery in 1992, with net outmigration reaching its zenith (nadir) in mid-1997.

But the trend after 1997, for six years, was much more favourable. There was still net outmigration, to be sure, but the rate of that outmigration decreased to recent-historic lows.

On the other hand, since late 2003, apart from one quarterly net in-migration, the rate of out-migration has been picking up again. The most recent quarter (ending January 2006) for which stats are available show a level of outmigration that is on par with the darkest days of the post-moratorium period.

It would therefore appear that the more favourable census population decline in the latest census, of "only" 1.5% population loss, is due almost entirely to more favourable demographic trends in the first half of the latest intercensal period; that is, from 2001 to late 2003.

But in- and out-migration are only two components of population change. (Well, there are two others, in- and out-migration internationally, but (a) they don't change the big picture much, and (b) figures aren't available for the same time-period and time-scale.)

You also have to consider natural population change, births and deaths.

For most of the post-Confederation period, Newfoundland and Labrador has had high net out-migration, but, until 1996, no net population loss. This was always because the province had a high rate of natural increase, that is, more births than deaths, to compensate for the out-migration losses.

No more.

The rate of natural population change, as shown in this graph, has been sliding from a substantial increase to an insubstantial one, to natural population decline.

(Clickificate to enlargimafy.)

Birth rates are declining, not only of their own accord, but because out-migration is depleting the population of people in their child-bearing years. Death figures are climbing as the large populations of people born in the early and middle parts of the last century are succumbing to old age and typically 20th-century causes of death for the middle-aged and early retirement-aged demographic brackets.

The natural population change trend is clear enough. (And while NL is the first to reach natural population decline, all other provinces, with the temporary exception of Alberta, are showing early or advanced signs of the same statistical slope.)

And the recent outmigration trend, if it holds, is demographically scary enough. It is too early to say how clear it is, but given events in 2006, there is no great hope that the trendline since 2003 has reversed itself.

If anecdotes like this or like this are anything to go on, it certainly hasn't.

If current trends continue, not only will the 2006 census be just a brief landing on the staircase, the 2001 census may look like the good ol' days. And with billions of dollars in population-retaining or -building economic activity now tied up in a tightly-knotted ball of spite, there is nothing in the next four years, until the 2011 census, that appears likely to change the demographic course.

So when Danny Williams says things like, "If there doesn't happen to be a job for someone in St. John's in an engineering firm, that's unfortunate... but there has to be some price paid in the short term", or "St. John's can take the hit," the media, the public, the business community, the provincial civil service, and members of his own caucus have to start asking themselves (there's no point asking him) some serious questions.

Starting with "What is the big guy smoking, and where can you get some?".

The problem with predicting the future, they say, is that it hasn't happened yet.

But here's a confident prediction: History, other than that written by one historian in particular, is going to be viciously unkind to the reputation of the Hon. Danny Williams.