November
12, 2002
Interpreting
Statistics
The
recent revision to provincial GDP data illustrates the kind
of problems one can encounter in interpreting statistics.
For some data series Statistics Canada puts out preliminary,
revised and final numbers although nothing is ever really
final. Years after the fact entire sets of data can be revised
using new methodologies.
On November
9th I wrote that "As a result of the revision by Statistics
Canada, BC's 2001 GDP growth after inflation went from a
positive 0.9% to slightly negative, but that is not due
to any downward revision for 2001 GDP. It is because the
estimate for 2000 was revised upward by $2 billion."
That conclusion was reached by dividing the revised current
dollar GDP data by the implicit price deflator for the previous
series. The result produced a 1997 constant dollar value
for GDP in 2000 that was $2 billion higher. Statistics Canada,
however, also revised the price deflator. I am grateful
to Bruce Little, business reporter at the Globe and Mail,
for pointing out to me that the real value for 2000 (constant
1997 dollars) "was not revised up by $2-billion from
the figure released last April. It was revised down from
$124.464-billion to $124.187-billion." How could that
be? The answer lies in changing the price deflator. Either
way, the economy peaked in 2000. With the new "Chain
Fisher Index" the 1997$ GDP for 2000 was revised to
$124.187-billion while the 2001 figure was revised to $123.912
billion." Consequently, real growth for 2001 changed
from a positive 0.9% to slightly negative as my article
reported using the old index. Using either price deflator,
the BC economy peaked in 2000.
It costs
$400 to purchase the complete set of revised provincial
economic accounts from Statistics Canada http://www.statcan.ca/Daily/English/021107/d021107b.htm.
Your free web service won't be analyzing all those data
unless more people respond to the appeal for support - see
http://www.StrategicThoughts.com/support.html.
Limited
free data are provided by Statistics Canada. The revised
GDP data in current dollars are available at http://www.statcan.ca/english/Pgdb/econ15.htm.
The constant 1997$ series is available at http://www.statcan.ca/english/Pgdb/econ50.htm
BUT while the free current dollar series reports the
revised data, as of November 11th, the free constant dollar
data are still unrevised (unrevised GDP and unrevised price
deflator).
Students
of business cycles know that the debate over their existence
and length depends on what statistical techniques are used
to remove the long term growth trend from the data. The
lesson from such exercises is, like the example of price
deflators, that the data are frequently not sufficiently
robust to support sophisticated analysis.
A regular
reader of these columns commented on my November 9th article's
reference to the index of help wanted ads. Statistics Canada's
latest
release noted "The index for Manitoba is not available
for October because of the Winnipeg newspaper strike."
The reader suggested that the BC data might also have a
higher error because of the dispute at the Times Colonist.
The Victoria metropolitan area is certainly important, but
it is not as large a portion of the population of BC as
the Winnipeg area is to Manitoba. The index weights each
newspaper's contribution by the 1996 metropolitan area population
and imputes the value if the paper is not available. Consequently,
some error might be introduced with the absence of the Times
Colonist but not enough to justify omitting the index for
BC. The help wanted index for BC has declined since July
(3 consecutive months).
November
9, 2002
BC
Economy Peaked in 2000
Statistics
Canada recently revised
its estimates of gross domestic product (GDP) by province.
The revision produced a story in the Globe & Mail with
the headline "B.C. economy shrinks for first time since
'82". It is true that as a result of the revision by
Statistics Canada, BC's 2001 GDP growth after inflation
went from a positive 0.9% to slightly negative but that
is not due to any downward revision for 2001 GDP. It is
because the estimate for 2000 was revised upward by $2 billion.
In the last full year under the NDP, B.C.' economy grew
by 7.2% (or 5.3% after inflation). That also means that
real per capita GDP increased by 4.5% in 2000 rather than
the previous estimate of 3.1%. Real per capita GDP is the
key indicator in the Campbell government's strategic plan.
It declined in 2001 and will probably show a slight decrease
again in 2002. That must be frustrating for Premier Campbell
but it is far more frustrating for families throughout British
Columbia.
In addition
to its revision of GDP estimates, Statistics Canada has
released other estimates that must cause concern as Finance
Minister Gary Collins prepares to release his Second Quarter
Report later this month. On a seasonally unadjusted basis,
employment in BC dropped by 14,600 in September and it fell
by another 18,100 in October. It doesn't ease the pain
for those families to know that when seasonally adjusted
those numbers are changed to an increase of 1,300 for September
and a drop of 2,900 for October. The help
wanted index for BC, a measure of jobs advertised in
newspapers, dropped to a level not seen since the recession
of the early 1980s.
The
Campbell government has been fond of pointing to robust
growth in retail trade but that too has hit the skids in
recent months. In August growth in retail
trade was only 4.8% compared to a growth rate of 6.4%
for all Canada. BC suffered a slight decline in retail trade
between July and August adding to the decline that July
showed relative to June. The estimate for September will
be out on November 22nd - about the same time Collins releases
his revised forecast for the second half of fiscal year
2002-03.
Construction
has been a bright spot in the BC economy. On November 6th
Statistics Canada released estimates for the value
of building permits. For the period January through
September this year compared to last, the value of residential
building permits increased by 38.3% in BC compared to 35.2%
for all Canada. Non-residential construction was not so
good. "Growth" in BC showed a decrease of 26%
compared to a decrease for all Canada of 6.5%.
It is
fair to point to the softwood dispute, the impact on tourism
from 9-11 and the 2001 recession when explaining why BC's
economy is sputtering. Of course, when Gordon Campbell was
in opposition, all economic ills were somehow blamed on
the provincial government. Premier Campbell asked that a
standard of relative performance be used to judge his government.
His strategic plan says "Our goal is for British Columbia
to have the best performing economy and the highest quality
of life in North America." If the economic forecast
by the Bank of Montreal is any indication, Premier Campbell
has a long way to go. That forecast placed BC dead last
in its estimates of economic growth for 2003 and 7th amongst
the provinces for 2004. Seniors who are being kicked off
Pharmacare, booted out of residential care and discharged
from hospital with no place to go probably question whether
their quality of life has improved in the New Era. We can
all ask "when will those tax cuts pay for themselves?"