Hong Kong's misleading data
SCMP April 19
It's no news that statistics lie. Some lie because they are innocently
based on wrong assumptions or poor data collection. Others are more
damnable lies contrived to deliver the results that those in charge
of them want.
I do not presume to know into which category fall some of the statistics
on inflation and deflation which are used to justify key policies in
places as different as Hongkong and the United States. But understatement
inflation or overstatement of deflation certainly come in handy for
Take a look first at Hongkong GDP data. According to the e Financial
Secretary in his Budget speech, GDP last year grew 3.3% in 2003 and
is forecast to rise by another 6% this year. That's an impressive showing
given that the population grew at less than 0.5% and the workforce not
However, take a closer look. GDP in current prices in 2003 actually
fell by 2%. The difference between that and the 3.3% "growth"
consisted of a fall of an astonishing 5.1% in the GDP deflator, which
is supposed to reflect price changes across the economy as a whole.
That is indeed bizarre.
In 2003, consumer prices as measured by the Composite CPI fell by just
3%. Even that figure is probably exaggerated (from a practical if not
economic statistics viewpoint) by a massive (24%) weighting of private
rental housing in the index. Most people live in their own homes or
in public rental housing. The private rental index is used as a proxy
for the whole private sector, which is not an unusual method but is
a poor reflection of the real situation in Hongkong.
The exaggeration (for practical purposes) of the GDP deflator primarily
derives from the huge weighting of imports and exports in the index
- even though most of this is low margin re-export business -- and,
perversely, from a fall in Hongkong's terms of trade. Export prices
fell by more than import ones. Prices of retained imports - the food,
fuel, consumer durables etc we use on a daily basis -- actually rose
by 2.9%. The statistical method has thus turned a real income fall into
an increase in the GDP deflator which has puffed up the headline GDP
growth. Look at the fine print of the GDP figures and it turns out that
real gross domestic income rose by 1.5% -- less than half the headline
Nor is this just a one-year aberration. The GDP deflator has now fallen
by an amazing 23% since 1998 whilst the CPI has fallen by only 15% in
the same period.
The government is even expecting these lucky GDP numbers to continue
in 2004. The forecast provides for a further 3% fall in the deflator
which will boost "growth" from a nominal 3% to a headline
6%. How the forecasters reach this conclusion in the face of CPI numbers
increasing in five of the past six months, rising prices in China and
an even weaker US dollar defeats me. But wishful thinking dies hard.
I am not suggesting this is deliberate manipulation by the Census and
Statistics. But the appropriateness of the methodology needs examination.
Exaggerating deflation comes in handy for the world's highest paid
government officials aiming to cut most public servants' salaries and
benefits other than their own. (And employ more PR people rather than
nurses to combat disease).
The official data claim that real wages rose 15% between September
1998 and September 2003. Much of this supposed rise in real incomes
of those in employment is a statistical mirage.
The main winners of recent years have been those who actually live
in private rental accommodation, or have large mortgages which have
benefited both from low interest rates and income tax concessions. Meanwhile
household incomes have been hurt not just by the level of unemployment
and a fall of 2% last year in nominal wages but in a continuing decline
in the workforce participation rate-down another 0.5% in 2003 to make
a cumulative fall of 3% since 1997.
It is clear that the man in the street has a much better grasp of the
real economy than fat cat peddlers of dubious data, Donald Tsang and
But Hongkong's financial leaders are in good company. Global bubble
blower Alan Greenspan continues to manipulate data to justify holding
real interest rates at zero or less to keep asset markets pumped up.
Note the way that Fed apologists emphasise the so-called "core"
CPI number when pointing to low inflation and a weak labour market to
justify interest rate policy. Convenient, is it not, to omit energy
and food prices as though they were irrelevant?
Meanwhile the "core" index itself has been manipulated by
use of so-called hedonic pricing. In simple terms this means that if
your new computer is 50% more efficient than its equivalent three years
ago, it is thus 50% cheaper. The fact that it is now impossible to buy
the slower version for 50% less is ignored. CPI statistics in the US
like those in Hongkong also suffer from the over-weighting of private
rental accommodation in an economy where most people own their own homes.
Rentals may have been static, but just look at the boom in capital values!
Artificially low inflation numbers have a double benefit. They justify
a lax monetary policy which helps keep the asset price bubble from imploding.
And they hold down government pension and welfare benefits which are
The bottom line is that politicians will always pick and choose data
for their own justification just as corporations cover up dreadful operating
results with extraordinary profits or blind the public and the ever
gullible financial wire services, with "pro forma" reports
which ignore accounting rules. The only thing that is surprising is
the willingness of highly-paid brokerage economists or job-secure academics
to look behind the headline data.
TOP OF THE PAGE