Here comes that sales tax meme
again. But whereas last year New York was losing $500 million a year in tax
revenues due to trafficking of counterfeit trademark goods, this year that number
seems to have doubled, to $1
billion a year. The New York Times reported the New York comptroller’s office
but as far as I’m concerned it still fails the smell test. So I downloaded the
report to see how they got at their numbers.
The first thing I noticed was that the report said that "New Yorkers paid
$23 billion for counterfeit goods during 2003", thereby depriving the city
of $380 million in unpaid sales tax revenue. These two numbers alone don’t seem
to add up: $380 million is a mere 1.65% of $23 billion, while New York City’s
sales taxes are over 8%.
So where is the comptroller getting his numbers? He starts off by estimating
global trade in counterfeit goods at $456 billion, or 6% of gross world trade.
For that number, he cites an estimate of 5-7% from the International Chamber
of Commerce. Then, he estimates the size of the US counterfeit trade at $287
billion, leaving us to assume that somehow he’s using the same estimate. He’s
not. In 2003,
US imports were $256 billion, while exports were $307 billion. Add those two
numbers together and you have US trade totalling $563 billion; 6% of that is
a mere $34 billion. (It’s worth noting that the comptroller is actually saying
that the size of the US market in counterfeit and bootleg goods is — rather improbably, one has to say —
greater than all of the USA’s imports put together.)
Alternatively, let’s say that the $456 billion figure is correct, and the US
accounts for its proportional share of that amount weighted not by foreign trade
(for the US is not a particularly open economy in that respect) but rather by
GDP. In 2003, US GDP was $10.9 trillion, while world GDP was $36.4 trillion.
On that basis, the US share of global trade in counterfeit goods would be $136
billion, still less than half of the number the comptroller actually uses.
So where does the $287 billion figure – the basis for all the comptroller’s
subsequent calculations – come from? All the report says is that "this
estimate is based upon an update of a 1996 estimate of US losses," footnoting
a 1996 article in that well-respected statistsical digest, Fortune
When you go to the article, it says only that "federal and industry surveys
indicate that America’s annual losses from the problem have quadrupled over
the past decade to a staggering $200 billion", without getting any more
specific than that. The comptroller’s "update" clearly consisted of
adding $87 billion to that figure to make it seem like less of a rond number.
But even assuming that the $200 billion number was accurate and not exaggerated,
there’s a huge difference between losses and sales. If I sell a bootleg copy
of Adobe’s Creative Suite for $20, Adobe’s loss is $1,200, while my sales are,
Nevertheless, the comptroller clearly says that New Yorkers are shelling out
$23 billion in cash for counterfeit and bootlegged goods, based not on sales
estimates but on loss estimates. He’s clearly comparing apples with oranges,
and he equally clearly knows that he’s doing it: the way that he buries the
numbers in his report shows that he’s rather embarrassed about their source.
Once he comes up with his spurious $287 billion number, of course, his deductions
are easy. New York City accounts for 4% of US GDP, counterfeiting is twice as
prevalent in New York as it is in the economy as a whole, therefore New York
City’s trade in counterfeit goods is 8% of $287 billion, or $23 billion. (In
fact, he even refers to it at one point as $22.9 billion, making the estimate
seem more accurate still.)
Nowhere does the comptroller explain why he’s using a multiplier of 2 on New
York’s share of trade in counterfeit goods, rather than 1.5, say, or even 3
or 4 or 10. There’s simply a bald statement: "The Comptroller’s Office
estimates that New York City’s share of the US counterfeit goods trade
is twice its share of gross product". Not even a footnote, this time. In
general, absolutely none of the comptroller’s estimates come from hard facts
on the ground: all of them take broader extrapolations from the country or even
the world as a whole, and then work out what New York’s share of that might
That, of course, is pretty stupid. Estimates of the global trade in counterfeit
goods are only as good as their underlying data, which necessarily must come
from component states and municipalities. If the best that New York can do is
to work down from the global estimates, rather than producing its own bottom-up
analyses, then the main lesson that we can learn from the whole exercise is
that none of these numbers really means anything at all.
If the comptroller had stopped to think, he might have realised that his numbers
seemed unrealistically large. His own report says that "the City’s
sales tax revenue per resident totaled $440 in 2000", which implies total
taxed spending per resident of about $5,300, at a sales tax rate of 8.25%. With
8 million residents, that in turn means about $42.7 billion in total legal spending.
If spending on counterfeit goods indeed is now $23 billion, then it would account
for more than half of legal spending, and more than a third of total spending.
No one, surely, spends more than a third of their total spent income on fake
goods. If the comptroller’s estimates are right, then for every dollar you spent
on hotels and restaurants, in bodegas and department stores, in wine shops and
music shops, on cars or bicycles or skateboards, you would, on average, spend
54 cents on counterfeit and bootlegged goods. Which is clearly ridiculous.
As ever, however, if an official agency throws lots of official numbers at
the press corps, the journalists will lap
them up and not ask any questions. The correct response to the New York
comptroller’s press release would have been to ask him why we should believe
anything he says, given the obvious garbage that he is spouting here. Instead,
the press simply repeated his numbers, without any indication that they might
be suspect. In the end, both the comptroller and the press end up looking stupid.