How statistics can mislead

RTE of Ireland reported in their online edition yesterday that an annual study of the cost of living* in the world's cities conducted by Mercer Human Resources, whoever they are, found that Dublin had moved up one notch from last year. They indexed city prices against those found in New York, assigning that city a value of 100.

From the RTE article:
In a study covering 144 cities on six continents, it measures the cost of more than 200 items in the housing, transport, food and entertainment sectors.
...
1 TOKYO Japan 134.7
2 OSAKA Japan 121.8
3 LONDON United Kingdom 120.3
4 MOSCOW Russia 119
5 SEOUL South Korea 115.4
6 GENEVA Switzerland 113.5
7 ZURICH Switzerland 112.18= COPENHAGEN Denmark 109.5
8= HONG KONG Hong Kong 109.5

10 OSLO Norway 105.3
11 MILAN Italy 104.9
12 PARIS France 102.213= NEW YORK CITY, NY United States 100
13= DUBLIN Ireland 100

15 ST PETERSBURG Russia 99.5
16 VIENNA Austria 97.8
17 ROME Italy 97.3
18 STOCKHOLM Sweden 96.8
19 BEIJING China 95.6
20 SYDNEY Australia 95.2

This listing would be next to useless to anyone wanting to compare the actual cost of living in one city with that in another. The largest item in most people's budget is the cost of housing, but it appears that Mercer didn't assign weighting values to the prices of the items they measured. If they had, Paris wouldn't have been listed above New York, it'd have been well below it. Apartment rents in Dublin are high, but from the listing someone might conclude that a New York apartment equivalent to the $1400 a month apartment that can be rented in Dublin would rent for roughly the same amount!
Londoners would best know housing costs there, but I'd find it hard to believe the true cost of living in London is anything like 20% higher than in New York.
Who cares about the price of a dress or an appliance when such a large chunk of your paycheck each month goes to pay the rent or the mortgage payment? That's my point.
Only a goofball, of course, would suggest that Oslo is 5% more expensive to live in than New York, or that Moscow is a whopping 19% more. The article didn't actually say that, but it implied it, and, I suspect, that is how most readers took it. Statistics can be misleading.
It'd be interesting to see a list of the top 20 cities based on the actual cost of living, not simply taking the prices of items in a number of categories into account. I couldn't possibly do it for the world's cities, but I could make a start with some of the major and not-so-major ones in the US, plus a few cities in Europe and one or two in Japan. Only one in Canada: rents and many things in Old Town Quebec are dirt cheap, relative to New York. (Nice town, too.)

*In fairness, the phrase 'cost of living' doesn't appear in the article, but with the opening sentence being 'Dublin has been ranked as the 13th most expensive city in the world' and with what followed, few people would think they were talking about something else. Evan, Donna, and I would know, but not everyone is Evan, Donna, or me.

Charles Riggs
There are no accented letters in my email address