ECB Publishes A Lie Dressed Up in Statistics – and the New York Times Falls for It
by Dirk Ehnts, Econoblog101
Here is the New York Times this week, republishing a Reuters article:
Peripheral Euro Zone Household Wealth Trumps Core – ECB Study
FRANKFURT (Reuters) – Households in many peripheral euro zone countries are on average wealthier than those in the bloc’s core, a study published by the European Central Bank on Tuesday showed.
Actually, that’s wrong. It should probably have read:*
Peripheral Euro Zone Household Wealth Does Not Trump Core – ECB Study
FRANKFURT (Reuters) – Households in many core euro zone countries are on average wealthier than those in the bloc’s periphery, a study published by the European Central Bank on Tuesday showed.
The problem is that the ECB study (available here) looks at median values. Median is not average. Here is an example from a website which shows you how to calculate the median:
Find the number in the middle of the series.
This is the median. 2, 19, 44, 44, 44, 51,56, 78, 86, 99, 99.
Well, as you can see a lot of information is omitted. Basically, the most extreme information is omitted. So, if you look at household wealth in different European countries and you take out the poorest and the richest of every country – what happens to the data? In countries with high inequality in wealth, the super rich are omitted. In countries where wealth is more evenly distributed this effect does not play a role. An example: let’s imagine two countries with 10 households each and look at the wealth of the households:
country 1: 0,5,5,5,7,8,10,10,10,20 (sum=100, average 10)
country 2: 0,0,0,5,5,5,5,5,25,150 (sum= 200, average 20)
Which country has the wealthier households? If you would go by average, country 2′s households have 100% more wealth than those of country 1. Now, how about the median? Since we have an even number of observations, we have to find the two in the middle and average them to get the median:
country 1: 0,3,5,7,7,7,10,10,11,20 (sum=100, average 10)
country 2: 0,0,0,5,5,5,5,5,25,150 (sum= 200, average 20)
So, that will be 7 for country 1 and 5 for country 2. Oh, those poor households in country 2! Oh, and by the way: how many people are there in a households? According to the OECD, early or mid-2000s data, in Spain that would be about 2.7, in Germany 2.3. And, what about the size of pensions?
I don’t know what is worse. The abuse of power at the ECB, where the authors of the study clearly try to mislead the public (don’t tell me that nobody expected this ‘misreading’ of the results) or the journalistic standard at a respected newspaper (NY Times) and a respeced news agency (Reuters) when it comes to statistical studies.
* I base my claim on the fact that GDP of the periphery countries has not risen above German GDP after WWII. Accumulating more wealth per household than a country with a consistently higher GDP per capita would be an extraordinary feat, given that Germany has been a net exporter almost all of those years.
