You might be looking at that title and thinking the Euro symbol doesn’t fit for an ‘S’ like the dollar symbol does. And that perfectly matches the topic of this blog post.

A recent survey polled the salaries of designers around the world. Designer and developer Ivana McConnell then took the raw data and built a very attractive visual representation of some of the statistics. The reputable Smashing Magazine consequently took a look at some of McConnell’s charts, focussed on the USA vs Europe chart and concluded that it "might be time to consider leaving Europe for a better perspective".

I beg to differ, and would argue that this chart and many of the others collated from the poll’s data are pretty much meaningless. I’ll explain why.

Living standards

Different living standards occur in each country around the world – even between cities in the same country (for example, compare London with Glasgow); different living standards incur different costs, which are then reflected in local salaries.

Which is the root of my problem with these results: this survey doesn't take into account the local costs of living, which always has a direct effect on salaries.

Consumer power

For example, one of the charts compares aggregated salaries from across Europe with those from the USA. Now, Switzerland and Scandinavia have high living costs – the highest in the world according to this website which compares their Consumer Price Indices (CPI). Or perhaps Purchasing Power Parity (PPP) is a better comparison, which shows that Switzerland and Norway are amongst the world’s 10 highest countries by PPP index, and that the 20 highest countries covers all Scandinavian countries. And yet cross the Alps from Switzerland into neighbouring Italy and both the consumer price index and Int$ (the measure of PPP) drop by a third. In theory, one should expect that salaries in Italy be proportionally lower than those in Switzerland. To lump these two countries together and compare them to a single country on the other side of the world is almost meaningless.

A similar drop is seen between Finland and neighbouring Russia.

So a comparison between aggregated countries against a single country is problematic, but this is also dependent upon the evenness of the underlying data.

How much is a vote worth?

I suggest that anyone considering the results of the poll also takes a look at the weighing of the votes: one can see that the number of votes from the high PPP-indexed countries is easily matched, if not overtaken, by countries which have a considerably lower PPP index.

For example, Norway, Sweden, Denmark and Switzerland have a total of 19 votes; Turkey, Russia, Italy, Spain and Portugal which all exhibit a sizeable difference in their PPP indices and therefore lower salaries should be anticipated, outweigh the richer countries with 23 votes.

And then there are Eastern European countries with small votes: Hungary, Czech Republic, Poland, Romania, Belarus, Ukraine, Latvia, Moldova, Bulgaria and Lithuania all lower the average European salary. Nothing wrong with them voting, it just skews the statistics if the raw salaries, without any normalising, are taken under consideration.

So comparing USA and European salaries is akin to comparing, say, Swiss salaries with those of all countries in the Americas. How would the USA (with around 53,000 Int$) like to have their salaries aggregated with those from Colombia (~11500 Int$), Guatemala (~6000 Int$), Bolivia (~5600 Int$) and Honduras (~4700 Int$)?

Conclusion

These figures are all but meaningless without some form of statistical normalisation.

As Einstein taught us, all things are relative. And without any relativity, statistics can be used to prove anything – 40% of people know that.