I have been off meeting company friends in Poland over the past few days, which is why this week’s output is shorter than usual. In my meetings one of the things discussed was the issue of cross-country comparisons (yes, they are every bit as geeky as yours truly). And my favorite illustration of how difficult these things can be is with the frequently used denominator disposable incomes. For my home country Sweden, for instance, people like to take household indebtedness, or house prices, in relation to disposable income and become flabbergasted by the extremely high debt ratios. Especially when seen in relation to other economies.
Note: These numbers might deviate slightly from official measures as it is incl NPISH and harmonized.
Yes, it looks ridiculously high and Sweden is certain to suffer the mother of all crises, right? – Well, yes, perhaps. And no matter how you turn the data, Swedish debt ratios are historically high. However, and given our objective to discuss the difficulty of making cross-country comparisons, what might be wrong with the above picture?
– The numerator is of course also candidate, to be sure. But the oft-used denominator of household (gross) disposable incomes indeed suffer from ‘irreconcilable differences’ producing a less than optimal match for cross-country comparisons. The main reason is the extent of public financing of different welfare services. In the US, for example, schools and hospitals are, to a large extent, privately financed. In Europe, particularly in Sweden, welfare services are publicly financed. Therefore, many (most?) statistical agencies publish a measure called adjusted (net) disposable incomes as well (OECD has some harmonized version of adjusted disposable incomes as well). While far from perfect, and probably still implying some form of underestimation of incomes in high-tax countries, by adding household consumption of typical welfare services to the “normal” disposable income, this measure facilitates comparisons between economies (and possible fiscal regime changes).
Looking at the same measure as above, but calculated with “adjusted” disposable incomes, produces a drastically different picture.
Note: A simple and perhaps even better alternative is to put debt in relation to GNI.
Undoubtedly, Sweden still stands out, both in comparison to other countries and from a historical perspective. However, on both accounts, developments have been less dramatic when using the adjusted disposable income measure.
To sum up: This short macro note was meant to highlight the very common mistake (been there, done that) of not fully thinking through choices of data and what those imply (because it’s never intentional, right?). In the words of Benjamin Disraeli, at least according to Mark Twain: “There are three kinds of lies: lies, damned lies, and statistics.”