A blog by Patrick Crozier

« State failure #25: State education | Main | Pet theory: why Bush will win »

September 03, 2004

How should one go about adjusting for inflation?

I found myself asking this in "Subsidy for the UK rail industry"

Usually, the answer is simple. Take the price, take the GDP deflator for the time, multiply and, hey presto, you have the price in today's money.

But there's a problem. As a blogger I want my posts to stand the test of time. I want them to still be read in 2024, 2044 and even 2104. But by then (assuming inflation continues to chunter away) 2004 prices will seem pretty meaningless.

So, rather than adjust prices to today's date, wouldn't it be more sensible to adjust them to some date in the past? That way you only have to make the adjustment once. So, rather than express everything in today's money, I could express it in 2003 money or 1803 money.

But if so, what date? Two dates spring to mind. The first is the year 2000. Nice round number. Reasonably recent. The second is 1900. This has the advantage of being at a distance from the inflations of today that are, incidentally, still going on. It also means that prices can be expressed in good old pounds, shillings and pence which would be a good way of preparing people for currency restoration.

The only problem is that it seems a bit weird. But if I take the lead and enough people follow suit, maybe, just maybe …

Actually, there is another problem. Knowing what prices were is not much use unless you also know how much people earned.



How about adjusting to % of GDP / GDP PC for large and small numbers resectively?

Posted by john b on September 7, 2004

Interesting idea. Let's get this right.

Big numbers - express as % of GDP
Small numbers - express as proportion of GDP per capita

Hmm. I'm grasping at straws here but the two seem to be qualitatively different.

Posted by Patrick Crozier on September 8, 2004

Convert it to gold at spot rate.

Posted by Guido Fawkes on September 9, 2004

Yup, you're right. I guess doing everything compared to GDP per capita would probably be fairest.

Spot rate gold could work, although you'd get some strange effects from gold market fluctuations.

Posted by john b on September 9, 2004

The strange fluctuations might be in the fiat currency...

Posted by Guido Fawkes on September 14, 2004

Correct for growth in M4 i.e. the "money supply", or to those in the know, the level of currency defraudment.

Posted by Rob Read on September 14, 2004