2 ms·
There are two other reasons for having a small but nonzero level of inflation. Your example (that you need a nonzero level in order to have something to reduce
by gumby 8y ago
There are two other reasons for having a small but nonzero level of inflation.
Your example (that you need a nonzero level in order to have something to reduce -- else your only alternative is asset sales which is a much worse option for many reasons) is probably the most important. Another is that measuring the economy is quite inexact (or coarse) and has a lot of latency, so you need a little "sloop" and you want that to be positive rather than negative for the reason you gave. This is no different than any other engineering consideration: you want your characteristic error to be slightly small or slightly large depending on how the piece you're making fits with another. Third, a little inflation encourages investment; if you leave all your money in the mattress it will lose (hopefully only a little) value over time, so you are encouraged to at least put it in a bank, if not into some more productive asset.
In general I think the last point is an important and good one, but a downside is that people who do not have much savings (live paycheck to paycheck) and who have no other defence against downward wage pressure do suffer what you nicely refer to as "de-facto theft through the back door" and that is an overlooked source of real misery. I'm not sure what central banks can do about it though; seems like it should be the responsibility of the actual government.
By the way ultimately the Fed's reductio job description (pace the "dual mandate") is to make the money supply match the level of economic activity as closely as possible, nothing more. If you don't issue new currency but more people are born, consuming more haircuts and building more homes, then you end up with deflation, which causes the economy to grind down and makes everyone miserable. If you issue to much you get inflation which causes the economy to grind down and makes everyone miserable. They have a small number of tools to do it and pretty crude observational tools but still have ended up more consistent than when currency was linked to stuff not at all correlated with economic activity (the availability of metals). Just look at the economic crashes (sometimes more than one a decade) of the 19th century triggered by the gold strikes of California, the Yukon and Victoria, as examples.