Just for the sake of argument, I'd ask everyone to consider the counterfactual. Let's talk about a normal year, for starters, to have a base to talk about. Sup

Just for the sake of argument, I'd ask everyone to consider the counterfactual. Let's talk about a normal year, for starters, to have a base to talk about. Suppose we have: 5% increase in nominal money supply Which breaks down into 3 parts: population growth, productivity growth, and inflation. So for example, in a normal year we might have: 1.5% increase in population So that means we have 3.5% leftover that either has to be accounted for by productivity growth or inflation. In an average year we might have: 1.5% productivity growth 2% inflation (There would actually be some small fractions here, but I'm simplifying the math because this is just a Hacker News comment. Feel free to work out the precise numbers yourself.) Typically, we know the population growth fairly well, and we think we know the inflation number fairly well, so the estimate for productivity growth amounts to "Whatever we have left over after we subtract out the population growth and the inflation." But what if we got inflation wrong? What if there were extraordinary events happening that make it more difficult to get an exact number on inflation? Then we would end up with a wrong productivity number. If we overestimate inflation by 1% then we underestimate productivity by roughly 1% (qualifiers: not exactly, but good enough for this comment). Remember, the inflation estimate is a weighted average of millions of products. A radical swing in one part of the index might have strange results, if the weighting is not exactly correct, or if the weighting relies too much on historical averages which are no longer relevant. All of which is to say, we need to wait a year or two to know if this trend is real. It could just be a statistical illusion.