Because economic output always builds on previous output. That is the definition of an exponential function, and to reverse those effects you need to take the log.
It's easy to see this - just look at the early years, and notice the curve - it's an almost exact exponential curve. Graphing an exponential curve teaches you very little about the underlying data.
A logarithmic view would be great for showing when the population and output stalled or increased faster, but the purpose of this graph is to show that the fact that the increase is dramatically exponential, and at what point in the curve we are right now.
Ideally both would be shown--but just showing the log graph would be misleading in this context.
I agree. I think that the point of this graph is to convey exponential growth, and simply that. It's not meant to be used "seriously," which'd make a logarithmic graph more suitable.
Ars is correct on this one (and also resdirector, the topvoted comment right now)- the issue is not money inflation, it's the simple fact that compounding growth will be exponential, always. So if your point is to show that growth rates increased in recent centuries, then you need to show growth rates, or show this but using a log scale.
Logarithmic scale would better show the percent change of population in the early years. Using linear scale hides the movement of the early years because the 20th century's magnitude is so great that the previous centuries look identical to each other. It's more dramatic this way, but perhaps misleading.
But that is the point of the graph - to show the ratio of history that has happened has happened recently. Or in other words, the point is not to show the growth patterns, but to show the nominal value.
I think that's where it's wrong, though; imo, "history" is better identified with the percent change, not with the nominal value. If someone improved some technology by 10% in 1850, and someone else improved it by another 10% in 1950, I'd call those historically equivalent contributions.
Comments
And yet another non logarithmic graph of money.
I'm really surprised to see such ignorance on The Economist. People who write about money for a living should know better.
The scale is % of economic output and years lived. Assuming the output is inflation adjusted, why would you prefer the scale to be logarithmic?
Agreed, since economic output is measured in 1990 Dollars, there's no need for a logarithmic scale here.
Because economic output always builds on previous output. That is the definition of an exponential function, and to reverse those effects you need to take the log.
It's easy to see this - just look at the early years, and notice the curve - it's an almost exact exponential curve. Graphing an exponential curve teaches you very little about the underlying data.
A logarithmic view would be great for showing when the population and output stalled or increased faster, but the purpose of this graph is to show that the fact that the increase is dramatically exponential, and at what point in the curve we are right now.
Ideally both would be shown--but just showing the log graph would be misleading in this context.
I agree. I think that the point of this graph is to convey exponential growth, and simply that. It's not meant to be used "seriously," which'd make a logarithmic graph more suitable.
Ars is correct on this one (and also resdirector, the topvoted comment right now)- the issue is not money inflation, it's the simple fact that compounding growth will be exponential, always. So if your point is to show that growth rates increased in recent centuries, then you need to show growth rates, or show this but using a log scale.
Why should it be logarithmic?
Logarithmic scale would better show the percent change of population in the early years. Using linear scale hides the movement of the early years because the 20th century's magnitude is so great that the previous centuries look identical to each other. It's more dramatic this way, but perhaps misleading.
But that is the point of the graph - to show the ratio of history that has happened has happened recently. Or in other words, the point is not to show the growth patterns, but to show the nominal value.
I think that's where it's wrong, though; imo, "history" is better identified with the percent change, not with the nominal value. If someone improved some technology by 10% in 1850, and someone else improved it by another 10% in 1950, I'd call those historically equivalent contributions.