4 ms·

It is possible to find an infinite number of differences between Switzerland and any other country. It is very tempting to look for reasons not to update your b

by nickff 11y ago

It is possible to find an infinite number of differences between Switzerland and any other country. It is very tempting to look for reasons not to update your beliefs in response to new data, but we must not give in to this temptation. Bayesian logic would dictate that events in Switzerland should cause us to be more skeptical of arguments in favor of inflation.

evanpw 11y ago
Absolutely, but to know how much more skeptical to be, you have to know likely this outcome was, given your prior hypothesis. And for that purpose, knowing ways in which this situation differs from textbook, across-the-board deflation is very relevant. P.S., I found this discussion (http://www.wsj.com/articles/john-cochrane-whos-afraid-of-a-little-deflation-1415919082 http://www.wsj.com/articles/john-cochrane-whos-afraid-of-a-l...) to be pretty convincing. His guess is that sudden, dramatic, Depression-style deflation will have effects that are a lot different from slow, predictable, Swiss-style deflation.
marcosdumay 11y ago
The same theory that says that deflation is bad is saying that this one is not big enough to produce much effect. The result is perfectly supported by the theory. Whatever mathematical names you bring on can not change this. Now, if you came up with examples of countries with decreasing nominal salaries and a growing economy, you could make a point. There's a paper with a few in another thread.
snowwrestler 11y ago
> Bayesian logic would dictate that events in Switzerland should cause us to be more skeptical of arguments in favor of inflation. Depends on prior evidence. Current thinking on inflation/deflation did not drop into economists' heads from space yesterday. It was developed by analyzing hundreds of years of real experience in real economies. Before we alter our beliefs too much today, we should at least be aware of the full scope of evidence on the topic.
- mike_hearn 11y ago
  Unfortunately it wasn't, nice though it'd be if that were true. Current thinking is almost entirely based on extrapolating the Great Depression out to all economies everywhere, even when actual empirical studies show that the extrapolation isn't valid (see the paper I posted above).

by nickff 11y ago

evanpw 11y ago

Absolutely, but to know how much more skeptical to be, you have to know likely this outcome was, given your prior hypothesis. And for that purpose, knowing ways in which this situation differs from textbook, across-the-board deflation is very relevant. P.S., I found this discussion (http://www.wsj.com/articles/john-cochrane-whos-afraid-of-a-little-deflation-1415919082 http://www.wsj.com/articles/john-cochrane-whos-afraid-of-a-l...) to be pretty convincing. His guess is that sudden, dramatic, Depression-style deflation will have effects that are a lot different from slow, predictable, Swiss-style deflation.

marcosdumay 11y ago

The same theory that says that deflation is bad is saying that this one is not big enough to produce much effect. The result is perfectly supported by the theory. Whatever mathematical names you bring on can not change this. Now, if you came up with examples of countries with decreasing nominal salaries and a growing economy, you could make a point. There's a paper with a few in another thread.

snowwrestler 11y ago

> Bayesian logic would dictate that events in Switzerland should cause us to be more skeptical of arguments in favor of inflation. Depends on prior evidence. Current thinking on inflation/deflation did not drop into economists' heads from space yesterday. It was developed by analyzing hundreds of years of real experience in real economies. Before we alter our beliefs too much today, we should at least be aware of the full scope of evidence on the topic.

mike_hearn 11y ago
Unfortunately it wasn't, nice though it'd be if that were true. Current thinking is almost entirely based on extrapolating the Great Depression out to all economies everywhere, even when actual empirical studies show that the extrapolation isn't valid (see the paper I posted above).