Causation is just a mirage…
March 7th, 2009
… and when you get a bit closer it’s just spurious correlation. If only economists could do randomised experiments like real scientists.
Categories: Cartoons
Comments are closed.
Recent Posts
- Drink heavily tonight
- Marginal rates, interest rates, and NZ monetary policy
- Why all the hating on DSGEs?
- If we think there is an implicit social dividend from land …
- Don’t stop limiting our competitors ability to compete say rivals
- Testing my economic knowledge
- But what is the problem?
- Housing tax changes – will they lead to a leap in rents?
Recent Comments
- Matt Nolan on Drink heavily tonight
- Eric Crampton on Drink heavily tonight
- Tihomir Ipotpaliev on It’s funny because it’s true …
- Matt Nolan on Why all the hating on DSGEs?
- Andrew Coleman on Why all the hating on DSGEs?
- Matt Nolan on If we think there is an implicit social dividend from land …
- Matt Nolan on Don’t stop limiting our competitors ability to compete say rivals
- Chris on If we think there is an implicit social dividend from land …
- rauparaha on Don’t stop limiting our competitors ability to compete say rivals
- Matt Nolan on If we think there is an implicit social dividend from land …
“If only economists could do randomised experiments like real scientists.”
You may want to check out the journal “Experimental Economics” and note that Vernon L. Smith got the Nobel Prize “for having established laboratory experiments as a tool in empirical economic analysis, especially in the study of alternative market mechanisms”.
@Paul Walker
Paul, thanks for your comment. I am aware of the field of experimental economics. However, as I’m sure you know, its results are often accused of being inapplicable to the ‘real world’. Furthermore, experimental economics hasn’t shed much light on the major problems in macroeconomics which are usually the focus of causation/correlation debates lampooned in this cartoon.
Hehe XKCD
@Paul Walker
Hi Paul,
Even Vernon Smith himself felt that current laboratory experiments are severly constrained insofar as what type of empirical evidence they provide.
He is confident that using experiments has shown that individuals to act in an “unconsiously rational” way (http://www.tvhe.co.nz/2008/11/01/quote-6-vernon-smith-unconscious-optimisation/). However, we already anticipate agents that act rationally in economics – but we still face the correlation vs causation issue because there may be “multiple ways” the agents may act rationally, or multiple types of rational expectations that agents may hold.
Laboratory experiments are great ways of testing our priors – but even with this the issue of correlation vs causation remains difficult.