What does the "notion of replication" mean in economics?

Question

What does the "notion of replication" mean in economics?

Answer 1

I assume you're talking about replication in the sense of being able to reproduce a study or experiment.  In a field like chemistry, all you need to do to reproduce an experiment is to follow the instructions: mix chemical A with chemical B, and see what happens.

In contrast, most economic studies are "natural experiments," where we look at things that have already happened and try to draw conclusions from them.  For instance, one could look at the countries of the world and see if there's a correlation between government spending on education and GDP.  To do this experiment again, some might argue that you'd need another Earth, so that you could look at a fresh set of countries.

Since we can't re-run the history of the world, we have to do different things to ensure that economic studies and experiments are valid.  First, economists describe their data collection and analysis techniques, so that others can make sure these were done correctly.  Second, other economists may try alternate data sources or analysis techniques to make sure the original conclusion is robust and independent of the methodology that was employed.  Third, economists often try to find other settings in which to test their theories.  In the example above, an economist may want to look at government spending on education across individual US states, rather than across countries.