An Introduction to Empirical Analysis
Econometrics is the use of statistical methods to analyze economic data, test economic theories, and evaluate government and business policy.
A typical empirical project involves a few key steps:
The type of data we have dictates the methods we can use. Here are the main types:
Many different units (people, firms, countries) at a single point in time.
A single unit (a country, a company) observed over multiple time periods.
Combines two or more different cross-sections from different time periods.
The same cross-sectional units are followed over multiple time periods.
You have data on the unemployment rate for every county in the United States for the years 2000 and 2010. What kind of data structure is this?
This is a Panel Data set. The key is that you are observing the same units (counties) at multiple points in time (2000 and 2010).
If you had data on a different set of counties for each year, it would be a pooled cross-section.
The goal of most empirical studies is to estimate a causal effect. We want to know the effect of a variable on another, ceteris paribus ("holding other relevant factors fixed").
Example: The Return to Education. What is the effect of one more year of education on wages, holding innate ability, experience, and all other factors constant?
This is incredibly difficult with nonexperimental (observational) data, because we can't run controlled experiments. We can't randomly assign education levels to people! People choose their education, and that choice is likely correlated with other factors that affect wages (like ability).