Internal vs External Validity: What Can Galileo and Newton Teach Us About Econometrics?

What do falling balls from the Leaning Tower of Pisa have to do with modern economic research? In this video, I explore the fundamental tradeoff between internal and external validity using one of science's most famous thought experiments.

Starting with Galileo's legendary experiment, I explain how perfectly controlled studies give us clean causal identification but limited generalizability. Then I show how Newton's theoretical approach—what economists call "structural modeling"—allows us to predict outcomes in entirely new contexts, just like predicting your weight on Mars.

The key insight: the best empirical economics doesn't choose between reduced-form and structural methods. Instead, it combines both approaches to understand not just what happened in one study, but why it happened and how those mechanisms apply elsewhere.

Key Topics Covered:
Internal validity: achieving clean causal identification through experimental control
External validity: generalizing results beyond the original study context
The reduced-form vs structural modeling spectrum in economics
How structural parameters (preferences, technology) remain stable across contexts
Why the future of empirical economics requires combining both approaches

Whether you're running randomized controlled trials, building structural models, or trying to understand what economic research can really tell us about policy, this video will change how you think about the credibility and applicability of empirical results.
Perfect for economics students, researchers, and anyone interested in how we can use data to understand human behavior and predict policy outcomes.

Related concepts: causal inference, identification strategy, generalizability, deep parameters, reduced-form econometrics, structural econometrics, randomized controlled trials (RCTs), policy extrapolation

Tyler Ransom is an Associate Professor of Economics at the University of Oklahoma. Subscribe for more videos on data science, econometrics, and research methods!

Editing credit: @neiljohnmanllios3064

#internalvalidity #externalvalidity #econometrics #causalinference #Galileo #Newton #structuralmodeling #reducedform #economicsresearchmethods #experimentaleconomics #identificationstrategy #generalizability #RCT #randomizedcontrolledtrial #empiricaleconomics #researchdesign #credibilityrevolution #policyevaluation #economictheory #structuraleconometrics #deepparameters #economicstutorial #PhDeconomics #appliedeconometrics #researchmethodology