Quantitative Economics Lectures: Computational Models and Numerical Experiments

Some economic models cannot be solved with a short algebraic expression. Quantitative economics uses numerical methods to investigate their implications, including how households respond to uncertainty or how a distribution changes over time. These lectures introduce simulation, state transitions and recursive decision problems. Connect the code's operations with the model's economic structure, then evaluate approximation accuracy and sensitivity. A computed result becomes informative only when its assumptions and numerical reliability are understood.

Here's a popular Quantitative Economics video — watch it, then summarize it with one click. Or paste any other YouTube link below.

— or paste your own YouTube link —

Want to keep the key points from a YouTube video? Paste the link below to turn it into structured notes.

Why Use Quantitative Economics Lectures: Computational Models and Numerical Experiments

Frequently Asked Questions

How does simulation differ from forecasting?

Simulation generates outcomes from a specified model and parameter choices. Forecasting aims to predict observed future outcomes, so a simulated path need not be a prediction unless the model is designed and validated for that purpose.

What is a state variable in an economic model?

A state variable records information needed to describe the current situation and determine feasible future choices. Wealth, employment status or capital stock are examples, depending on the question being studied.

Why use a Markov chain?

A Markov chain describes transitions between states using probabilities that depend on the current state. It offers a tractable representation of processes such as employment changes, provided that its conditional-memory assumption suits the application.

What does value-function iteration calculate?

Value-function iteration repeatedly updates an estimate of the best attainable discounted payoff. The resulting policy describes choices by state, while convergence and approximation checks establish whether the numerical solution is sufficiently accurate.

Can a calibrated model prove a causal mechanism?

Calibration chooses parameters to align selected model implications with evidence or external information. Several mechanisms can fit the same targets, so calibration alone does not establish that the proposed causal explanation is correct.