Bayesian econometrics is the branch of econometrics that treats unknown model parameters as uncertain and updates beliefs about them using data.
$$$$
Core logic
The method combines:
- a prior distribution for the parameters,
- a likelihood from the observed data,
- a posterior distribution after updating.
In shorthand:
$$
\text{Posterior} \propto \text{Likelihood} \times \text{Prior}
$$
The result is not a single point estimate only. It is a probability distribution over plausible parameter values.
Why economists use it
Bayesian methods are especially useful when:
- samples are small,
- models are complex,
- prior knowledge matters,
- the researcher wants full probability statements about parameters or forecasts.
Macroeconomists often use Bayesian methods in DSGE models and time-series forecasting because priors can stabilize estimation when many parameters are hard to pin down from data alone.
Practical interpretation
Suppose an economist is estimating how strongly inflation responds to unemployment. If the data are noisy, a purely data-driven estimate may jump around. A Bayesian approach can incorporate prior knowledge from earlier studies and produce a posterior distribution that reflects both sources of information.
Knowledge Check
### What distinguishes Bayesian econometrics from a purely classical approach?
- [x] It combines prior beliefs with data when estimating parameters
- [ ] It ignores data and uses theory only
- [ ] It never produces forecasts
- [ ] It treats all parameters as known constants
> **Explanation:** Bayesian econometrics updates prior beliefs with observed evidence instead of relying on the sample alone.
### What is the posterior distribution?
- [x] The updated distribution for parameters after combining priors and data
- [ ] The raw data set before estimation
- [ ] The same thing as the likelihood alone
- [ ] A measure of only model fit
> **Explanation:** The posterior summarizes what the researcher believes after the evidence has been incorporated.
### Why are Bayesian methods often attractive in macroeconomics?
- [x] Because complex models with many parameters can be hard to estimate from data alone
- [ ] Because macroeconomics contains no uncertainty
- [ ] Because priors replace all empirical work
- [ ] Because Bayesian methods are used only for household surveys
> **Explanation:** Priors can regularize estimation and improve inference when models are highly parameterized or data are limited.