Mathematical and Quantitative Methods
This field (JEL category C) comprises the mathematical and statistical methods of economics — above all econometrics, the application of statistical inference to economic data for measurement, testing, and forecasting.
Scope
It covers econometric theory and methods (regression, time-series, panel, and microeconometrics), mathematical and computational methods, game theory as method, and experimental design, providing the quantitative toolkit used throughout economics.
Sub-topics
- General
- Econometric and Statistical Methods and Methodology: General
- Single Equation Models • Single Variables
- Multiple or Simultaneous Equation Models • Multiple Variables
- Econometric and Statistical Methods: Special Topics
- Econometric Modeling
- Mathematical Methods • Programming Models • Mathematical and Simulation Modeling
- Game Theory and Bargaining Theory
- Data Collection and Data Estimation Methodology • Computer Programs
- Design of Experiments
Core questions
- How can economic relationships be measured from data?
- How can causal effects be identified and estimated?
- How should economic time series and panels be modelled?
- How are economic hypotheses tested rigorously?
- How can models be used to forecast?
Key concepts
- Regression and estimation
- Identification and causality
- Hypothesis testing
- Heteroskedasticity and robust inference
- Stationarity and cointegration
- Simultaneous equations
- Forecasting
Key theories
- The founding of econometrics
- Frisch (who coined 'econometrics') and the Econometric Society set out to unite economic theory, mathematics, and statistics.
- The probability approach
- Haavelmo recast econometrics on explicit probability foundations, enabling statistical inference about economic relationships and the simultaneous-equations program.
- Robust inference
- White's heteroskedasticity-consistent ('robust') standard errors made valid inference possible without strong distributional assumptions.
- Time series and cointegration
- Engle and Granger's cointegration and error-correction framework transformed the modelling of non-stationary economic time series.
History
Econometrics emerged in the 1930s with the Econometric Society (Frisch) and the Cowles Commission program, founded statistically by Haavelmo's probability approach (1944). Time-series econometrics (Box-Jenkins, then Engle-Granger cointegration), robust and microeconometric methods (White, Heckman), and the modern 'credibility revolution' in causal inference have successively reshaped the field.
Debates
- Structural versus reduced-form/experimental methods
- Economists debate the trade-off between theory-driven structural models and design-based, quasi-experimental approaches to causal identification.
- How to handle non-stationary data
- Spurious-regression concerns motivated the cointegration framework and ongoing debate over time-series specification.
Key figures
- Ragnar Frisch
- Trygve Haavelmo
- Halbert White
- Robert Engle
- Clive Granger
Related topics
Seminal works
- frisch-1933
- haavelmo-1944
- white-1980
- engle-granger-1987
Frequently asked questions
- Is econometrics the same as statistics?
- Econometrics applies and extends statistical methods to the special problems of economic data — observational data, simultaneity, and the need to identify causal economic relationships.
- What is identification?
- Identification is whether a parameter of interest (e.g., a causal effect) can in principle be recovered from the data and assumptions, separate from how precisely it is estimated.