2 Model specification
Model specification refers to: (1) variables selection, (2) causality direction and (3) functional form selection
Based on economic theory (specific subject of research) we should select appropriate variables, and assume causality directions between them in advance
| \(~~~~\)dependent | \(~~~~~~~~~~~~~~~~~~~~\)independent(s) |
|---|---|
| \(y=\) height | \(x=\) age |
| \(y=\) consumption | \(x=\) income |
| \(y=\) demand | \(x=\) price |
| \(y=\) production | \(x_1=\) labour, \(x_2=\) capital |
| \(y=\) interest rate | \(x_1=\) money supply, \(x_2=\) inflation |
| \(y=\) crop yield | \(x_1=\) temperature, \(x_2=\) rainfall, \(x_3=\) number of sunshine days, … |
- Variables in TABLE 2.1 are all numerical. However, standard econometric analysis requires continuous dependent variable, while independent variables are allowed to be qualitative (non-numerical) as well!
Qualitative variables can also be included in the equation on the right-hand side by using binary values of \(0\) and \(1\). These variables are known as dummy variables!
- In the most simple application a single dummy variable with two categories can be included, e.g. we could examine if there is a difference in average weight between males and females, i.e. if the weight (variable \(y\)) depends on the gender (variable \(x\))?
\[\begin{equation}y_i=\beta_0+\beta_1x_i+u_i~~,~~~~~~y=weight~~,~~~x=\left\{\begin{array}{cl} 1& for~males,\\0& for~females\end{array}\right. \tag{2.1} \end{equation}\]