Fixed Effects Model
- This is a statistical way of controlling for omitted variable bias when using panel data. The method is so-called on account of the fact that it holds constant ('fixes') the average differences between the determinants of a variable by using dummy variables. Thus, for example, the effect of geographical location on hospital costs might be controlled by having a dummy variable to represent each hospital's location and this would pick up the (often unobservable) effects of otherwise omitted variables (local wage effects, distance from major suppliers, etc.). The regression analysis could then focus on the non-geographical deter minants of hospital cost, such as size or ownership, having controlled for the other effects. The method in effect distinguishes between 'within group' variation and 'between group' variation where, in the example just given, the effect of 'group' (i.e. the location of the hospitals) is taken out to enable the analyst to focus on the variations between hospitals in similar locations.