E3ME’s origins lie in post-Keynesian economic theory. This is a defining characteristic of the model and contributes to several of E3ME’s unique advantages over other macroeconomic models. For example, E3ME includes a realistic representation of the financial system and pays particular attention to unemployment and the labour market.
Following from the early writing of Keynes, E3ME accepts that there is fundamental uncertainty in the economy. Economic agents are not able to optimise their decision making and base behaviour on their limited knowledge. This approach contrasts with that in standard Computable General Equilibrium (CGE) models, in which agents are assumed to have perfect knowledge and behave in an optimal manner. E3ME also rejects assumptions of perfect competition and fully flexible prices.
The result is that in E3ME the economy does not automatically operate at full capacity. The level of output is determined by the level of aggregate demand, which is usually less than the level of potential output. This approach is consistent with reality, as data on involuntary unemployment and economists’ attempts to measure the ‘output gap’ show.
This means that E3ME can be used to assess scenarios that include stimulus or austerity measures, as well as policies relating to the efficient use of resources in the economy. Previous analysis has shown that interactions between the real economy and the financial system are key to determining model results.
Because E3ME rejects theories of optimising behaviour, an alternative way of modelling human behaviour is required. Behavioural parameters are obtained from econometric estimates based on time-series historical data. This is why E3ME is often referred to as a macro-econometric model.
The econometric approach makes E3ME highly empirical in design. The assumption that behaviour in the past can be extrapolated into the future is sometimes questioned and it is acknowledged that the degree of uncertainty around model results increases when long-term scenarios with structural change are assessed. However, the econometric estimates still provide a best, non-biased guess of economic behaviour and do not rely on assumptions about optimisation that behavioural economists have repeatedly disproven.