It was shown in Chapter 5 that the s tabil ity 0 f eq uil i bri urn depends on the values of the parameters like a, sand k. Here, we have seen that if OJ can be assumed constant, there is an additional condition (8.85). This can perhaps be interpreted as follows: if an equilibri urn value of Dj is calculated using . Harris and Wilson (1978) methods, as in Chapter 5, say as ojqUll then the ability to achieve a stable equilibrium in a simulation will require (8.85) to hold. Since it is a condition on para-are not j-dependent, this presumably
will be difficulties in simulation in any cases where it is not satisfied. White (1977). for example. has reported simulations of this type which have not converged.