The models discussed in this part have features in common with the non-economic empirically oriented models of Part 2, and with the monocentric simulation models of part 1. With the former they share a committment to actual data from which model equations are to be estimated. With the latter they share a committment to microeconomics from which model equations are theoretically derived. At the same time, these econometric models go beyond the other model groups in a number of ways. First, there is the use of econometric techniques to estimate the models' equations. As a result the availability of data and its aggregation play an important role in determining the models' specification. Second, these models have been developed in response to certain urban policy problems and they are used to evaluate appropriate government programs or planning options. Third, these models are not concerned with optimization (as were the models of Part 3) but with market forecasting exclusively. Fourth, the solution of the models on the computer, in most cases, requires complex numerical procedures. Fifth, in all cases the models' development undergoes a number of successive major refinements and improvements incorporating attempts to deal with limitations encountered in previous phases of the models' development.