In Report 2 we have given breakeven variances and c.v.'s for different loss functions, corresponding to uncertainty in estimated true population total. There we considered the U.S. divided into 51 areas: D.C. and the 50 states. Here we likewise consider only the 6 New England states ("Division 1": Maine, NH, VT, MASS, CT, RI). For these states we do calculations for each of 3 artificial populations considered by Isaki, Diffendal and Schultz (IDS) in their studies of small-area estimation. For each of these 3 populations, as discussed in IDS's SRD Technical Report 87-02, "Report on Statistical Synthetic Estimation for Small Areas", there are artificial census and true population counts, based on the experienced census counts and substitutions.