Jules Siedenburg (in this issue of DPR) has misunderstood the purpose of our 1994 book, which was to differentiate between long-run forces of change and the impact of particular projects. The methodology we adopted is rehearsed in this response, as it authenticates our data. His attack on our use of price data is rebutted. From our data we expanded our original model and derived the policies that had encouraged farmer investment and technology change. Our work can be rebutted if these policies fail, not by the mere existence of continued degradation or failure to adopt particular technical recommendations. We have since tested our model in comparable areas of dryland Africa, and the cited works by others generally support our findings.