An econometric framework for estimating a two-regime dairy structural system is presented. Failure to account for switching between regimes due to government price intervention raises the problem of selectivity bias. Further, since a structural system of equations is involved, the problem is not limited to the market associated with the intervention. Rather, bias from a single source can distort all equations in the system. The ramifications of not correcting for the bias in policy analyses are investigated.