I suppose part of the difficulty is that no matter how good your process, you still need an 'objective function', and that will always be defined by ideology.

I suppose part of the difficulty is that no matter how good your process, you still need an 'objective function', and that will always be defined by ideology. I agree that having the measurement of results at least partially based on hard data would be better though. Using measures that are less crude than simply "Is GDP higher this year?" might help as well. Unfortunately in politics you are unlikely to get any kind of consensus on what this 'objective function' ought to be. Perhaps this is partly why smaller more homogenous societies like the Nordic countries are often considered to be better governed [0], it is much easier for them to come to a consensus about what society's goals should be, and therefore it is easier to strive efficiently towards those goals. [0] http://www.economist.com/news/special-report/21570835-nordic-countries-are-probably-best-governed-world-secret-their http://www.economist.com/news/special-report/21570835-nordic...