Theoretical arguments suggest that codetermination has scope for real efficiency gains in terms of both productivity & job satisfaction, but the latter is difficult to measure, while the former is mostly likely to show up in long-run data sets that are difficult to acquire.