Is it possible that some non-trivial part of that number comes not from increased efficiency but from losing some energy-dependent industries?