While some of it may be correct, that's a pretty terrible paper to use as evidence of anything, not only because it's incredibly biased but also because almost all of it is based on explorations of case studies. I have read a lot of these and contributed to some: it does this because it's pushing policy objectives. It's also from 17 years ago.