That [starting a small rock-smashing business] may help her stave off extreme deprivation, but it won't turn Malawi into a developed country. That's certainly not how Europe or Asia became industrialized. Without massive state-driven investments in, among other things, infrastructure, legal institutions, health, and education, the markets in these countries will be too stunted for "entrepreneurs" with microloans to do much more than set up rock-smashing businesses or sell bananas on the side of the road. I really don't want to denigrate that, but as a poverty-reduction strategy, it's no substitute for proper economic development.Now, to a certain extent, this is obviously true. The history of 'proper economic development' has, by and large, been the story of massive state investments in things like that. But the history of failed development has also largely been a history of massive state investment. And one central reason for this is that while massive state investment, in any given case, may or may not be good for the area as a whole, it is almost always good for somebody, and that somebody usually has some say in the matter.
That's not to deny that there's something compelling about the "big push" idea of virtuous circles of development and the multiple equilibria available to a given economy; there obviously is. But if getting from here to there were remotely easy, we wouldn't have people like Bill Easterly making a career out of pointing how badly we've fucked it up for half a century.
One reason why bottom-up development has never 'worked' could simply be that any time you get a reasonably strong state, you're going to have very strong incentives to have that state Do Something. Exactly how much of the resulting development (in the good case) is a result of the Something that gets Done, and how much is from the industrious rock-smashers going about their business, is difficult to say; it's made still more difficult by the fact that state intervention will naturally divert entrepreneurial energy towards those things that the state is pushing forward or propping up.
So, three points. First, the way things have been done in the past might not be the best model to follow; we don't want to encourage others to duplicate what were, at the time, maybe-not-even-necessary evils (enclosure is the obvious bogeyman here, with intellectual property its contemporary heir). Second, microcredit has one clear advantage over massive state investment: it is unlikely to further entrench wretched political-economic equilibria. Third, the sort of growth microcredit is likely to encourage is likely to be the sort least well-tracked by standard national-income accounting, and we should recognize this in our assessments of it.