One idea he discussed was that of "isomorhic mimicry" - basically, in nature when something takes on the appearance of another thing but without any of the underlying, more complex form. And his proposal is that donors and governments too often take institutional reforms from elsewhere, but that although these bear a resemblance in formal terms, they do little in terms of substance as they fail to reflect the full original set of institutions.
For anyone who has worked in a developing country, the fact that governments adopt policies and strategies, often for very little to happen is nothing new - it's basically the "implementation deficit", or saying you'll do things but not doing them. In some cases this may be unintentional, although in many cases it is strategic - a memorable description of this kind of thing was given by The Economist some years ago:
"Over the past few years, Kenya has performed a curious mating ritual with its aid donors. The steps are: One, Kenya wins its yearly pledges of foreign aid. Two, the government begins to misbehave, backtracking on economic reform and behaving in an authoritarian manner. Three, a new meeting of donor countries looms with exasperated foreign governments preparing their sharp rebukes. Four, Kenya pulls a placatory rabbit out of the hat. Five, the donors are mollified and aid is pledged. The whole dance then starts again." The Economist (19 August 1995).So calling it isomorphic mimicry is really just giving a complicated name to something we all recognise when we see it. Still, maybe it is a useful concept to think more clearly about the distinction between superficial policy appearance and actual policy substance.
The second idea presented was Pritchett's classification of types of policy problem. To oversimplify a little, instead of lumping together policy issues by their sector or target audience or some other common factor as is commonly done, he distinguishes between different types of policy reform according to the number of face-to-face transactions required for the policy to be implemented. That is, where there are few transactions required, and less room for subjectivity or negotiation, it is generally easier to implement a policy. Although he goes into a more detailed typology, the main extremes are what he calls "logistical" problems, that is getting vaccines to patients, or building schools; to far more complicated ("wicked hard"!) problems such as tax collection, where the number of face to face transactions are high, there is room for negotiation, so there is far more room for things to go astray. His view is that performance has been relatively good on the former type of problems, and weaker on the latter, precisely because of their nature.
While this is also actually a pretty simple idea, it is nonetheless a useful insight: we often fail to think properly about the requirements for certain policy reforms to meet with success on the assumption that all policy reforms are more or less the same. Perhaps, then, such a way of thinking about these would assist us in taking the steps to try and turn some of these trickier, high-transactions policies into more manageable bits.
But what we usually think about when we talk of failure to implement these days is political economy. More and more is being written about how a better understanding of context, actors, their history and motivation can help understand why governments agree to certain policy reforms and then fail to implement. (A whole bibliography of readings is available here, while a recent paper summarises the issues here.) The implication is that by understanding motivations, we can better alter the incentives at the margin to achieve better policy results. Although the political economy aspect was not explicitly mentioned in the presentation, it would seem that indeed the two approaches may be highly complementary: not only do you need to understand the types of actors and their motivation etc., but also the nature of the policy itself.
The next question is harder though. Even once you understand all that, how much can you actually say or do to improve the policy outcome? Or do you just understand better why it didn't work the first time around...?