There is a new edition coming out in a couple of weeks, but until then, the data on coverage from the 2010 State of the Humanitarian System caught my eye.
On average, total humanitarian contributions equalled over 85% of total stated requirements in 2007 and 2008, compared with 81% in 2006 and only 67% in 2005. However, the needs of affected populations have gone up as well, and are still not matched by resources, so the result is a nearly universal perception of insufﬁciency, despite quantitative evidence of progress.
So; yet again; things are getting better, but they are still really bad. Unmet humanitarian need means that people are dying for the sake of what is really pennies in the grand scheme of things (overall funding for the sector is around $7 billion, considerably less than, say, the $50 billion US residents spend on their pets each year).
If you take the data at face value, the simple framing of the problem as a quantifiable one with some kind of definite limit, I find incredibly encouraging. Perhaps the really interesting question though is what that data really means. "Humanitarian need" is presumably some kind of function of the actual need on the ground, but processed through the humanitarian agencies on the ground who report those needs to funders. And the agencies are locked in a very long repeated strategic game with those funders - for instance agencies have incentives to overstate need in order to get more funding. But donors know this. And agencies rationally prefer funding before a disaster happens rather than after, but it is presumably harder to both fund raise and demonstrate impact when you have averted a crisis (i.e. nothing happened) than when you responded to one has started (i.e. something is clearly happening). And what on earth is the line between humanitarian need and development need?
You may have noted a couple of "presumably"s in that last paragraph. Anyone know of any studies on game theory and the political economy of humanitarian fundraising?