If you only read one thing this week…


Evidence based poverty reduction and the Millenium Villages by Nick
November 30, 2011, 11:28 am
Filed under: If you have 30 mins, measuring results

So – here’s the dream – figure out how to reduce poverty, scientifically. Find ways to do this that you can show, with numbers, actually work. What a seductive idea. And how slippery and evasive. Let’s look at one of the flagship projects, the Millennium Villages.

They are the poster child of Jeffrey Sachs of the Earth Institute at Columbia, along with UNDP. The idea is to take Sach’s theories of poverty traps (if you need to brush up on this stuff it’s all in ‘The End of Poverty’ by Sachs), and design a series of pilot interventions that demonstrate, empirically, that an integrated approach to rural development with consistent adequate funding over time can achieve the goal of lifting communities out of the poverty trap (and meet the Millennium Development Goals while we’re at it).

There are 13 of the sites, in 10 countries in sub-Saharan Africa, plus a bunch of similar type projects all over the place. They started as a five year project in 2006 and just got renewed for another five. So – where are we as far as empirical proof of poverty reduction is concerned? Well, it’s, erm, a mixed bag, and sort of depends on how you look at it…

ODI did one of the earliest reviews of it in 2008 (read the full pdf here) – a little early to really show impact on poverty reduction, they do claim crop productivity increases (although only alongside additional agriculture inputs, so there isn’t really a lot of income increase). There does seem to be some malaria reduction, and that’s good, although it’s not entirely clear how cost effective this public health program is compared to traditional ones, since the whole thing is a big integrated package.

ODI signs out with some rumblings about sustainability of gains and issues of scaling up, but they are basically optimistic. Not so the Soils, Food and Healthy Communities group, who claim that their Ekwendini project villages that use sustainable agricultural techniques and crop diversification achieve better agricultural gains with far less cost.

Fast forward to this week, when we get the first independent evaluation of the MVPs, by Kenyan economist Bernadette Wanjala of Tillburg University. She looked at 236 families in the project and 175 control families in the same districts (read more here). She was hoping to find that the MVP project (which spends the equivalent of 100% of local per capita income per year on each beneficiary) would be able to show some empirical increase in recipient income. Unfortunately, she couldn’t find any evidence for it.

How can that possibly be? Well, let’s dig in a little more. She did find a 70% increase in agricultural productivity, which did tend to increase agricultural income in those families (at a substantial input cost). The problem was that those families that spent more time farming seem to be spending less time doing other profitable activities, and their income in those other areas fell, canceling out benefits.

The MVP group claims that ‘incomes are rising’, and that the project is ‘enormously successful’, but unfortunately doesn’t publish their data on incomes.

So where does this leave us? The MVPs are a bold attempt to test theories of change on a village level, and five years on we’re still arguing about what the data mean. It’s encouraging to see donors and academics entering this field of research, and perhaps not surprising that the answer is “it’s complicated – we need more research”.

Advertisements

Leave a Comment so far
Leave a comment



Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s



%d bloggers like this: