Jeffrey Sachs on the Millennium Villages Project
What this covers
Jeffrey Sachs of Columbia University defends the Millennium Villages Project against criticism from development economists in this conversation with Russ Roberts on EconTalk. Sachs presents the project as a proof-of-concept for integrated rural development across roughly ten clusters in Africa, involving approximately 500,000 people and financed at around $120 per capita annually—half from the project itself and half from local partners. He argues that this Integrated Development Strategy attacking hunger, disease, and illiteracy simultaneously produces synergies that single-issue interventions cannot match. Roberts, while acknowledging certain successes, presses Sachs on whether the project's ambitious transformational claims can be credibly measured and whether its gains will prove sustainable once external funding ends.
The conversation ranges across the mechanics of agricultural productivity, primary health care delivery, and the prerequisites for functioning markets. Sachs credits the project with catalyzing Africa-wide scale-ups in bed nets for malaria control and training Community Health Workers at minimal cost, arguing these represent genuine Spillover Effects in Impact Evaluation rather than village-level achievements alone. He points to technology innovations—off-grid solar prepaid systems and digital soil mapping—now adopted across multiple countries as evidence of a market test. Roberts counters that government adoption is a weaker signal than private Revealed Preference, questions whether productivity gains translate into sustainable livelihoods when prices fall, and challenges whether top-down crop and input choices constitute real market decisions. The two also spar over whether the project's success should be measured by present improvements or the durability of prospects for the next generation.
Sachs argues that the Millennium Villages Project's integrated, goal-based, low-cost development strategy has produced scalable, life-saving successes (especially in malaria and primary health care) that critics have misrepresented, while Roberts counters that the project's ambitious transformational claims cannot be credibly evaluated and may rest on top-down choices and unsustainable hopes.
- Sachs claims integrated development at ~$60/capita can deliver health, agriculture, education, and infrastructure gains with strong synergies
- Sachs credits the project with helping drive Africa-wide malaria control scale-up rather than just village-level effects
- Roberts questions whether productivity gains create sustainable markets and whether the project can be independently evaluated
Genuine market sustainability requires voluntary exchange where villagers buy and sell their own choices, not external choices made for them through aid or government programs.
- What a functioning market does is allow you to better yourself by bettering others—you help other people by doing something they value and they pay you for it—so the central sustainability question for the project is whether there is a genuine market in which villagers can buy and sell, not just temporary aid-funded activity.
“what I wrote after that interview was: We better ourselves by bettering others. And what a market does is it allows you to help other people by doing something that they value and they pay you for it”
- Even with good intentions and good outcomes, giving farmers productivity-enhancing seed involves decisions—which seed, which crops—that the farmers did not choose through market processes or a vote and were not given money to decide, so these are not really market decisions and that non-market character may affect the project's effectiveness.
“that isn't what they chose to do on their own. I don't think they picked the kind of seed... there are decisions that are made that are not really market decisions. And it seems to me that is going to affect the effectiveness of the project”
- Economists know what markets are capable of once established and that they are buttressed by culture, norms, and unobservable factors, but we do not actually know how to create markets or how to help them emerge—they have to be allowed to emerge—which makes top-down attempts to seed market activity epistemically fraught.
“I don't think we know very much as economists about how to create markets. We certainly understand what they are capable of doing once they're established... we don't know how to help them emerge”
- Even though crop yields rose two- to three-fold, this may not be a scalable route out of poverty because productivity increases lead to lower prices and agricultural profit margins are typically small, making it difficult and challenging to get large numbers of farmers above subsistence simply by making them more productive.
“The increases in productivity are going to lead to lower prices. Profitability of agricultural tends to be--profit margins are quite small. It seems it's going to be difficult and challenging to get lots of farmers above subsistence”
Aid promises that fail to deliver durable transformation are deeply cruel, especially when they exploit emotional compulsion rather than test sustained outcomes.
- It is one of the cruelest things in the world to come to a group of people, set their hearts on fire promising the magic of expertise, wisdom, and money will transform their lives, and then through one's own hubris smash that dream of a better life—and such optimistic transformational arguments tend to win precisely because they are emotionally compelling.
“it's one of the cruelest things in the world to come to a group of people, set their hearts on fire saying I'm going to change your life... and to take that dream... and to smash it, and through your own hubris”
- The fundamental question for evaluating the project is not whether people are currently alive or better off during the intervention, but whether the prospects of the next generation—the 14-year-old boy or girl in those villages—have genuinely and durably improved; if the project encouraged belief that broad outside money would transform lives and it does not, that outcome would be tragic.
“Have the lives of the next generation, of the 14-year old boy or girl in those villages, have their prospects improved? It's great that they're alive. I don't disagree with that.”
Impact assessment must separate the cost-effectiveness of narrow interventions from the broad expensive ambitions claimed for entire integrated projects.
- The fundamental economic question is not whether bed nets reduce malaria—they do—but whether the cheap, narrow intervention everyone accepts is superior to the much more expensive integrated approach Sachs is selling; the challenge is to distinguish the bang-for-buck of the cheap intervention from the broad ambitious claims made for the project as a whole.
“The question is whether that inexpensive, as you concede, approach is superior to the integrated approach which you are selling”
- Roberts welcomes Sachs back to the program.
“Jeff, welcome back to EconTalk.”
Government adoption of programs is a weaker test of merit than sustained private donor funding based on genuine cost-effectiveness.
- Government choice to adopt a set of approaches is a weaker form of market test than continued voluntary donor funding, because governments choosing particular programs reveals less about genuine cost-effectiveness than would private donors continuing to fund the project on its merits.
“you're talking about governments choosing a particular set of approaches. I would be a little more impressed maybe if the donors were continuing to fund it”
Spillover benefits to neighboring areas complicate causal attribution when comparing project villages to control regions.
- Because bed-net distribution and other interventions spread more widely than just the project villages (partly based on lessons from the villages), a comparison of child mortality within the villages versus neighboring areas may not capture the true impact of the project, since the neighbors also benefited from the scaled-up intervention.
“I'm happy to give you credit that the adding of bed nets in the villages made it easier to add them outside the villages, so that a comparison of child mortality within and compared to neighboring villages may not be suggestive of the real impact”