What will it take to end poverty?

As a member of Giving What We Can I have pledged to give 10% of my income to the most effective charities I can find that are working to alleviate poverty in the developing world.

I take this pledge seriously. Accordingly, I've been thinking about how to eliminate extreme and moderate poverty from the globe. The World Bank defines extreme poverty as income below US $1.25/day and Jeffrey Sachs and others define moderate poverty as income below US $2/day, which is the median poverty line in developing countries.

First the good news. Major progress has and is being made. The World Bank reports the number of people living in extreme poverty has fallen from 1.9b in 1981 to 1.4b in 2005 (The Developing World is Poorer than We Thought, But No Less Successful in the Fight Against Poverty). This reduction occurred during a period in which the world added 1.9b people.

Unfortunately, at this rate it will take until 2072 to eliminate extreme poverty. And what is worse, the number of people living in moderate poverty has increased slightly from 2.54b in 1981 to 2.56b in 2005. This probably has something to do with all the people we added; so the projected drop in the World's population growth rate from 1.8% in 1981 to 1.2% in 2005 and to a projected 0.5% in 2049 will likely help. The other part of the story is that the drop in extreme poverty is entirely due to progress made in China. Without China, the absolute number (but not the percentage) of people living in extreme poverty has actually increased.

The scale of the problem

The same World Bank report finds the mean income of the extreme poor is US $0.87/day. With 1.4b in poverty, this puts the amount of money needed to lift everyone out of extreme poverty at US $190b/year. For moderate poverty the mean income is US $1.21/day and lifting everyone out of moderate poverty would cost US $750b.

So the rough magnitude of the problems we face is around US $500b/year. This makes a hidden assumption that it is possible to transfer wealth to just the people below the chosen poverty level and nobody else and without any overheads in doing so. If it is hard to correctly target the poor, or the overheads of doing so are large, the costs could be significantly higher. All dollar values we have been discussing so far are at purchasing power parity. But a dollar goes further, roughly twice as far, in the developing world as in the developed world. This means the actual US dollar amount of funding required could be less. This is important. A dollar given to aid the poor does not equal a dollar going into the pockets of the poor. It could be more or it could be less.

The current approaches

The Hudson Institute in 2006 published the following estimates of US private sector globally focused philanthropy:

US NGOs $9.7b
US corporate international giving $4.9b
US foundation international grants $3.4b
total $18.0b

Based on this, a rough estimate of global charitable giving for international development is US $50-70b/year. This is a lot, and a lot of good can be done by carefully targeted charitable giving, but the prospects of getting this up to the US $500b/year mark seem slim.

Estimates of official development assistance (ODA) from governments and multilateral institutions such as the World Bank made by the OECD:

ODA to Least Developed Countries (2011) US $44b
Total ODA to all countries (2011) US $178b

Total ODA is starting to get towards the right magnitude, but with only 25% of it going to the least developed countries, a much better targeting of poverty would appear at first glance to be warranted.

Brainstorming solutions

If private philanthropy and ODA probably won't get us to a solution, what will?

Positive feedback

Positive feedback could play a huge role in eliminating poverty by allowing smaller investments to go to scale. Microfinance is one example. Microfinance doesn't have the same direct payoff as global health, and it has an equivocal track record, but microfinance has managed to go from nothing to serving 100 million borrowers with US $25b in loans in less than 30 years. Are there other ideas like microfinance, or possibilities for fine tuning the microfinance model to better lift people out of poverty?


New forms of taxation could readily provide the funds needed to eliminate poverty. In the UK there is talk of a Robin Hood Tax on financial transactions with 25% of the proceeds going to combat global poverty, and the WHO is considering a global tobacco tax, to say nothing of the wealth transfer represented by initial allocations to countries under cap and trade CO2 pollution schemes. Global GDP is estimated at US $70b, so a tax on the order of 0.7% would be sufficient to end poverty. Technology too might enable the implementation of new forms of within country taxation in the developing world, so that those living on $5/day could help those living on $1/day.


Private remittances (money sent back home to the family by immigrants) were expected to total US $406b in 2012. Numerically this is of the right magnitude to solve the problem, but the money needs to be going to the poor. In addition to helping the immigrant's family, immigration is of direct benefit to the immigrant. Migration can also contribute to trade based economic growth. Silicon Valley's Indian population played an important role in the emergence of India's IT industry, and a similar story can be told for the Chinese export industry. The possibilities for immigration to eliminate poverty warrant further study. Three options worthy of consideration are better targeted immigration (universities and governments implementing a preferential option for the poor when they make admission decisions), increased immigration, and a more radical opening up of borders (Nobody is Illegal, Welcome to Europe).

A note of caution: according to some economists remittances are negatively correlated with country level economic growth. They say this is due to moral hazards brought about by asymmetric information, but could it not also be due to brain-drain effects, or the money flowing in when the need is greatest. In any case the goal shouldn't be economic growth for the sake of economic growth, but improving human welfare, which migration and remittances presumably do.


The field of development economics is vast and distant. There must be answers to some of our questions there, but where to begin looking? If only it could be summarized by a few key bullet points? Understanding how China has managed to reduce extreme poverty from 84% to 16% from 1981 to 2005 and moderate poverty from 98% to 36% would appear key. Is it a population growth rate story? Or is it due to switching from a centrally planned to a market economy? If the later, why did poverty increase in the former Soviet Union under a similar transition? And once what happened is understood, how reproducible is the Chinese experience elsewhere? Encouragingly, the progress in China appears to have been driven primarily by policy rather than an infusion of money. Less encouragingly, one of the key policies appears to have been the de-collectivization of agricultural co-operatives, something that has limited applicability.


Cell phones are now ubiquitous in the developing world. What prospects do they or other technological advances hold for reducing poverty? In an increasingly information driven world, the Internet allows knowledge workers to be located anywhere, effectively doing an end run around border regimes (Samasource, Digital Divide Data).

Rather than increasing the income of the poor, another option would be to reduce the amount of money the poor need to live comfortably. Genetic engineering and high yield crops might be one approach. Another is scientific advances in the treatment of the diseases of poverty (OneWorld Health).

Other thoughts

Low hanging fruit

Global health makes it to the top of the effectiveness lists of both GWWC and GiveWell. Picking the low hanging fruit first makes a lot of sense, but we must not delude ourselves into believing that this can eliminate poverty. It can prevent the slide down the poverty ladder from moderate to extreme poverty or even death. But lack of access to healthcare is only one part of the story for someone living on $1/day.


If ODA is missing the mark today, the obvious thing to do is focus on getting ODA to target the things that matter (RESULTS/ACTION).


It seems possible to eliminate poverty. We have outlined something like half a dozen different plausible approaches. Some reasonably well funded; some barely funded at all; each with different uncertainties. Nearly all of them represent giving opportunities. It would be foolish to place all of our eggs in one basket. Rather it makes sense to make giving decisions so as to ensure a diversity of approaches are taken. The weights to assign each approach are currently unclear. How should one weight an independent set of uncertain probabilities to maximize the likelihood of success?


We live in exciting times: for the first time in recorded history the possibility of ending extreme and moderate poverty is clearly visible. In order to succeed in a reasonable time frame however we are going to need to think and give outside the box.