Horizon 2025: the future of aid (and a potentially epic nerdwar on poverty numbers)

I’ve been starting to feel like an unpaid publicist for the Overseas Development Institute recently. It’s not my fault – ODI keeps publishing horizon 2025 coverreally interesting stuff (and anyway, I’m not always nice about it). You’re likely to hear a lot about their latest paper, Horizon 2025: Creative Destruction in the Aid Industry, by Homi Kharas (Brookings Institution) and Andrew Rogerson (ODI), so I urge you to read it, rather than just winging it based on this blog (not that you would dream of it, of course).

The paper first takes the ‘MICs v LICs’ discussion on where poor people actually live, which Andy Sumner has been leading for the last couple of years, and projects it forward to 2025. Futures work usually offers several scenarios because the future is so difficult to predict, but Horizon 2025 opts for a single ‘base level 2025 scenario’, extrapolating recent trends as follows:

• High per capita income growth and falling population growth in large, dynamic, middle income countries (MICs) shrink the global poverty pool drastically.
• Income stagnation and high fertility rates in selected low-income and fragile countries re-establish them as the main locations of global poverty.
• Growth in emerging economies dominates global growth and they account for most new trade and foreign capital flows to poor countries, along with sizable increases in aid-like flows, competing for influence with traditional aid donors.
• Availability of public and private resources for development, coupled with the fall in global poverty, imply that dramatically more funding is potentially available for each poor person.

globalpov fragile v stableThis is all good stuff but is arguably skating on fairly thin ice as the ‘baseline’ is at the optimistic end of possible scenarios. It uses IMF growth projections and average expenditure per person from GDP data (known as the national accounts), rather than the now burgeoning amount of household poverty survey data, and assumes that inequality remains constant as countries grow (when it often rises). That means that the 2025 baseline effectively maximises per capita income and minimises poverty, compared to other estimates such as that by Karver, Kenny and Sumner for CGD (here) and Andy for IDS (here).

Andy’s now ubiquitous killer fact is that three quarters of the world’s poor – well, 79% or 2 billion poor people (by $2 poverty) – now live in MICs. You hear this quoted everywhere, with the implication that internal politics and distributive struggles (rather than aid) hold the key to future development. Kharas and Rogerson argue that the MIC thing is actually a blip (OK, a twenty year blip, but still). By 2025, growth in most MICs will have lifted almost all their poor citizens above the $2 a day poverty line (I said they were optimists). Instead, the vast majority of the world’s remaining poor will live in FRACAS – fragile and conflict affected states (see chart 1): the long-term dividing line will between between fragile v stable states, rather than LICs v MICs, which chimes nicely with my own work on effective states. They will also largely live in Africa, which should please Paul Collier, as it endorses his Africanist ‘Bottom Billion’ thesis.

Using a different source of data – household poverty surveys and multiple scenarios – Andy Sumner finds a quite different pattern. Across a range of scenarios, there is a 50/50 LIC/MIC split in poverty in 2020 and 2030. Given that some of today’s LICs will be MICs by then, it’s possible that just one-third of the world’s poor will be in the remaining LICs. Meanwhile, estimates for total global $2-a-day poverty by Andy’s method in his paper with Charles Kenny and Jonathan Karver at CGD (see here) run from as low as 600m to as high as 1.6bn (there’s also multiple scenario data for other poverty measures such as malnutrition there too).

All in all, I’d say that Horizon 2025 is not the last word: there’s plenty of room for doubt and arcane methodological debate over the numbers, with significant consequences for the way we think about the future of development. I see a nerd war approaching – wonderful!
But whatever the final outcome of that war, it seems to me that Horizon 2025 is really onto something. In terms of reducing poverty, the focus in most MICs will be domestic policy, not aid. Donors and INGOs may play a minor supporting role in those domestic struggles, but the core aid business is likely to retreat to fragile states, (where, incidentally, we will witness a titanic clash between the chaotic, unpredictable nature of fragile states and the increasing demands for measurable, attributable impact).

Where Andy and Kharas and Rogerson do agree is on the affordability of ending poverty. Andy provides a set of estimates that suggests Horizon 2025 chart 2the costs of ending poverty might well be affordable (as a percentage of GDP) for most MICs by 2020 and certainly by 2030, leaving just 16-28 LICs in 2030 needing foreign aid to end poverty.

Kharas and Rogerson calculate the costs of eradicating poverty in a similar rump of fragile states, assuming a certain amount comes from domestic sources and the rest from aid. Their argument is that by 2025, mobile phone-based banking will be ubiquitous so it really will be possible to deliver cash direct to the poor, however dysfunctional the state. As poverty falls and GDP grows, a tech-based $2-a-day safety net becomes increasingly affordable, accounting for just a fraction of 1% of global GDP (see chart 2).

Next, the authors think through a further set of issues on the future role of aid. They look at three intriguing and believable ‘disruptors’ to the traditional aid system:

1. New channels for aid via philanthropy or ‘disintermediated’ direct giving
2. The rise in South-South cooperation, and with it, the blurring of distinctions between aid and cooperation based on mutual interest
3. Climate change and the rise of climate finance (both for mitigation and adaptation) either in parallel to, or as a competitor to, poverty-focused aid

Going into any detail would be too much for an already overl-long blogpost (you really should read the paper), but they run a stress test against existing aid agencies to see who is best/worst prepared for this 2025 world. On the resulting ‘traffic light’ bar chart (below), red is the worst prepared. Bad news for Spain, but Canada (for once) is deemed to have got it right, and they’ll be dancing in the corridors of GAVI and the Global Fund.

I suspect this may not be the last post on this.

Horizon 2025 chart 3

Subscribe to our Newsletter

You can unsubscribe at any time by clicking the link in the footer of our emails. For information about our privacy practices, please see our Privacy Policy.

We use MailChimp as our marketing platform. By subscribing, you acknowledge that your information will be transferred to MailChimp for processing. Learn more about MailChimp's privacy practices here.


4 Responses to “Horizon 2025: the future of aid (and a potentially epic nerdwar on poverty numbers)”
  1. Heidi Monk

    Hi Duncan,

    Actually I quite appreciated both your post and the article.

    I’m concerned with the ODI paper’s understanding of development only being about poverty and poverty only being about money (therefore development is only about money). It is the subject of the paper and make only a cursory acknowledgement that there might be a bit more to the picture. (On a side note, if they are going to spend so much energy talking about the quantity of fiancial resources, they should probably save some breath for their quality. I thought we’d established that quality of the expenditure matters quite a bit… ie Cuba, Kerala, etc).

    I’d also like to hear your thoughts on the effect of going down the CCT or basic income route. Is it a way out of public provision of services? Does that mean tiered services? The more money you have the better your services? (I realise that this is the starting point, but the CCT route doesnt seem to address equality issues). This was a slight tension in the paper: MIC governments will be doing more to address development, and CCT/direct income transfers will increase. Do these not pull in different directions?



  2. Charles kenny

    Duncan–tnx for shout out. Any estimate of 2030 poverty including ours depends hugely on growth estimates for a few big countries and growth estimates 20 years out are incredibly unreliable so I’d take all of them with oceans of salt. Also, obsessing too much over two more or less arbitrary lines ($2pov and low mid income) and who is just below or just above them can be taken to far. Low middle income countries and people living on $2.50 a day are still poor…

  3. There is something disconnected from reality about this paper by my friend Homi Kharas and his co-authors: the peculiar manner in which they consider (or don’t) the impact of climate change. The impact of a rapidly destabilizing climate on economic growth is not included in the upbeat growth scenario through 2025. However, the impact of climate concerns on the aid system is considered in some detail. The authors write:

    “Failure [to avert rapid climate change] could prove catastrophic in many ways, including
    unjust and unbearable further burdens of adaptation on the
    weakest countries, with their negative externalities such as
    forced migration and increased cross-border conflicts over
    resources. In such a dead-end scenario, one minor corollary
    will be the massive diversion of national effort and external
    support toward spiralling needs for damage containment in
    such country contexts. This fact alone would rather quickly
    force development aid to refocus on disaster management
    and resilience programmes, to the exclusion of many of the
    priorities outlined elsewhere in this paper.”

    Well, yeah, but hello?? the continued rapid economic growth in the MICs that the authors are counting on to lift hundreds of millions of people out of poverty between now an 2025 also seems mighty unlikely, given that all indications are that runaway climate change HAS ALREADY BEGUN.

    Thinking about the impact of climate disaster on aid flows without also considering more carefully the likely impact on growth itself seems to miss the (burning) forest for the (beetle-infested) trees.

  4. Dr Green writes,

    “…there’s plenty of room for doubt and arcane methodological debate over the numbers, with significant consequences for the way we think about the future of development. I see a nerd war approaching – wonderful!”

    These views are perhaps part of the problem for the world’s poor.

    It is not “arcane” to point out that the numbers are on spending, not poverty.

    It is not engaging in a “nerd war” to point out that the statistics are widely abused, by the World Bank, by economists, by charities and by politicians.

    They are abused in all assertions and implications that they show lack among, or benefit to, real people. In reality there is zero assessment of either specific prices faced by the poor or need.

    You don’t have to be a “nerd” to see the risks in ignoring food needs, rent needs, medical needs and so on when assessing progress or policies.

    Which development experts can tell the difference between important and unimportant problems in the use of statistics to guide public policy?

    Everyone agrees that poverty is lack. It isn’t the level of resources whatever the needs. So resources – let alone the spending numbers – don’t measure poverty.

    Pointing that out is not just contributing to a “methodological debate”, or an argument about “definitions of poverty”. It is to state categorically that what the public is being told is not what is actually compiled.

    If they were poverty numbers, or “absolute poverty numbers” or “how the world’s poorest fared in the 1990s” or “consumption” in the ordinary sense of that word, or “consumption poverty” then there would have been some method for inferring from “spent more” to “got more in return” to “met needs more adequately” and perhaps to “fared better economically overall”.

    There is no such methodology.

    Meanwhile, governments act on the basis that they have data on which policies help the poor, when there is an obvious risk of skewing, and policies being adopted which harm people and/or cause unsustainable economic development through fundamentally unproductive transactions, by ignoring need (and assets and debts).

    As can be seen in other comments on this blog, Dr Ravallion of the World Bank has failed to engage in a “methodological debate”.

    He has failed to explain how he knows, for example, that there is not a significantly increased need for spending on rent in some countries.

    He doesn’t have a methodology for inferring global poverty.

    He has spending numbers (with imputed monetary value of own consumption and as alternates to spending, in some cases numbers on income).

    He can’t say how he gets from “spent more” to “filled needs better” because he has not looked at needs.

    And if the World Bank which compiles the data don’t have a methodology for inferring global poverty, or poverty comparisons for policy purposes, neither does anyone else.

    Is it hard to see how this is risky for the poor?

Leave a Reply

Your e-mail address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.