I am prepared to abandon these beliefs at the first sign of trouble.

In light of a recent shift into my late twenties and the arrival of the new year, I felt it would be reasonable to write down some of the relevant things that I have to come to think I believe. These beliefs are not necessarily backed by hard, empirical evidence and I may be prepared to abandon many of them in the future. Still, it might be useful to clarify some of these thoughts, as many will seem quite obvious to frequent readers, while others will appear counter-intuitive.
 
1. There are two, largely divergent views on the ultimate purpose of aid which are more difficult to reconcile than most people realise.

One of these is `aid as a safety net’ or `aid as an immediate welfare-improver’ – this is aid intended to make life a little better for poor people in poor countries, with little scope for improving their lot outside the immediate effects of the aid itself. The other view is that aid has a potential for being a `structural transformer’ or a `growth-enhancer’, kicking trapped countries into a virtuous development cycle. Owen Barder has done a good job of delineating them on his blog here and here.

The more immediate, short-term welfarist view accommodates aid that is focused on project-based, small-scale interventions and the careful measurement of impacts. The transformational view lends itself towards the development of systems, institutions and possibly the more diffuse, diagnostic-based approach to policy. These drastically different goals are often conflated, usually due to assumptions that fail to bridge the micro-macro paradox (i.e., bednets have been proven to improve health at an individual level, so bednets will obviously lead to more development).

It is possible that the goals do not only contrast, but inherently undermine each other. Even so, there needs to be more dialogue about how they might be reconciled in future
 
2. Even after years and years of debate, many proponents and practitioners do not have a clear idea of what metrics of successful development/aid matter, nor the trade-offs between them.

The discussion over what matters rages on: do we measure income or capabilities? Happiness or educational attainment? These basic arguments seem to have gone on without too much concern for the intellectual bases of the metrics in question. A popular policy paper might, for example, lead to a craze of using `happiness’ as a new way of measuring welfare outcomes, without reasonable discussion about what this metric really measures. Understanding these foundations is important, because it allows us to be more precise about what we want, how we measure it, and how different policies might affect it.

Furthermore, we should frequently question why we care about our chosen metrics and be wary of focusing too much on proxies for our indicator of choice. Proponents of specific causes have difficulty seeing beyond the frontier of their personal set of metrics. Studies have shown that deworming is an extremely cost-effective way of getting children to attend school, leading to a big push for scaling up such projects. Yet we have little knowledge about the eventually effectiveness on income, happiness, or other measures of welfare that we might actually care more about than attendance.

A lack of precision and certainty about metrics suggests a lack of precision and certainty about goals – at which point the discussion gets even more murky than it already is.
 
3. The concepts of country ownership, autonomy and responsibility have never been popular with donors and continue to suffer, despite the resolve fo the Paris Declaration.

On my most cynical days, I feel that most donors and NGOs would be much, much happier if recipient governments vanished in a puff of smoke, leaving them free to administer to the poor as they see fit. Even when my skepticism is diminished, it is still apparent that the givers and proponents of aid believe that donors are, by default, more aligned to the needs of the poor than their own governments.

This behaviour is visible out in every facet of the West’s approach to aid: global goals like the MDGs, which do not directly stem from the desires and aspirations of recipient citizens, inherently stifle the natural process of planning and goal-setting in recipient countries. Top-down programmes, like those popular in global health, create vertical funds which make it more difficult for local decision-makers to make appropriate trade-offs in policy and resources.

It is true that not many governments in developing countries are good, efficient, capable, or benevolent – and it is likekly that too many resources have been wasted on grand-scale government aid. Yet if we never start treating recipient governments like `adults’, able to manage their own resources, to make their own plans and the same trade-offs that our own institutions make every day, they will see little incentive to act the part.
 
4. Despite rhetoric about enhancing civil society’s role, many donors still see themselves as the key players in holding developing governments to account.

There is a fuzzy, muddled contract between citizens and functioning governments – a feedback mechanism which, at the very least, can keep the state concerned about some minimum level of development. The elements of this contract comprise some basic elements: taxation, some method of selecting out awful governments and some degree of outcome transparency. This contract isn’t always completely democratic – in which case the feedback may not be entirely equitable (lobbying, for instance, in the US), but it still pretty essential in keeping governments from being entirely incompetent or malevolent.

Resources drive this relationship – the government depends on its citizens for survival. This is why large deposits of natural resources are often seen as a curse, as they allow governments to exist without a reasonable mandate. Aid also has the potential to act as a `resource curse’ , an outcome discussed in detail in a paper by Todd Moss and Arvind Subramanian a few years ago.

I believe most donors recognise this problem, but rather than thinking of innovative ways to strengthen the natural chain of accountability between citizen and state, they are more concerned with micromanaging recipient government policy and behaviour. Rather than being compelled to make improvements in education because they want to appease a testy electorate, recipient governments are compelled to make improvements because Donor X wants them to. The outcomes may be similar in the short term, but in the long term recipient governments are learning to follow the wrong carrot and be hit by the wrong stick.
 
5. The aid/development blogging community is becoming less apt at meaningful debate than it is at self-congratulatory criticism of outsiders.

While social media developments like Twitter have brought us closer as a community, such proximity has done little to foster new and interesting ideas. With some exceptions, 2010 was a rehash of old ideas, with much blogging (including this blog) dedicated harping on those who lacked appreciation of ideas we found sacred. We spent very little time testing those ideas and bouncing them off of each other, instead joining each other in various wars on outside laymen. While I wouldn’t expect to see the same cutthroat back-and-forth visible in strictly academic circles, I do believe you can find more heat on the Star Wars message boards than you can in the average development blogging debate.

We should aspire to be islands of critical thought which more likely to spar with each other than slap each other on the back. Part of the joy of blogging is sharing new, interesting ideas, then viciously tearing them apart.
 
6. Randomised control trials are one of the most important tools we have in developing coherent, effective policy, yet we must be careful to treat their conclusions with the same careful, critical eye all research deserves.

I work on an RCT – I believe in them as a gold standard for identifying treatment effects, but I am wary of those who believe that a couple of successful, short-term RCTs constitute enough externally-valid evidence to feed into general policy prescriptions. This is no different to how I feel about evidence using less robust identification strategies.

Part of the answer?  Replicate, scale up, replicate, scale up, all while continuing to assess impacts. Institutions like the IPA are dedicated to this necessary approach.

We also need to recognise that there are going to be some questions that are more difficult or impossible to answer with RCTs (we are probably not going to be able to assess the effectiveness of cash-on-delivery aid with RCTs, for example). Yet, being out of the reach of the gold standard does not mean that these questions should be ignored. The compelling evidence that rigorous impact evaluations reveal should become another tool in the policy-maker’s toolbelt, but should not be regarded as the ultimate tool.