What does “Why Nations Fail” mean for International Aid?

Written by Neil McCulloch

In August, I spent part of my leave reading “Why Nations Fail” by Daron Acemoglu and James Robinson, two American political scientists and economic historians who have made a reputation over the last decade by publishing a series of papers suggesting that it is the quality of the institutions that a nation constructs that is at the core of long-run economic success. Their fascinating and well written book is a bold attempt to put forward a theory of how political institutions determine economic success (or failure) and why countries end up with the political institutions that they have. They do this by way of a colourful romp through the economic history of numerous countries including Australia, Bolivia, Botswana, Britain, China, Colombia, DRC, Ethiopia, Paraguay, Somalia, South Africa, the United States, and many more.

Their basic thesis is that both political and economic institutions can be either extractive or inclusive. Extractive political institutions are one where a tiny elite hold the power and put in place mechanisms to ensure that this continues to be the case e.g. by systematically disenfranchising others in society. Similarly, extractive economic institutions are those which ensure that the lions’s share of the country’s wealth and annual output is captured by elites, with the rest of the population subsisting on a much more meagre income. Historically, extractive political institutions give rise to extractive economic institutions, which in turn help elites to remain in power – a vicious circle. However, they argue that such countries have tended to have poorer long-run economic performance, because elite interests are threatened by, and therefore block, the creative destruction that is essential for innovation and productivity improvements.

Acemoglu and Robinson show that some countries managed to break out of this vicious cycle into a virtuous cycle of relatively inclusive political and economic institutions. Inclusive political institutions provide all (or at least most) with a voice in the political decisions; this makes it more difficult for elites to rig the rules in their own favour and therefore results in more inclusive economic institutions (e.g. concerning property rights and contract enforcement) that encourage innovation and investment. They argue that Great Britain was the first to make this leap by first enshrining greater political inclusiveness in the Glorious Revolution of 1688 which then placed it in a better position to exploit the opportunities created by the industrial revolution over the following two centuries.

Finally, they argue that switches from one regime type to another tend to occur at “critical junctures” e.g. major external “shocks” such as the Black Death, which killed perhaps half of the population of Europe in the 14th century; or the opening of the Atlantic sea routes in the 17th and 18th centuries. But the outcome of these transitions depends on the particular composition of institutions at that time (and the way in which these institutions have evolved or “drifted” over time): in some cases a transition from extractive to inclusive institutions can take place; but they document many others cases where an opportunity for change was lost e.g. when extractive colonial institutions were simply taken over, on independence, by a local elite but run in a very similar way.

The implications for international aid

If correct, Acemoglu and Robinson’s argument has major implications for aid.

First, much aid takes the form of technical assistance. The underlying assumption of such aid is that the problem is one of ignorance (or, more politely, “low capacity/skills”). However, Acemoglu and Robinson argue that this misses the point – the lack of capacity in a given area is often deliberate, reflecting systematic neglect of investment in that area because the area threatens the interests of the elite. For example, if the elite are avoiding tax, then the last thing they want is a clean and efficient tax office; if they are politically powerful they can ensure that the tax office remains weak. No amount of external TA will solve this problem because the fundamental issue is one of political interests, not technical capacity.

Second, if the fundamental problems are political, then, they suggest, it is extremely difficult for external actors to influence them. Policy or political conditionality will be unsuccessful – most governments are unlikely to implement changes that challenge entrenched interests simply to access donor funds. Even if external actors have significant influence, changes are likely to be superficial and unsustainable.

But are Acemoglu and Robinson right? 

I think that their historical and political analysis is, for the most part, spot on – but their chapter on aid is the weakest part of the book showing a disappointingly simplistic understanding of what aid agencies actually do. As a result, they draw far too simplistic conclusions for aid policy.

What A&R get right

A&R’s description of Aid is right in two important ways. First, they point out that domestic politics is the key determinant of a nation’s long-run progress. There is sometimes a tendency in the development community to assume that external factors (international commodity prices, availability of capital etc) are the critical determinants of a country’s performance. Whilst these factors may be important, A&R (and many others before them) make a compelling case that it is how a country’s politics constructs its institutions that determine its long-run success.

Given this, it is surprising how much aid is focussed on providing technocratic solutions to capacity problems. Much of aid is seen as matching “skills gaps” in the receiving country with the skills available in the aid providing country. Sometimes we fail to ask why things are the way that they are i.e. to uncover the politics that gave rise to underinvestment in a particular area. More broadly, development partners as a whole put much less effort into the analysis of the politics of the countries in which they operate than the technical aspects of the development problem.  Such analysis, when done, is often confined to assessments of government “demand” and possible political “risks”, rather than a thorough political economy analysis of who the actors are, what are their interests and how this has shaped the design and evolution of the relevant institutions.

Moreover, the neglect of political analysis is often deliberate, since, to be acceptable to the recipient country, development aid cannot be seen to be “political”. Of course, this doesn’t make sense – understanding the political economy can help programs to be more effective even if they are in no sense “political”. Part of the reticence to undertake such analysis also reflects different cultures between aid agencies, who have a predominantly apolitical technocratic mandate to “save lives” and Foreign Affairs officials who do value political analysis, but regard it as a tool of international relations rather than of development.

The other area where I believe A&R are right, is their focus on exclusion, as well as poverty reduction. The mandate of AusAID and other aid agencies is, rightly, to reduce poverty. However, the depoliticisation of the development narrative over the last 20 years has pushed issues of inequality and justice to the periphery of much aid. Yet development is not just about improving the living standards of the poor, but also about enhancing rights, reducing inequalities in opportunity, and promoting what A&R call “inclusive institutions”. In this AusAID is somewhat ahead of the game compared to some other donors, with its strong focus on the inclusion of disenfranchised groups e.g. women and the disabled.

What A&R get wrong

Whilst I believe we can learn a lot from A&R’s core argument, there are a number of areas where I think they get it wrong.

First, they dismiss programs that attempt to achieve institutional reform within individual sectors, providing an example from India where higher level politics undermined an attempt to put in place sensible incentives for good service delivery.  Whilst there are surely many examples of programs failing because of the wider political environment, broader political and institutional reform is not always best achieved by a frontal attack on the vested interests of existing elites. Indeed, institutional reforms are often achieved through an incremental process of small victories on specific issues and in specific sectors. One has to start somewhere, and providing support for local organisations pressing for specific inclusive institutional changes in their area or sector is often a good way to catalyse broader changes.

Second, A&R dismiss technical assistance on the basis that it assumes that “ignorance” is the problem. But doing so makes two errors. First, it ignores the important role of technical assistance to support the ability of key reformers, both in government and outside, to press for institutional reforms. Reform champions often have few staff with the necessary skills and training to achieve the reforms that they want to introduce; technical assistance can play a key catalytic role in getting reforms off the ground in such a context. Second, dismissing technical assistance assumes that ignorance is never the problem. Sometimes it is. That is, there are (many) development issues where there is no conspiracy to benefit an elite by keeping things bad for everyone else. Rather, there is an enthusiasm to adopt the relevant knowledge and technology; in such circumstances technical assistance can accelerate the adoption and absorption of more effective ways of doing things.

Finally, A&R make the argument that all societies with extractive political institutions will ultimately perform poorly economically, even if they have a short period of economic growth, because such institutions prevent the creative destruction that drives long-run growth. In particular, they warn against the “Beijing consensus” arguing that China’s economic performance will eventually come unstuck because of its extractive political institutions. Unfortunately, their argument here sounds more like ideology than evidence. One can think of numerous examples of lengthy periods of growth in countries in which the political institutions have been dominated by elites (not only China which has grown at breakneck speed for more than 3 decades, but also Vietnam, South Korea, Singapore, and, of course, Indonesia). While there is clear evidence than creating inclusive economic institutions is necessary for growth, the evidence that inclusive political institutions are essential to achieve this is, sadly, far less compelling. From a growth perspective, it would appear to be sufficient to have contestability between competing elites.

This blog is a part of a series on ‘Why Nations Fail.’ For other blogs in the series, see here.

Neil McCulloch is the Lead Country Economist for AusAID in Indonesia.  The views above have been expressed in a personal capacity and do not necessarily represent the position of AusAID.

image_pdfDownload pdf

Neil McCulloch

Neil McCulloch is the Lead Country Economist for AusAID in Indonesia.


  • Interesting post, and on the whole I’m inclined to agree with you. I have yet to read A&R’s book. However, I note that your counter-arguments (1) and (2) do not come equipped with supporting evidence. I think that both such arguments could easily be over-played. E.g. if challenged I suspect many an aid bureaucrat could claim that (1) and/or (2) applied to every single one of the programmes s/he funds.

    To conclude: it seems to me that you have defended some aid, whilst leaving probably the majority as unlikely to lead to significant economic development. A better defence is that a lot of aid will not generate economic development but, done properly, can at least make life substantially less miserable to millions of people who live below the poverty line. Alas I think too much aid is not ‘done properly’ but that is another story …

    • “A better defence is that a lot of aid will not generate economic development but, done properly, can at least make life substantially less miserable to millions of people who live below the poverty line”…That’s exactly what elites in extractive political insitutions want aid to provide…for political stability so they can continue to extract wealth and power while aid agencies continue to maintain political stability in their country. Aid therefore can be viewed as keeping these institutions in power.

  • I’d like to support your statement (on last third para) on small victories. Sometime when we feel that no way to win ‘big battle’, small victories will be a perfect strategy. Unfortunately it is very often that we have lost when we busy fighting the ‘small battle’. We don’t realize that the ‘small battles’ that we fight were not lead to the big one. Good understanding on political environment, broader political and institutional reform therefore will help us not to lost in the ‘small victories’ strategy.

  • I have enjoyed reading this set of interesting and constructive reviews. Thank you, Neil.

    I thought your point about our failure to ask why things are the way that they are, and to put sufficient effort into the analysis of the politics of countries receiving assistance, compared to the technical aspects of the development problem alone, especially important. This point about analysis echoes the views of Gerard Guthrie in his book ‘The Progressive Education Fallacy in Developing Countries’ (Springer, 2011) that evident failures in education in development likely reflect insufficient analysis of local cultures and their impact on learning and teaching.

    The case for better analysis in public policy formation has also been very well-made by Gary Bank’s, Chairman, Productivity Commission, in his 2009 address to the Australia New Zealand School of Government in February 2009 available at: http://www.pc.gov.au/speeches/cs20090204

Leave a Comment