Blog

Context: There has been a move away from big cross-sectional studies – they can’t tell you why countries do things and how it works out by looking across contexts; we need to do something more specific. But that then became that you have to spend eight years studying a country before you can say anything.
We pitched ourselves in the middle ground, saying that context matters, but it actually matters in quite similar ways, depending on the configuration of power – different political settlements. We wanted to know how different kinds of settlement deal with different development puzzles, such as avoiding the resource curse, generating growth, delivering services or protecting women’s rights.
We found that we could identify different trajectories of development in different configurations of power. Take oil governance in Uganda v Ghana. They both found oil about the same time, in about the same amounts. Ghana rushed to market on deals that were questionable, whereas the more autocratic Uganda played a longer game and got a better deal.
But we followed up on that and looked at what happened next. There, Ghana way outperforms Uganda – money is coming in, even from worse deals, and going out on public services. In Uganda the oil is still in the ground. The president is trying to drive harder deals, and wants a refinery as well. This seems to be more about ideas than interests. The president would have made more money for him and his cronies if he’d gone to market, but he held back. That’s about his vision for the country.
Ideas Matter: Political science has had a blind spot on ideas. Political settlement analysis tends to imagine that every ruling elite is just interested in its material interests and maintaining themselves in power. But actually we find that ideas matter – on nationhood, who belongs, redistribution, the relationship between state and capital, but also at the level of solutions. Elites only buy into external policy ideas on things like domestic violence or cash transfers when they solve a pressing political problem, an existential crisis, or they fit their ideological frame.
Ideas interplay with interests. But it’s very rare that ideas trump the politics of survival.
Civil Society: It is prominent in some areas of our research, for example on tackling domestic violence or in public service provision in urban India. But it is heavily absent in other areas, such as the rapid recent spread of social protection. There it was donors building coalitions with bureaucrats, and winning over the more powerful blockers. I think civil society will become more important at the next stage of policy implementation – for example, we are seeing constituencies of older people developing around social pensions. And some unusual forms, like traditional authorities supporting local teachers on education reform. I think public authority is a useful term there.

 

Capacity: We found the state is still the only institution that can manage globalisation, protect women’s rights, deliver services at scale, build political order. It wasn’t an easy message, given the current backsliding around democracy. But we found that the international agenda had become much more obsessed with making states transparent and holding them to account, without making sure they were capable of doing the things that they could then be held accountable for!
Coalitions: In conditions where rules don’t play out as expected, you need fixers and coalitions. They can be narrow – e.g. politicians and bureaucrats agreeing to protect pockets of effectiveness, or much broader ones with private firms: there are both good and bad forms of cronyism. The coalitions have to be based on credible deals, built on mutual reciprocity. They need to need each other.
But although early growth take-off deals can be based on politically cronyistic deals, these then block the next round of capitalists that need to come up and make growth more dynamic.
Pockets of Effectiveness: We’ve shown that you can have high levels of public performance in the two different kinds of political settlement (competitive and dominant), but through different routes. Rodrik showed us the big macro data that dominant systems have bigger booms, but also bigger busts, whereas competitive systems produce a smoother ride. What we’ve added is much more on the causal mechanisms for our much smaller data set.
Sequencing: We’re asking countries in Sub-Saharan Africa to become both high capacity states and highly democratic at the same time, and it’s just really difficult – that’s not how it happened in England or the US. What we have at the moment is donors doubling down on democracy, in part due to the democratic backsliding and rise of authoritarian populism. Liberal OECD states are doubling down, saying ‘we have to talk about inclusion louder than ever now, and it has to be democracy first and foremost’. Another group says it’s all about order, and we’ll back that, even if it comes at the expense of democratic rights.
We don’t think either response is correct, either ethically or on the basis of the evidence. We’re trying to map a kind of third way that says that state capacity really matters, and that can be part of a progressive agenda. A lot of evidence points towards sequencing, but it sticks in the throat to go to countries that are being repressed and say that. We don’t find compelling evidence that repression is required for development.
Some of our work shows the ways you can build capacity and accountability at the same time, especially at the local level. You find parts of Ghana and Uganda delivering high quality education or healthcare, through local-level coalitions.
My most exciting/inspirational result of nine years of progress? The recognition that we now have a brilliant generation of scholars that are going to keep talking about politics. A remarkable set of authors from the Global South, Ghana, Bangladesh, India, Uganda. They will keep on talking about politics and how it shapes development. That’s my favourite takeaway from all of this.