Political research is good for you (in the end)

Image

This blog was born as a twitter thread in 2020, was updated as a blog in 2024, and here it is once again in 2026. 

I’ve been commissioning ‘applied’ political analysis and research in big, mainstream development sectors since the early 2000s. 

By ‘applied’, I mean practical, actionable, and in plain language. The label can vary – political research, political science, political economy, political economy analysis (PEA), political economy research (PER), political analysis. I’m no purist. ‘Research to understand and tackle political constraints to reform, development, and growth’ is probably most accurate.

Most of the time, I end up preaching the benefits of political research to the already converted. This piece is intended for a different audience: those in the ‘big sectors’ of health, growth, education, livelihoods, and climate who remain sceptical of PEA.

Way back when….

The first proper political study I commissioned examined a proposed investment to mitigate arsenic contamination in water supplies. This was already a draft business case with a well-argued technical case. 

But the political study concluded that the accompanying institutional reforms were unrealistic. They relied on traditionally dominant (big spending) forces in government becoming subservient to a small new unit. My study delivered an unpopular message, as it undermined months of work, and I annoyed people by commissioning it. But I knew what I was doing. The study formalised concerns that were already whispered but were being ignored.

This is not at all to diminish the scale of the problem or the need for action. Twenty years later, arsenic is still in the news as a serious threat to health. But the proposed strategy did not look politically realistic, and so risked being far less effective than promised. And also therefore a waste of precious aid funds. 

Basic rules

This study, and other early work by ‘political economy’ consultants, examined knotty institutional and political constraints in big spending sectors. My ground rules for commissioning such studies were simple:

1. Have a clear analytical framework and method (i.e. the work should not be a ‘freestyle’ political description). There were a few frameworks to choose from. There are more now, and they are not hard to find.
2. The team should include a bona fide political expert who has previous experience doing such work. It should not be composed solely of sector experts moonlighting as political experts.

This second rule can irk people. When we selected a consultant or political researcher, our sector experts often complained: ‘she doesn’t know anything about this (our!) sector!’ They’d suggest a well-known sector name instead, often a retired bureaucrat from the same department. But our governance gut feeling was that fresh eyes and strong political analytical skills were more important than deep sector expertise, and the political experts would in any case draw on sector specialists as they worked. Sometimes we combined both in a small team – but for us the inclusion of a political expert was non-negotiable.

Even today, plenty of ‘political economy’ studies break both of these rules. The resulting reports don’t tend to be very good.

Political weather versus political climate

It was also important to go deeper and more structural than the current political ‘weather’ of a country, sector, or ministry, which is often made by the personalities, rumours, and ‘noise’ that preoccupy corridor chat. This meant getting into the long-term, political economy ‘climate’ – the underlying distribution of power and interests.

Such studies ruffled feathers – if they didn’t, they were probably pointless. But they could have positive effects on plans, such as when they advised that certain technical or institutional reforms were politically unfeasible (e.g. required turkeys to be pretty chilled about the Christmas menu). Intervening to arrest failure and stop money being wasted is a good thing. Intervening to make strategies more likely to succeed, and investment more effective, is even better.

A secret service….

But the outputs – studies, policy advice – were rarely made public. They helped those who commissioned them feel and perhaps act smarter, but they were private, uncontested, and did not spur open debate. To me this was a wasted opportunity. How could such short-term, private gains address long-term, collective pain if other partners ploughed on regardless?

I’m a ‘public policy open debate’ person. I think this should be enabled by open data, published research, and public opinion at all levels. Civil society, researchers, and governments should talk about and contest domestic policy in their own countries, as should international agencies and their government partners.

Big R Research - love it, hate it, debate it

In time I moved to commissioning ‘Big R Research’ – that which is put out by universities and think tanks – including practical ‘political economy’ research on growth, health, education, urban development, and sub-sectors in between. Big R Research is by nature published. Love it or hate it, at least everyone can debate it.

Such political economy research is typically commissioned through competition, so there is less tendency for the funder to only work with their long-favoured experts. It is also quality assured via peer review throughout the research process. Researchers ultimately perform in public and are assessed by their peers. Reputations depend on this. 

This feels procedurally better to me, but practically it falls short if the only people who end up reading the work are other political researchers. Academic research can be an inward-facing, expert niche, in part because the publish-or-perish incentives of academia generally favour the novel over the useful. As a result, many political researchers struggle to communicate in ways that sector players (health, education, growth, skills, climate) can understand, or act on (that is why I say ‘applied’). 

But the best research can have big, real-world impact. And it does not cost much, particularly when compared with the sector budgets at stake – regardless of whether they are funded through taxation, borrowing, aid, or philanthropy. 

As an example, see this compilation of OECD aid data (up to 2022) graphed by the good people at One Data. Look at the budget at stake in health!

The (potential) RoI of PEA and PER (research)

My experience is that the cost of commissioning such practical political research is a tiny fraction of the sector programme investments on which it typically focuses. In theory, if you can identify binding political constraints, and work out how to fully or partially overcome them, you should be able to unlock greater effectiveness in that sector.

This could be measured. And many potential customers/funders rightly ask for some credible estimates of input/output, or return on investment (RoI) of PEA. Unfortunately, such data is hard to come by. This is in part because practical political economy research remains peripheral. Few if any big sector players have ever said: ‘come and help fix us - and measure by how much’.

It’s also because the expectations of rigor in causal claims in development are high and getting ever higher. It is, for example, extremely hard to offer counterfactuals for activity undertaken without ‘practical political economy’ research and action, to accompany and compare with those that have it. (I asked a professor of political economy for advice on how to start looking for evidence of effectiveness and estimates of RoI, and they said, ‘Now that is a really good question!’. It is hard!).

The shortage of clear RoIs or proof of effectiveness for ‘practical political economy’ is too often used as a reason to steer clear of it. But this confuses a gap for a trap. The lack of data should lead decision-makers to the conclusion that PEA needs funding and testing, not to looking for something that doesn’t yet exist and triumphantly shouting ‘aha!’ when they don’t find it. Plenty of other punts and fads in development get funded and then tested – PEA should be no exception.

In lieu of quantitative data, here’s a worked narrative example of what the RoI of PEA could look like.

In 2022, a political economy research project identified serious political and corruption constraints in a sector with over $1bn multilateral investment – private sector skills training for garment sector factories.

The sector investment strategy included performance-related payments for private sector training organisations, which would get paid a final one third of their fee when a trainee graduated and secured new employment in a factory.

Breaking it: unfortunately the political economy research found that the system was being gamed with fake verification of new employment. This undermined the effort to actually develop better skills as the training organisations would get paid regardless. The research estimated that between 10% and 30% of the investment might be wasted in this way. Research exposes a failure – so ‘breaks it’.

Making it: the research also suggested a better strategy for ensuring that real skills were gained. Instead of working through training organisations, the programme should target hiring companies that had operations sophisticated enough to use greater numbers of more skilled workers. This would expand their business and profit, so they’d have an incentive to care about the quality of trainees. Research identifies an alternative strategy - so ‘makes it’!

The research cost £130,000 – less than 0.02% of the total external sector investment. So even if it increased overall effectiveness by a small percentage – such as 1%, or even just reduced losses by 1% – the returns and/or savings would be significant. 

If the research had not happened, would anyone have noticed or publicised the problems? Unlikely. Does this research look worth it? I think yes (but sadly the £130,000 came from ‘my’ small governance research budget in a government department, rather than the multilaterals’ sector budgets for skills investment).

There is still too little practical political research happening, with large areas of policy and development spending untroubled by it. The supply of good research is constrained, and demand from sectors, while growing, is still low. Both need attention.

Opening up to political (economy) research: the 5 stages of grief

Political research can often seem to be an awkward optional extra, pushed by peripheral governance people. Looking back, commissioning new political research in big sectors in some ways echoes Kübler-Ross’s ‘5 stages of grief’: denial; anger; bargaining; depression; and finally acceptance.

A typical case starts with us annoying political types suggesting that we commission political research in key sector ‘X’, because years of investment have not yet delivered change at the desired pace or scale (or because we get a whiff of something more sinister going on). The sector experts (who control resources and investment decisions) will often respond by pointing to our old friend ‘Lack of Political Will’. But as Heather Marquette has often said, it may be more useful to think of active political won’t.

And so begin the five stages of grief.

1. Denial

Sector technocrats may say: ‘Political research? No time for that / we don’t need it.
We are experts – it is REALLY technically complicated. I am sure that someone is doing this already. No, I can’t name a particular study. Or any researchers’.

(This may seem artificial, but I had this exact conversation when proposing new research on the politics of massive scale tree planting schemes – the kind that, if successful, would be visible from space. My hunch was that this is a big ‘if’).

Despite their protests, we commission political research by people who may not have deep knowledge of that sector, but are experienced, practical political researchers. They ask some awkward questions about areas/processes/relationships that are critical, but stuck. They ask about power and ‘small p politics’.

(For tree planting, such questions included: why do politicians keep announcing big new schemes, when media reports show the largescale failure of existing schemes due to a lack of investment and attention to detail?)

2. Anger

There may be annoyance that the political research exposes over-simplistic technical approaches that are expensive to implement, directly contradicted by real political incentives, and have been unsuccessful in the past (plant trees – again!).

There may also be complaints that political research is better at ‘breaking’ (pointing out all of the problems) than ‘making’ (offering a menu of politically viable options to work around, overcome, or reduce the constraints). This is often justified. Some political research is glibly destructive, rather than practical and action oriented.

3. Bargaining

But political problems have broken cover – the research creates debate. Sector experts may say ‘we knew about this, but couldn’t mention it, or did not have the mechanisms through which to raise it.’

There is acceptance that some political constraints are real, though there can also be rejection of others that make things look too broken.

4. Depression

There are new insights about how politically stuck the sector really is. This makes previous plans look a bit simplistic. This is depressing. Sector experts may have also been riding high on their privileged access to key players in the sector. These same people may now be described as obstacle to change, rather than as the positive movers and shakers they used to be. (I write this with empathy and sadness – I have been there!).

5. Acceptance (the dream)

The political research is mentioned in the next sector plan. If it has been made public, it may even be mentioned in other organisations’ sector plans and discussed in policy debate – both private and public. This is the dream! Political constraints are now on the table (not just whispered about in corridors).

It is recognised that political constraints reduce the effectiveness of sector investments, and therefore there is a strong logic for moving beyond ‘technocratic’ approaches.

A year on (more of the dream…)

The bruised but healing sector experts say ‘can you design and fund some more political research or quick analysis for us?’. 

Politics has broken through, with further research and analysis building on the earlier work. Political researchers, and those commissioning it, get invited to more meetings. An increasing number of interventions get more effective.

Of course, the kind of people who fund political research typically have pretty limited and niche budgets. The ideal budget to dip into is the big sector budget for which political constraints are a drag on effectiveness. We try to persuade them that they should spend their own money as a means of ‘de-risking’ their investment, including noting that international business spends heavily on ‘political risk’ analysis. Why not the public/development/ aid/philanthropic sector?

So, is political research really good for you?

I’ve been telling people this basic message for many years. When I put it out on Twitter back in 2020, I got some very positive messages from some senior people in aid, but no cheque in the post. So I set out to create the change I wanted to see in the world myself. 

Slowly an idea formed in my mind. What the field needed was an open-access platform, funded by philanthropy rather than government, hosting quality-assured, plain-language, actionable Political Economy Analysis (PEA) that could inform mainstream policy and practice, and normalise its use in growth and development. Such a platform could address both supply and demand constraints (and retire the ‘five stages of political research grief’). It could demonstrate that political research really is good for you (in the end). 

Today (2026), this project has a home at the Blavatnik School of Government, University of Oxford. It has the renowned Professor Stefan Dercon, former chief economist of DFID and author of Gambling on Development, as its academic director, alongside a brilliant core team. And it has a name: Open PEA.

Welcome to the revolution. We are open. Everyone is invited.