Previous
This short series titled Evidence as Politics in Policy explores a simple but often overlooked point: evidence does matter in policymaking, but never on neutral ground. In all political systems, evidence is used strategically, selectively ignored, unevenly protected, and shaped by democratic and non-democratic pressures, institutional design, and political incentives. Across these articles, I argue that the challenge is not simply to produce more evidence or communicate it better, but to understand the political, economic and social conditions under which evidence is used, contested, or sidelined — and what that means for anyone interested in evidence-informed policy.
The reflections in these articles have been drawn from a series of interviews with political scientists in Latin America, Africa and Asia as part of our ongoing effort to better understand the political economy of evidence use.
Read until the end for summaries of the interviews.
If we only look at the most politicised parts of government, it is easy to conclude that evidence never really stands a chance. This is how I feel, sometimes.
Electoral pressures are short-term, public opinion is volatile, ministers come and go, policies are reversed for symbolic reasons, patronage distorts appointments, and interests intervene in procurement. And this is before we take AI and misinformation into account!
In such environments, evidence appears fragile, selective, and often expendable.
And yet that is not the whole story.
In fact, there are plenty of examples of evidence making a difference, and civil servants, policymakers, thinktankers and activists making it possible – against all odds.
Even in highly complex, technocratically weak and extremely politicised political systems, some parts of the state still manage to use evidence with a degree of consistency. They are rarely the whole government and do not pretend to escape politics altogether, yet they still manage to retain more room to plan, justify, and implement decisions based on data, technical criteria, and professional standards.
Political scientists often refer to these as pockets of effectiveness, pockets of excellence or islands of excellence: institutions that manage to function relatively well even when the wider system around them is unstable, politicised, or weak.
They matter because they show that the use of evidence is not an all-or-nothing question.
Unlike the universe, it is unevenly distributed: it depends on where one looks, who is protected, what is at stake, and how much autonomy an institution is allowed to preserve. +
Not all institutions are equally exposed
One of the clearest lessons from our discussions with political scientists across regions is that governments are not internally coherent in their use of evidence.
Some ministries and agencies operate under such intense political contestation that long-term planning becomes almost impossible. Others, often dealing with high-density and more developed fields such as finance, trade, revenue, or high-level economic coordination, are expected to maintain a higher technical standard because the costs of failure are simply too high.
The reason is not that the people who work there are less political than everyone else – they may be; but this does not protect them from political reshuffling. It is that governments are generally less willing to gamble in these sectors. Fiscal decisions, trade negotiations, revenue systems, and macroeconomic management are too consequential, too visible to international actors, and too vulnerable to costly mistakes.
In Peru, the Central Bank+, the Ministry of Economics, the Ministry of Foreign Affairs and some parts of social development ministries or autonomous bodies have maintained a more professional, evidence-driven standard than other sectors. They have been able to rely on stronger technical teams, better policymaking routines, and closer links to independent communities of expertise.
However, this has not made them immune to politics. It has just provided more room to operate with discipline.
The contrast with Ministries of Education is telling. As I mentioned before, education sits much closer to everyday political conflict: parents, teachers, unions, ideological disputes, debates about national identity, and repeated leadership changes have something to say about education.
Ministries of Education can become sites of constant reversal, with each new minister seeking to establish a new policy brand, often at the expense of continuity.
This distinction also plays out across different levels of government. In decentralised systems such as Spain, Indonesia, Nigeria, or Brazil, the national government may produce evidence and issue policy guidance, but local authorities can still ignore it. Local political leaders have their own incentives, their own constituencies, in many cases their own rivalries with the centre – and sometimes access to their own data. This means that pockets of evidence use can exist at one level of government while being absent at another.
They cluster, unevenly, in particular institutions, sectors, and levels of government.
If these islands of excellence exist, how do they survive?
Protection matters
The most consistent answer from the interviews is that they are protected – tooth and nail and by a broad cordon sanitaire.
Their room to manoeuvre is rarely the product of institutional design alone. Instead, it depends on deliberate political shielding from the top: a president, prime minister, a powerful political coalition or interest group may decide that a particular agency must be allowed to function with relative autonomy because it is central to a priority they care about: raising revenue, stabilising a capital city, managing trade, or maintaining credibility with external partners.
The example from Peru above illustrates this particularly well. And in Uganda, the Uganda Revenue Authority and, historically, the Kampala Capital City Authority are often both presented as institutions that could act as pockets of effectiveness because they were deliberately protected from disruptive political contestation.
Professional standards create room to act
Protection alone is not enough, of course. It only matters if there is something worth protecting.
These pockets of effectiveness also tend to have stronger professional norms, more technically capable staff, and better access to data and analytical support. They are often staffed by civil servants or advisers who understand how to translate complex problems into defensible policy choices and who can justify those choices publicly.
Technical ministries often retain their effectiveness because they are closely linked to national statistical institutes, established administrative systems, or part of wider, powerful professional ecosystems. For example, technocrats in the Central Bank, the Ministry of Finance or the tax authority are part of national and international economics and finance professional networks – themselves with significant power.
Some of these bodies have developed (and protected) independent hiring processes to avoid capture. In Peru, the Central Bank and the Ministry of Foreign Affairs hire through postgraduate courses delivered in partnership with top independent universities.
The result of this technocratic corps is that policy decisions have a firmer empirical basis, and it makes it harder, though never impossible, to replace technical judgment with improvisation.
Unfortunately, while none of this happens overnight, it can be unravelled in a blink of an eye.
Their strength is also their fragility
It would be a mistake, however, to romanticise these islands of excellence as they are often more fragile than they appear.
Pockets of effectiveness are not stable settlements. They depend on an unusual alignment of technical capacity and political protection. Once that alignment breaks down, they can quickly lose their autonomy and become subject to the same short-term pressures, patronage, and conflict that affect the rest of government.
We’ve seen these pockets of professionalisation unravel in Peru in just a few years.
What these pockets teach us — and what we still need to learn
Orazi Bellettini, the founder of FARO in Ecuador, had the idea to write a case study on how Ecuador reached the 2002 World Cup. He argued that Ecuador had qualified by doing what they were meant to do. In a country where everything else was failing, this proved that a better outcome was possible!
I agree. These institutions matter because they show that even in difficult political environments where the incentives I described in the second article of this series push evidence and expertise to the sidelines, evidence can become part of routine decision-making when certain conditions are in place: political shielding, strong professional norms, technical support, and a clear sense that failure would be costly.
They also show that the challenge is atemporal: the defence of excellence cannot stop.
But to extend what works on these islands of excellence, we need to better understand how these pockets emerge in the first place. Are they created deliberately through political design, or do they develop gradually through bureaucratic professionalism and institutional habit? Who shields these institutions, for what purpose, and under what political conditions? What kinds of sectors are more likely to generate such pockets, and are some policy domains simply more hospitable to evidence because they are less electorally exposed or more tightly tied to international scrutiny?
Just as importantly, we need to study their fragility. What causes these pockets to endure beyond particular ministers, presidents, or political coalitions? What makes them collapse? Can their practices be transferred to other ministries, or are they too dependent on highly specific political settlements? And what happens when decentralisation, coalition politics, or administrative turnover disrupt the protections that once sustained them?
Finally, there is also a broader comparative question: if pockets of effectiveness appear in very different political systems, what does their comparison tell us about the real conditions under which evidence use survives in government?
Anyone interested in funding this type of research? Contact me!