Research on VoxDev in 2025 has shed light on getting stuff done, paying for it, the changing development landscape, and a whole lot more.

The world is changing rapidly. Can research inform what comes next?

12 months ago, I identified ten key takeaways from the development economics research I read in 2024. That post ended up being our most read piece from 2024, so I guess I’m doing this every year now.

Below, I have outlined some of the main lessons from the articles, podcasts, and VoxDevLits we released in 2025. After my ten key insights, I have included a few quickfire honorable mentions - we had more content than ever before this year, making it particularly tough to cut down to ten! I end by reflecting on last year’s blog.

Do you want insights from delivered to your inbox each week? Of course you do, so sign up to our newsletter.

1. Getting stuff done (aka bureaucracy): Boring, lifesaving

Bureaucracy is a topic in need of a rebrand – a boring word with huge importance. I like to think of it as how governments can get stuff done. Good ideas might not be worth much by the time they’ve been implemented poorly, so putting bureaucrats in a position to succeed is crucial. In fact, it can literally save lives.

In Chile, Pablo Munoz and Cristobal Otero show that merit-based recruitment and better pay attracted better-trained managers to public hospitals, which in turn reduced mortality rates. In India and Pakistan, Gemma DiPoppa and Saad Gulzar show that properly leveraging existing bureaucratic structures can substantially reduce crop fires, lowering child deaths without the need for new institutions or subsidies.

Improved technology can also help bureaucracies be more effective. Early on during the HIV epidemic in Malawi, the government decided to adopt an electronic medical records system, developed by local NGO Baobab. Laura Derksen, Anita McGahan and Leandro Pongeluppe found that this dramatically improved patient survival, as the system made it easier for staff to track down and contact lapsed patients. For more on this topic, our VoxDevLit on Bureaucracy summarises everything you need to know about getting hiring right and making sure bureaucrats have the right incentives.

2. Money’s tight: How can governments pay for stuff?

Governments everywhere are looking to raise funds. As highlighted above, getting the personnel right in the tax authority is critical. Communication strategies with taxpayers are also vital. In our VoxDevLit on Taxation, Anders Jensen, Anne Brockmeyer and Lucie Gadenne highlight these as two ‘building blocks’ of a well-functioning tax authority.

In the past at VoxDev, we’ve highlighted examples of tax policies that worked well. In Colombia, Rachid Laajaj, Marcela Eslava, and Tidiane Kinda show digital reforms at ports in Colombia encouraged growth and increased tax revenues. In the mid-2000s, Indonesia re-organised its tax administration structure by allocating larger firms to dedicated tax offices. Chatib Basri, Mayara Felix, Rema Hanna and Ben Olken show this dramatically increased revenue at low cost to the government.

This year, research has reinforced the importance of taxpayers perceiving the system as fair. Michael Best, Luigi Caloi, Francois Gerard, Evan Kresch, Joana Naritomi and Laura Zoratto look at property taxes in Manaus, Brazil. They find that inequities in the taxes which households owe, inequities created by a system of crude proxies of land values, dramatically reduce compliance.

3. Governments have SO MANY options to improve lives

Well, duh! What point am I trying to make here? I’ve sometimes felt that all of the focus on the (very real) constraints that policymakers face robs them of their agency.

It certainly makes sense to account for these constraints if you are a researcher wanting to translate your research into impact. But in terms of what societies can expect of their leaders, is it too much to ask for a leader who just gets stuff done, good stuff? And what would that good stuff look like?

Research on VoxDev has shed light on so many government policies that were a success. Personally, these are my favourite papers. For a start, it’s much easier to refer to this type of research when discussing policy.[1]1. Which feels more convincing: a government did this and it worked vs* a small randomised pilot in this village did this and it worked*.

So, after that ramble, here are just a few of the examples I’d highlight to an ambitious leader wanting to improve lives.

• In Burkina Faso, an extraordinary vaccination campaign reached over a million children in two weeks, and transformed lives for a generation (Richard Daramola, Md Shahadath Hossain, Harounan Kazianga, and Karim Nchare).
• In India, free sanitary pads were distributed at schools through the SHE Pad Scheme – dropout rates reduced dramatically as a result (Sumit Agarwal, Liu Ee Chia, and Pulak Ghosh).
• Embrapa, a publicly-funded R&D programme which focused on agricultural innovation suitable for Brazil, massively improved agricultural productivity (Ariel Akerman, Jacob Moscona, Heitor Pellegrina, and Karthik Sastry).
• China implemented the New Rural Pension Scheme between 2009 and 2012, which provided rural residents over 60 with a basic income transfer of about 30% of their annual income. This policy had huge knock-on benefits on younger generations, and ultimately even boosted growth (Qingen Gai, Naijia Guo, Bingjing Li, Qinghua Shi, and Xiaodong Zhu).
• And here are nine more examples of successful policies from Colombia, India and Mexico.
• And here are seven more win-win policies in development.

4. Education: A good investment for growth, poverty reduction, better health and more

Whilst we are on the subject of good options, we have featured a whole host of research on education over the past year, applying a range of methodologies from large-scale RCTs, to economic history, to growth accounting. They all highlight the importance of education to children, children’s future children, and society as a whole.

Amory Gethin shows that education has been a key driver of growth and poverty reduction. He estimates that, since 1980, education accounts for almost half of overall income gains across the world. Alexandra de Pleijt and Ewout Frankema study Southeast Asia specifically. They document new evidence on the crucial role of higher education in Southeast Asia’s take off - it wasn’t just good policies in the 1970s, early investments in education paid off massively over the long run. And evidence goes further back, all the way to the 1870 Education Act, which created England’s first publicly funded schools targeting working-class children. Ben Milner shows that this policy change massively boosted social mobility.

In Ghana, results from a 15 year RCT show that the impacts of subsidising secondary schools extend to the next generation. Esther Duflo, Pascaline Dupas, Michael Kremer, Elizabeth Spelke, and Mark Welsh find cost benefit ratios of 1:30 (and potentially as high as 1:110) from a programme that gave full scholarships to rural students so they could attend high school. A key benefit that became obvious in the follow-up was that female recipients of scholarships, and their children, went on to have much better health outcomes. Even if we only judged this programme by its impacts on infant mortality (of the children female recipients had later on), at a cost per life saved of $170-260 it would be a competitive public health policy.

Is all this that surprising? Maybe not. The importance of education does feel intuitive, obvious even. But somehow, when the chips are down, cuts are being made, and budgets allocated, the education field seems to lack the dramatic, life or death, urgent framing that other areas have to hand. So, we would do well to remember that much of the world’s progress is built on a foundation of education.

5. More tech ≠ more learning

There are lots of effective policies and good investments in education. Merely expanding access to new technology is not one of them.

The One Laptop per Child programme was a widely publicised initiative that gave students in over 40 developing countries a personal laptop. Santiago Cueto, Diether W. Beuermann, Julian Cristia, Ofer Malamud, and Francisco Pardo followed up 10 years later in Peru and found that laptops did not improve academic achievement in the long run.

In our new VoxDevLit on EdTech, Abhijeet Singh summarises papers across Peru, Uruguay, Costa Rica, Honduras and China. They all reach the same conclusion - these programmes simply don’t improve learning. And it’s not just laptops, evidence on the expansion of internet access also shows mixed impacts.

You would hope a strong evidence base like this would change policymakers’ minds. Well, Stefan Dercon elaborated on Kenya’s experience where this evidence failed to stop the President from committing to a laptop policy against the advice of researchers, who knew there were far better alternatives.

And what does this mean for the question on everyone’s minds in education, how will AI impact learning? Despite the noise, and assumptions by leading labs, it’s far too early to tell and we need a lot of research. But history tells us one thing, having lots of children using new technology guarantees nothing.

6. Industrial policy is back on the menu

Industrial policy is back in fashion. In fact, new measures from Réka Juhász, Nathan Lane, Emily Oehlsen, and Verónica C. Pérez show that it is advanced economies leading the charge.

So, the Overton window has shifted. But how should developing countries choose who to target? Tristan Reed shows that targeting is difficult, but doable. Actually, many export promotion strategies have successfully coordinated public inputs to help sectors grow. This approach should not just copy wealthy countries and does not need to involve picking specific firms. Instead, countries should target industries in the export sector that have achieved scale or show good growth potential. And once they’ve picked a sector, what next? Tristan Reed recommends a WTF ‘what’s the failure’ approach, i.e. fixing market and government failures rather than defaulting to protection through tariffs, or subsidies.

7. Trade getting harder will be bad for development

We have featured more and more and more and more evidence on the benefits of trade for development:

• In China, workers at exporting firms, particularly those exporting to high-income destinations, earn and learn more – Xiao Ma, Marc Muendler, and Alejandro Nakab.
• Also in China, granting firms the right to trade internationally boosted productivity, and gains built over time as firms learn by trading – Yunong Li, Yi Lu, and Jianguo Wang.
• In Mexico, NAFTA led to lower prices for consumers, whilst producers saw larger profit margins - Felipe Brugués, Ken Kikkawa, Yuan Mei, and Pablo Robles.
• In India, falling tariffs reduced the importance of being politically connected for firms, improving efficiency by reducing misallocation due to politics – Sebastian Jävervall and Roza Khoban.

And we know that no country has climbed the income ladder without participating in global trade.

Yet, it now seems questionable that a decades long trend of falling trade barriers and fewer import substitution policies will continue. As Penny Goldberg recently noted on our podcast:

“Growth miracles seem less likely because two of the key mechanisms – access to big, lucrative markets and willingness to share technology across borders – no longer seem to be here.” Penny Goldberg on VoxDevTalks.

Along with offering fewer chances for transformative development, our more fractured world will generally be less efficient. For example, this article provides an intuitive framework for thinking about food security specifically. Tasso Adamopoulos and Fernando Leibovici show that while countries can access cheaper food through trade, this leaves them vulnerable to trade disruptions. As trade risks intensify, retreating from global markets is one way that food-importing countries might respond.

For more on trade, check out our VoxDevLit on international trade by David Atkin, Amit Khandelwal and Co-Editors, updated this year. A key emphasis for policymakers, to capture the benefits of trade, is complementary policies which address domestic barriers.

Another aspect of global integration is foreign direct investment, and you’re in luck, as we also now have a VoxDevLit on this. Amongst many other insights, Stefania Garetto, Nina Pavcnik, Natalia Ramondo and Co-Editors outline some key steps for developing countries looking to attract FDI:

• Capital account liberalisation.
• Promoting the expansion of multinational banks, which offer an established network of financial services and expertise.
• Trade liberalisation.
• Investment promotion initiatives that target outreach efforts build an international image.
• Cultivating a suitable workforce in targeted industries.

8. Conflict: On the rise, also bad for development

Another worldwide trend not helping development is the global rise in conflict and fragility. Over one billion people now live in economies experiencing fragility and conflict.

Both climate change and natural resource booms are emerging as significant conflict triggers. Oliver Vanden Eynde and Juan F. Vargas show that extreme weather disasters systematically increase violence (especially in agriculture-dependent regions), and surges in the value of minerals or oil can ignite conflicts over resource grabs.

Institutions have an important role to play. Laura Mayoral and Hannes Mueller show that good institutions, think democracies with inclusive political processes, checks and balances, and effective state capacity, can act as engines of peace. And beyond the broad role of institutions, Elena Esposito and Austin L. Wright show that specific policies can help steer individuals away from conflict. Evidence shows that education, employment opportunities, and cognitive behavioral therapy for at-risk youth can all reduce violence by changing incentives and outlooks.[2]2. Footnote: 2025 marked the first year of the new CEPR ReCIPE programme (Reducing Conflict and Improving Performance in the Economy), which is funding fresh research into conflict. Check out our full coverage of this work here – ReCIPE on VoxDev.

9. Infrastructure: Build it, then maintain it

Between 1997 and 2014, Ethiopia transformed its road networks through a large-scale road building programme. Tasso Adamopoulos shows that this had a host of benefits and accelerated structural change. And this Ethiopian programme wasn’t just about laying tarmac and walking away; it also focused on quality and upkeep. In that period the share of main roads in good condition rocketed from just 17% to 73%.

Building infrastructure of any variety is only the first step, as infrastructure only delivers value if it keeps working. Research on sanitation infrastructure, summarised by Britta Augsburg, Andrew Foster, Molly Lipscomb and Co-Editors in our VoxDevLit, drives this point home. Desludging and maintaining latrines have significant health benefits. And when there aren’t subsidies and follow-ups, latrines fall into disuse and health gains are eroded.

This is also the case for electricity. Poorly maintained electricity infrastructure, common in low- and middle-income countries where many utilities operate at a loss, leads to outages and voltage drops. In our VoxDevLit on electricity infrastructure, Robyn Meeks and Meera Mahadevan highlight studies showing that improving the reliability and quality of electricity supply has huge economic payoffs – firms become more productive, and consumers gain confidence to use services more.

Let’s build the infrastructure. Let’s also budget and plan for the less glamorous work of maintaining it.

10. Prevention is cheaper than the cure: Anticipating problems before they become entrenched

We featured a wide range of economic history research this year, including a piece by Andrea Matranga and Timur Natkhov which explored why serfdom proved to be so persistent in Russia. They find that a short-term trade for security ended up embedding an extractive institution for centuries.

It can be hard to step back and think long term, but tackling future problems today is often far cheaper and easier than waiting.

In our VoxDevLit on organised crime, Maria Sviatschi, Santiago Tobon, and Nicolás Cabra-Ruiz highlight the many serious consequences of criminal organisations which become embedded in society. While policy options do exist for governments looking to tackle organised crime once it is established, the reality is that governments face huge challenges, while societies bear huge costs. Nipping it in the bud is by far the best solution.

Political polarisation is on the rise everywhere, all at once. This undermines democracy, particularly in contexts with weak institutions, fragmented media markets, and ethnic, religious or regional-based politics, weakens accountability and depresses investment. Cesi Cruz, Horacio Larreguy and Ernesto Tiburcio outline what works for reducing polarisation, and highlight the importance of pre-empting backlash when intervening in such settings.

Deforestation is another area where prevention is the name of the game. But how do you get communities to forgo the benefits they would acquire from clearing forests, so that the whole world benefits from the services those trees provide? Allan Hsiao and Francisco Costa assess a range of options in their VoxDevLit. ‘Payments for Ecosystem Services’ (PES) are a key pillar of these efforts. Santiago Izquierdo-Tort, Seema Jayachandran, and Santiago Saavedra explored PES in Mexico, and show that smarter contract design can more than quadruple cost-effectiveness.

Some quickfire insights that almost made the list

Whilst I have your attention, here is some other stuff I thought about including, but didn’t, so I could stick to ten (for some reason).

• Carbon offsets often don’t work, or worse. We have been working on a series of economics explainer videos that will start in January, one of which focuses on carbon offsets, so subscribe to our YouTube and stay tuned!
• I have written quite a few blogs this year. The following was probably my favourite, but it was also one of the least read, so now’s your chance to read it! Common misperceptions: what people get wrong about the world and why it matters.
• Over the years, we have featured evidence on a range of coordinated, well-designed aid efforts that succeeded. Emaan Siddique brought this together in a blog: foreign aid success stories.

Some reflections on 2024’s insights

As I wrote in last year’s post, I hoped that 2025 would be a year where we started to make some serious progress on understanding the impacts (current and potential) of AI in low- and middle-income countries. I summarised what we do know in this blog: AI and development economics. If I only had two words: not much.

Unfortunately, the debate on AI is rapidly outpacing researchers’ attempts to generate evidence, particularly in LMICs. We are launching a new podcast in January, and the second season (February), will focus on exactly this topic, trying to establish what we do know.

Last year’s blog also touched on the problems faced by the agricultural sector in Africa. We continue to learn more about the pros and cons of different policy options (irrigation infrastructure, post-harvest loans, lowering travel costs, index insurance, fertiliser). Joshua Deutschmann, Maya Duru, Kim Siegal, and Emilia Tjernström’s research in Kenya suggests bundled programmes might be our best shot at boosting productivity.

Ok, time to conclude

The world is changing fast. You would like to think that research has something to say about what should come next, to positively shape future paths of development. We certainly feel that way at VoxDev, and we have lots of exciting plans for 2026 to cover even more of this work. As I have teased throughout this blog, we will be releasing videos, a new podcast, a bunch more Lits, and a whole lot more on top of that.