Golden Age or Depressed Decade? Breaking down the World Bank Global Economic Prospects

How leaders are deciding to spend their next ten years

Jun 12, 2026

Article voiceover

0:00

-13:47

It can feel as if the lights are going out across the global economy. Outside big tech and winner-takes-all platforms, that is not the whole picture. New market positions are forming inside the largest, hardest, longest-running problems on earth - and the economics of holding those positions have shifted decisively in the past three years. This article is a map of where structural opportunities sit, and what is required to plug into them.

The growth engines of the last thirty years – from cheap capital to ever‑expanding trade and grant‑funded development – are either contracting, rationalising, or becoming more competitive. At the same time, the economy is being pulled toward three forces: AI, energy transition, and regionalisation, all of which run straight through hard problems. Taken together, they point to an architectural shift in how the world works. Hard problems – those that cross borders, are not owned by any one institution, and affect millions or billions of people – are becoming the most important market positions of our time.

The problems are not new. What has changed is the economics of solving them and the international environment they sit inside. And what is emerging is how leaders look at the world and decide where to spend the next decade of their attention.

Across health, financial inclusion, land rights, education, and climate infrastructure, three things are now true: outcomes can be measured with far greater precision; people and places can be reached at a fraction of previous costs; and capital can pay for verified results, not just the intent to deliver them. Taken together, that turns hard problems from obligations into structural vacancies in the global market.

A Darker Macro Picture

At the same time, the wider macro picture has darkened. The World Bank’s June 2026 Global Economic Prospects presents a balance sheet of a lost decade: global growth slowing to 2.5%, the weakest pace outside outright recession in almost 20 years, and nearly half of developing economies failing since 2019 to narrow the income gap with richer countries.

By the end of 2026, one quarter of developing economies, one third of low‑income economies, and half of fragile and conflict‑affected states will be poorer than they were before COVID‑19. Private investment growth in emerging and developing economies has more than halved relative to the 2010s.

The Emerging Forces

And yet, inside that bleakness, three strong forces are gathering: AI, clean energy, and a surge in regional trade. Each is, at its core, a response to a hard problem.

AI is shorthand for a productivity crisis that standard tools have failed to fix. Even if it “under‑delivers” against the current hype, the Bank’s own estimates suggest that broad adoption would still lift global growth in the 2030s above the average of the 2000s, making it the most prosperous decade since the 1970s. Energy transition is no longer framed solely as climate policy; clean energy now accounts for two‑thirds of all global energy investment, reaching a record US$2.2 trillion in 2025, driven as much by energy security and resilience as by emissions targets. Regional trade agreements have proliferated from over 300 in 2020 to nearly 400 today, now governing around 60% of global trade and binding developing economies into denser regional architectures with clearer rules on investment, standards, and services.

If we zoom out, what we see is: growth is increasingly constraint‑driven. The constraints are hard problems.

Hard Problems: Where Potential, Profit, and Purpose Meet

Hard problems have always existed. What has changed is the economics of solving them. The collapse of the old development architecture in 2025 - USAID dismantled, UK aid redirected, bilateral flows into Africa falling by around 40% in 2026-27 - did not make the problems go away. It left a structural vacancy. The institutional model that assumed sovereign states would sustain grant‑based aid at scale, multilaterals would coordinate delivery, and funding continuity was the baseline condition for design simply no longer fits inside the remaining fiscal space.

The numbers are stark. Global military spending reached US$2.7 trillion in 2024 - ten consecutive years of growth, 2.5% of world GDP, and roughly thirteen times total global development aid. Net official development assistance fell 23% in real terms in 2025, the largest annual drop on record, and is expected to decline further, stripping away one of the last buffers supporting basic services in many low‑income countries. More than half of the world’s low‑income countries are in or at high risk of debt distress. The SDG financing gap has widened to between US$4 and 6.4 trillion per year.

Meanwhile, demand for essential services has not moved. Six hundred million children are in school and not learning. 1.1 billion people lack formal land rights. Five billion cannot access basic legal help. In 2024, 4.9 million children died before their fifth birthday, most from preventable causes that existing interventions could have avoided. At current trajectories, around 27.3 million children will die before age five between 2025 and 2030. These are tragedies. And, they are structured markets where the cost of the status quo already exceeds the cost of proven solutions.

That crossing point - the moment when solving becomes cheaper than leaving a problem unsolved - is what I call the Crossover Point. It is a calculation, not a metaphor.

In homelessness in the UK, Nicholas Pleace’s work for Crisis established that one person sleeping rough costs the public purse around GB£20,128 per year across emergency services, healthcare, temporary accommodation, and legal costs, while a successful early intervention to prevent that homelessness costs around GB£1,426 - a 14‑to‑1 ratio. The crossover has already happened. The question is whether our models, and our capital, behave as if it has.

In child survival, Living Goods’ digitally enabled community health model (the DESC model) delivers a 27-28% reduction in child mortality at US$3.09 per person per year - cheaper per outcome than what governments were already paying. That turns under‑five mortality from an unquantified tragedy into a contractable unit of survival a minister can defend to parliament.

In financial inclusion, M‑KOPA built pay‑as‑you‑go credit rails for a segment traditional finance had written off as “too risky to serve”: 400 million people across sub‑Saharan Africa without access to formal credit. By using mobile money and real‑time behavioural data, they made the loan itself the mechanism for generating credit history rather than the reward for having it, extending over US$2 billion in asset finance to seven million customers and turning “credit invisibles” into a bankable consumer segment.

Doing Good? Or Good Business for the Greater Good.

What these organisations have in common is that they hold structurally defensible positions inside hard problems where the crossover has already occurred, the chokepoints have shifted, and outcomes are contractable. They’re not driven to “do good”.

Underneath, four structural shifts are doing the heavy lifting.

First, the global aid architecture has collapsed, yet emand remains. The old model has exited the field, leaving in‑market stakeholders to look for new partners, new models, and new capital that does not carry the conditionality or timelines of the previous system.

Second, the cost of delivery has collapsed. The cost of frontier language‑model inference, the cognitive layer for tutoring and legal documentation, fell by about 98% between 2023 and early 2026, dropping from roughly US$60 to around US$0.75 per million tokens. National‑scale digital identity registration, once US$20 or more per person, now runs at under US$1 per person. Satellite imagery suitable for monitoring and evaluation has become cheap enough to replace large parts of field‑verification budgets. Deworming a child still costs around US$0.50 and yields an estimated US$169 in lifetime economic returns per dollar invested. Each of these cost curves changes which models are commercially viable.

Third, the chokepoints have dissolved. Accreditation, centralised data gatekeeping, correspondent banking, and institutional chains of trust were features of the old architecture, not of the problems themselves. DESC community health workers match or exceed clinic outcomes without being professionally registered. Legal empowerment organisations provide land tenure and contract enforcement through paralegal networks that never touch a formal court. Conflict‑zone evidence can be secured via cryptographically verified provenance, rather than institutional custody.

Fourth, outcomes have become contractable. We can put a price on a single well‑being‑adjusted life year (a WELLBY) at GB£15,900, endorsed by HM Treasury and used in social value calculations. We can price one child reading fluently at US$12 per verified outcome unit. High‑integrity carbon credits now require verified land tenure and attract investment‑grade capital. Remittance flows to Africa exceeded US$100 billion in 2024, surpassing aid; if routed through outcome contracts at current unit costs, they could finance hundreds of millions of verified health or land outcomes.

All of that runs over digital rails - UPI in India handling 16.6 billion transactions in February 2026 (US$274 billion in value), Pix in Brazil reaching 170 million users and processing more daily transactions than Visa and Mastercard combined, with the UPI architecture now licensed to Peru, Namibia, and Trinidad and Tobago. Proven models can now travel without headquarters.

So where does this leave each of us?

The World Bank warns that, barring a miracle, the 2020s will be a lost decade for dozens of developing economies, with per‑capita income in EMDEs excluding China and India not regaining its pre‑pandemic relative position versus advanced economies until after 2028. It also notes that between 2019 and 2025, the number of people facing severe food insecurity in EMDEs (excluding China) rose by around 220 million, with a further 35-70 million at risk of being added by 2028.

At the same time, it points to AI, clean energy, and regional trade as the forces powerful enough to make the 2030s a golden era for job creation and growth if we are deliberate now. Those forces do not sit on top of hard problems; they run through them.

Essentially, we get to decide our focus. And the action of our next decade.

Deciding Your Next Decade

I am interested in this as a decision lens for leaders who already have the resources and ambition to build something that shifts a hard problem in a measurable way:

If you are a CEO, your exposure to hard problems is already on your balance sheet - in supply chains that depend on undocumented land, workforces without basic health coverage, or markets where financial exclusion caps your customer ceiling.

If you are an investor, you are watching models that look “too impact‑shaped” for traditional mandates but whose unit economics are moving faster than your portfolio assumptions.

If you are a founder or a former development leader, you may be circling a problem you care about, funding pieces of it, joining boards, speaking on panels - and still not have a structure that tells you whether there is a real market position inside it.

Hard problems, solved at global scale, with verified outcomes and a cost base that drops every year – that is not a gap in the market. That is the market. It is how we find the market inside the mission of doing work that is necessary, needed, and, for a number of leaders, the next focus of their career.

The lights are going out. And there’s glimmers of light shining though. Against this backdrop, the question I am sitting with, and that I invite you to sit with alongside me, is:

For the problem you care about most, has the crossover already happened - and if it has, are you prepared to treat that problem as your primary market position in the decade ahead?

Soft Power, Hard Problems with Sophie Krantz

Discussion about this post

Ready for more?