The future of social protection: Adapting to a world in flux

Behind this fiscal squeeze, a deeper design problem is poised to shape what happens next: the basic architecture of social protection belongs to another era. The classic model was built for a stratified world where formal unemployment was a predictable, insurable risk, and the poor formed a relatively static, coherent bloc whose characteristics were stable enough to be reliably targeted with household surveys carried out every few years. But as work becomes fragmented and shocks become the rule rather than the exception, the traditional poverty relief model is failing. In economies dominated by climate volatility or gig work, narrow targeting is futile because income swings are frequent and a single event can impoverish a previously secure family overnight.

As we move towards 2026, these factors are accelerating a few key trends that are reshaping the system. The first trend is the shift away from both narrow income-based targeting and the sweeping ambition of Universal Basic Income (UBI), a minimum unconditional monthly payment for every adult. While UBI has been widely discussed as a potential solution, its cost of 20–30 per cent of gross domestic product (GDP) across most world regions make it fiscally unrealistic for most economies. Instead, governments are turning to categorial targeting or ‘bounded universalism’. This term, while not yet standard in the literature, captures an emerging approach: granting benefits to entire categories of people who are universally recognized as vulnerable, such as all children or all people with disabilities. By doing so, policymakers avoid the complex and expensive bureaucracy of means testing without incurring the massive fiscal cost of UBI. The emerging pattern is that countries start with predictable payments to specific age groups or climate-vulnerable regions to act as a prosperity floor that helps families withstand shocks. This approach, of which Universal Child Benefits (UCB) are a prime example, represents a pragmatic middle ground between traditional targeting and UBI, and is likely to remain at the centre of social protection advocacy from agencies such as the International Labour Organization (ILO) and UNICEF.

Running alongside this move towards bounded universalism, we can expect an acceleration of a second trend: algorithmic targeting. This is becoming increasingly common as governments grapple with the tension between tight budgets and the reality that traditional targeting often excludes large shares of those it aims to reach.  As digital capacity grows, more governments turn to artificial intelligence (AI) and big data to decide who gets support - mobile phone records, payment histories, and satellite images are increasingly used as proxies to estimate who is poor or severely affected by a shock.

While this can be an appealing way for finance ministries in debt-distressed nations to stretch budgets, the danger is that its expansion could be driven by what is fiscally viable rather than what produces the best socioeconomic outcomes despite well-known risks. When carefully designed and combined with strong human oversight, such tools can sometimes identify and reach people in need more quickly, especially during shocks.  But without strong safeguards (transparent criteria, independent audits, effective appeal mechanisms and robust data protection), algorithmic targeting tends to favour those with a digital footprint, miss the poorest and most remote, shift the burden of proof onto the vulnerable, and can be misused to impose surveillance and behavioural monitoring.

A third decisive shift concerns social protection financing in economies where most workers are outside traditional payroll systems. In the Global South with already large informal sectors, the surge of hybrid non-formal work is further straining traditional employer-based payroll models. In its place, semi-contributory financing is emerging as a potential structural bridge. These systems, where the state matches or tops up the modest contributions that low-income workers can make, have the potential to bring the missing middle of informal and gig workers into the safety net at scale. Simplified monotax regimes, where small businesses pay a single flat fee that covers both tax and social security, are part of the same evolution. Over the next decade, semi-contributory schemes could steadily replace pure contributory systems in many countries, rebuilding the revenue base while turning people from passive recipients into active contributors.

In addition to these structural, technological and financial changes, anticipatory policymaking is becoming a defining feature of social protection, enabled by the growing availability of data and advances in predictive analytics. Rather than merely responding to existing crises, governments are beginning to design systems that act on early signals, shifting the question from who receives support to when it should arrive. By linking climate and market information directly to social programmes, governments will increasingly set pre-agreed thresholds, such as specific rainfall levels, river heights or food prices, that automatically trigger cash transfers before crises peak. A thin but growing evidence base already shows returns outperforming purely reactive responses to emergencies. However, this trend is likely to remain financially fragile, as anticipatory systems depend on pre-arranged financing at a time when all sources of funding are under pressure. Their future turns on scaling instruments such as parametric insurance (coverage that pays out automatically based on data triggers rather than long damage assessments) and catastrophe bonds that can guarantee liquidity when triggers are hit.

A further, quietly powerful shift could stem from mobility. Portable social protection, the ability for rights and benefits to travel with people, is essential for climate migrants and platform workers whose livelihoods cross borders. Currently, 627 bilateral and 19 multilateral social security agreements exist, but only one in four international migrants benefits from them, and 84 per cent of bilateral agreements involve European countries – leaving South–South migration corridors severely underserved. Pushback against migration in developed countries means that we are unlikely to see much progress on this front beyond specific regional blocs, yet the tension between portability and exclusion will be decisive in determining whether or not children are protected.

For children, these trends are not abstract. Rather, they dictate the reality of whether they receive food, health care and schooling. The decisions made in the next few years regarding the design of the social protection systems of the future (digital platforms, ID systems, eligibility rules, and financing) will lock in trajectories for children (and therefore global development outcomes) for 2030 and beyond.

The infrastructure now being designed could signal which future we are moving towards, but the path forward is not a simple binary choice. The digital platforms, eligibility rules and financing models emerging over the coming years will likely blend several distinct trends. The challenge is to integrate these pathways (universal child benefits, semi-contributory systems for informal and hybrid workers, anticipatory triggers for climate risks, and portable benefits for migrants) into a coherent system that prioritizes resilience. While algorithmic targeting is likely to be in the mix for fiscal reasons, it must be carefully managed; if left unchecked, it risks creating systems that look efficient on paper but remain exclusionary in practice. By anchoring these technological and financial shifts in solidarity and rights, guaranteeing a universal floor for every child, we can ensure that twenty-first-century volatility is not pushed onto children’s shoulders. In this integrated future, technology still plays a central role, but as a tool that makes delivery faster and safer, rather than an automated gatekeeper.