In the world of international development, why are we all so obsessed by innovation?
At the right time, social security innovation is important, vital even, but why is there a bias among donors towards the new? This bias can sound progressive, but it often comes at a cost, usually at the expense of supporting governments to build a basic social security system that includes the necessary nuts and bolts of the core delivery infrastructure. Supporting governments to build core systems usually requires relatively simple, practical approaches to enable delivery at scale. At times it involves innovation, of course, but this should not be at the disproportionate exclusion of other priorities.
Over-complicating social security
Unfortunately, the quest to build basic infrastructure – including comprehensive, lifecycle social security systems – can get lost in the search for the new. There are, of course, plenty of examples of donors successfully supporting core infrastructure. (I say this as, until recently, an employee of the UK’s Department of International Development). In social security, there are numerous occasions when donors have helped build elements of core social security systems including management information systems, single registries, payment systems etc, which do, indeed, often rely on innovation. But, examples of donors prioritising the development of comprehensive systems, including schemes that deliver on a national scale and are government-funded (so will last), are harder to find.
Too often, donors accompany some system support with financing small, over-complicated programmes that have little chance of reaching national scale or being sustainable in the long-term. Social security programmes are increasingly asked to address multiple agendas. This isn’t in itself a bad thing: international evidence shows that, as well as providing a minimum income, social security programmes can increase asset levels and labour market participation, improve access to health and education, reduce inequality, and generate economic growth, among other things. But, too often the existence of multiple agendas means we seek to complicate programme design, prioritising innovation over the basic nuts and bolts.
We complicate social security programme design in a number of ways. We connect the receipt of social security payments to conditions on school or health care attendance (when there is little evidence on the rationale for imposing conditions); or we require programmes to have the in-built capacity to scale up quickly in the event of a crisis (but how can a system be scaled-up that hardly exists?). Or else, programmes must have additional complex livelihoods support (when we know that many people will invest in livelihoods anyway, if they receive a regular and predictable benefit); or they target transfers at the poorest in a bid to save resources and maximise the impact (when we know that targeting the poorest is not possible to undertake with any degree of accuracy and that universal schemes have much large impacts on inequality). And so on. Too often we – with the best intentions – bend programmes to our agendas and, as a result, bend them out of shape. In doing so, we miss the wider priorities inherent in developing comprehensive national social security systems.
Donors can often become distracted by small-scale, over-complicated programme design. We too often create something we think we can defend through our organisational layers of approval, where incentives are often more about being innovative and having a short-term impact on corporate objectives within the typical 3 to 5-year programme cycle. International donors tend to employ people – very committed people – who have proven themselves in the academic world more than anywhere else. Consequently, they – we – lean towards the idea that things can be figured out technically, at a desk or around a meeting table. We create incentives for each other to find new technical solutions, to innovate, to ‘solve’ the problem. As a result, we too often get distracted; worse, we distract governments from prioritising simple, core, national systems.
Indeed, we often undermine the building of national, state-financed systems by designing schemes that governments are unwilling to scale-up. For instance, we know that, in a democratic context, governments are unlikely to invest national resources in poverty-targeted schemes, since they will not be popular with the majority of the electorate. Nonetheless, we still insist on them. We need to heed the lessons from the recent demise of Mexico’s iconic poverty-targeted Prospera scheme and the Government of Mexico’s move to establishing universal – and probably very popular – old age pensions and disability benefits.
Keep things simple – stick to the basic framework – and follow the example of governments
It may be possible to incorporate some additional elements to social security programmes. The impact of social security schemes may benefit from the provision of additional information, for example on good nutrition, or by connecting recipients with other services. Or, programmes that are already in place can be used to provide additional support during crises, as was done in Nepal after the 2015 earthquake.
But, as someone who now spends more time with governments than donors, I would argue that design should be kept simple. The pressures government officials are under, from donors with diverse opinions and sceptical and cash-strapped Ministries of Finance are usually considerable. If donors are to provide further support, I think two conditions should be met: (a) additional elements should only be considered when it can be demonstrated that the basics of programme and system architecture are in place or are being put in place (and are aligned to the political economy of a country, so they are sustainable); and, (b) additional support should be located within the appropriate sector of government. Otherwise, the hard-enough task of developing core national social security systems will become impossible.
Social security systems are built on a well-established model. Social security schemes provide a minimum income and protect us all against challenging lifecycle circumstances such as unemployment, disability and old age, or the costly (but joyful) task of raising children. They depend on basic processes which change little: identifying a group in the population to support, registering and enrolling them, paying them, then checking periodically that they are still eligible. These processes rightly involve innovation – using digital methods in information collection and mobile phone networks for making payments, for example. But, the framework for social security stays largely unchanged, whether across time or across countries.
Despite the model of building social security systems being well-established, it is not easy to design and implement national social security programmes. Building a national system is still a huge feat of organisation. Apart from building the required political support, and making fundamental high-level design decisions, it depends on large-scale information gathering and processing as well as requiring significant capacity on the ground. We all know this, but we can forget that this makes it important not to over-complicate things. The evidence shows relative simplicity – for example by using universal selection – can also help with accountability and political support.
Do-ability and longevity are the top priorities for governments in designing social security systems, and should be for everyone. In countries I have spent a lot of time working in – such as Nepal, Kenya and Uganda – the largest social security programmes, either already in place or on the way to being introduced, are government-run, government-financed, simply targeted and reach (or will reach) national coverage. Do-ability and longevity have been prioritised.
Things are changing, but not fast enough
There are some positive signs of change. For example, there is a general move in international development towards Adaptive Management, Doing Development Differently, Thinking and Working Politically etc. All of this will – and should – guide us to being better-grounded in practical and political realities. And, in the world of social security, there is a greater acceptance worldwide – and an ambition to achieve – universal programmes and universal systems. In fact, a recent World Bank book on social security talks of the importance of the political, institutional and financial considerations and asserts that “ignoring these spheres may lead to technically sound, but practically impossible, choices and designs”. These are all positive signs.
But they are shifts that have not yet been reflected on the ground. There is still a fundamental imbalance in donor approaches. For example, why does no major donor support a national disability benefit in developing countries? This is a core part of social security systems in any country. And, how many donors support universal programmes of any kind on the ground? A change of approach requires donors to spend more time than they currently do supporting governments in sorting out the basics, the system plumbing. It is about prioritising – even celebrating – the more mundane, and making sure we stay within political and practical realities. This, perhaps, requires patience and the humility to understand that we are small cogs in a big wheel. But building on solid foundations is what will deliver real impact in the long-term. Let’s keep searching for innovation, for delivering things more efficiently and effectively. But not at the expense of core system-building. Let’s keep innovation in perspective. In terms of the fundamentals, we already know what works.