The familiar old debates about centralised versus decentralised public services are already emerging ahead of the next general election. Yet we lack any meaningful evidence of what’s best done centrally (once) and what locally (many times over).

If the pattern of previous government re-organisations is any indicator, each new localised organisation will duplicate leadership teams, management hierarchies, administrators, clerks, technology, systems, accountants, buildings etc. Such duplication pointlessly wastes limited and precious public resources, depleting the very frontline services that most matter to people.

So how do we get the democratic upside – more locally accountable and responsive services – without also eroding those very same services by burdening the public purse with all the waste inherent in massively duplicated overheads?

This is where the practices of modern, digital organisations and platform-based systems come into play. Whilst modern digital organisations are enabled by technology, their real significance lies in a much more important cultural shift away from functional business units. They have restructured around users’ needs and the processes that best support their services — building entirely new types of organisations and services on the back of modern technology.

To start with a prescriptive one-size-fits-all model — “Central is best”, “Local is best” — is to start in entirely the wrong place. Organisations can be highly efficient at small, medium and large scale — if their purpose is well understood and they are well designed. This is why digital platforms are an integral part of modern organisations. They provide a set of flexible building blocks that can be used to compose functions, processes and services in cost-effective ways that better meet users’ needs.

To realise these same benefits, government needs to mature its approach of the last 20 years. Technology has been applied repeatedly to refurbish the visible part of its operations — online transactional services. Meanwhile reform of the underlying invisible iceberg of creaking processes, functions, roles, technology and organisations that sit behind them has barely started.

These "invisible” elements remain the most significant challenge to the improvement of our public services. They drain time and effort from the frontline, burdening frontline workers’ and distracting their attention away from where it matters into massaging the public sector’s self-serving internal overheads instead.

These problems will only get worse if the pendulum arbitrarily swings once again between centralised and localised organisations. Understanding why and where the right approach makes sense is essential to this process: why this particular function or process is best run once centrally as a common platform service shared by all, or why this one must be local, meeting a very specific community need.

An understanding of these opportunities, and what “Government as a platform” truly means, is essential for any politician wishing to achieve the right balance of what is best done centrally and what locally. Moving to government as a truly digital organisation, drawing upon platform-based, componentised services, is essential for the relevance, timeliness and efficiency of our future public services. Its merits, significance and implications need to be far better understood well ahead of the general election.