Twenty years ago this month, ministers in the newly elected Government were steaming ahead with the development of radical proposals for devolution.

Spurred on by a young and enthusiastic Prime Minister, their efforts persuaded Parliament to enact legislation that brought about significant and lasting constitutional change.

Historians of the early Blair years highlight the importance of the Scotland Act and the Government of Wales Act, both passed in 1998. Given they resulted, in 1999, in the establishment of the devolved Scottish Parliament and the National Assembly for Wales, these Acts deserve to be celebrated. A significant amount of power was shifted from Whitehall to Edinburgh and Cardiff in short order.

At the same time, they pushed forward radical reform of city region governance in London. In the face of bitter opposition from some local government traditionalists, Prime Minister Blair orchestrated the creation of an entirely new kind of metropolitan governance for the capital.

Ministers in the Department of Environment, Transport and the Regions (DETR), a forerunner of the present Department for Communities and Local Government, deserve credit for their bold thinking.

At the time, I was an academic adviser on local government to ministers and I was struck by their genuine interest in learning from abroad, and by their willingness not just to entertain radical ideas, but also to knuckle down and develop solid, deliverable reforms.

Their intellectual curiosity led to proposals designed to create an entirely new kind of strategic metropolitan authority for Greater London, one headed by a London-wide directly elected mayor. Few would claim that the Greater London Authority (GLA), established in 2000, is a perfect model of metropolitan governance.

But the GLA was, from the get-go, a breakthrough in developing strong, effective, transparent and accountable city region governance. 

Last year the Local Government Association (LGA), impatient with the piecemeal and unprincipled approach to devolution being pursued by the Government, invited me to conduct a short study designed to widen the public conversation about devolution in England. 

The brief asked me to provide an international review of different models of sub-national governance, to assess them according to explicit principles of good governance, and to draw out lessons for councils and others involved in developing devolved governance arrangements.

Presented at the LGA Conference last year my report, English Devolution: Learning lessons from international models of sub-national governance, benefited significantly from inputs from senior city leaders in England. 

They helped me develop the six criteria I then used to assess the quality of sub-national governance in other countries:

• Civic leadership
• Effective decision-making
• Transparency and efficiency
• Accountability
• Public involvement
• Business engagement

Four highly respected examples of metropolitan governance emerged from the analysis: Auckland Council, New Zealand; Greater London Authority, UK; Portland Metro, Oregon, USA; and the Association of the Region of Stuttgart, Germany.  These city regions are chosen to illustrate a range of design options for devolved city region governance.

What pointers emerge for today's debates about English devolution? 

First, the Greater London Authority model of city region governance came out well in the international evaluation.  Involving direct election of a metropolitan mayor and, just as important, a London Assembly comprising 25 members, to represent diverse communities in London and to hold the executive to account, the Greater London model of sub-national governance is highly respected internationally.

Second, the research shows that it is perfectly possible to have excellent sub-national governance without introducing directly elected metro-mayors. Two of the four highly respected models of sub-national governance featured in my report – Portland Metro, Oregon, and the Association of the Region of Stuttgart – eschew the directly elected mayor model. 

It follows that the Government's insistence on introducing directly elected mayors for city regions, or for that matter other sub-national territories, represents an evidence-free policy stance.

Third, my international comparison shows that the combined authority model of metro-mayor governance is wholly inadequate. It fails to perform well on many of the criteria.

What to do? 

It is clear that devolution policy in England needs a radical rethink. International experience, and the relative success of the GLA model of city region governance, suggests that effective devolution must pass three litmus tests.

First, the new units of sub-national governance must have substantial tax-raising powers. The politicians elected to these authorities need to have the fiscal power to make a difference to the quality of life of the people in the area they govern and they need to be held to account for their performance. Metro-mayors without significant tax-raising powers are bound to be featherweight mayors. 

Second, any sound model of city region governance requires a strong elected assembly as well as a strong executive.

Third, the international evidence shows that locally elected politicians can enhance the quality of life of local people when they have the political space to operate as powerful local leaders. 

It follows that the key challenge for devolution policy is to give place-based power a really significant boost so that local leaders have the authority to work with other stakeholders to shape the fortunes of local people.

Robin Hambleton is emeritus professor of city leadership at the University of the West of England, and director of Urban Answers