Failed State: Devolution vs Decentralisation
This article is prompted by reading Sam Freedman’s book Failed State, which has attracted considerable attention among the chattering classes.[1] I had thought to write a review of the book because I agree with a lot of his diagnosis but, equally, I believe that he is profoundly wrong as soon as he attempts to go beyond pure description. I will write that review in due course. As a start I think that a detailed examination of key issues may be more interesting.
Freedman argues that the UK is a vastly over-centralised polity. This is hardly a novel idea. The same point has been made repeatedly for several decades. And yet most policies during that period have increased centralisation. Even worse, Freedman’s model of decentralisation is devolution to Scotland and Wales and the creation of city or regional mayors.
I cannot think of anything less likely to promote the potential benefits of decentralisation. My father was Scottish (from the East Neuk of Fife). I have lived in Scotland for nearly 4 decades, while I have family who live or spend a lot of time in Wales. Few outside the Central Belt in Scotland or South Wales believe that devolution has done anything to improve their life. By almost all external indicators the performance of the Scottish and Welsh Governments has been dismal, especially in education and healthcare. Their administrations are inept and highly centralised with minimal recognition of local initiatives or local differences. Not coincidentally, there were strong movements in Orkney and Shetland and in the Borders to leave Scotland had the SNP’s Scottish Independence referendum succeeded.
Regional devolution with city or regional mayors is the sort of thing that think tanks in London (or Edinburgh) promote, but this has little to do with genuine decentralisation. I can write about this from both an academic and a personal perspective. Over more than 2 decades I wrote many articles on what economists call fiscal decentralisation and the economics of local government.[2] I have also advised several countries, notably in Latin America, on these issues. I have lived for extended periods in the US and, more recently, have spent a lot of time in Italy and have family living there. I will use Italy as my main point of comparison in this article. The lessons would be similar had I used France or Germany as my comparator.
Italy has only been a unified country since the 1860s. Before that the dominant powers since the mid-1500s were the Austrians in the North and Spain in the South, while the Papacy governed the Papal States in Central Italy and tried to weave a course between Austrian, French and Spanish influence. With limited or no central authority, government was predominantly local with powers exercised by the communi (communes or municipalities) though larger territories such as Milan/Lombardy or Tuscany might exercise regional powers.
That structure remains today. We live in the commune of Bellagio in the province of Como. It has a population of about 5,000 following a recent merger with a neighbouring commune. Bellagio covers an extensive rural hinterland going up into the mountains in the centre of the peninsula that has Bellagio at its tip. The commune is responsible for a range of local services from maintaining roads, collecting rubbish, providing water supply, operating libraries and cultural facilities, overseeing planning decisions, local police, etc. The crucial point is that it works! Local services are generally good, and facilities are well maintained. Responsibility for pre-school and school education is shared by the commune (up to secondary school) and the province (Como). The region (Lombardy) is responsible for health services, transport and tertiary education, though again it works with lower tier authorities. The commune is financed by a property tax with block grants from the provincial or regional government and charges for specific services – water, rubbish collection, etc.
This arrangement is very similar to the way in which local district or town councils used to operate in the UK. I grew up on the edge of the Cotswolds, in Warwickshire near to Stratford-on-Avon. The rural district council was similar in size, population and operation to Bellagio. That all changed in 1974 when several small councils were merged into the much larger Stratford-on-Avon District Council, and further as core services were transferred to Warwickshire as the upper tier authority.[3]
The story is similar where I live in Scotland, which falls within the historic county of Peeblesshire. The county had a population of under 15,000 and covered a large area, mostly with a very low population density. The council understood the issues of providing local services in very rural areas. In 1975, the county was transformed into the lower tier Tweeddale District Council with the Borders Region as the upper tier authority. Finally, in 1996 the lower tier districts were abolished and the regional authority was transformed into a unitary authority - the Scottish Borders Council.
Most of the functions previously carried out by Tweeddale District Council have been centralised and are operated out of offices more than 60 km from where we live. One consequence is that local control over the kind of services provided by Bellagio commune has disappeared and, sadly, most of those services have been run down drastically in rural areas.
The point of these comparisons is that there is no longer any tier of truly local government in the UK in the sense that this is understood in Europe, the USA and Latin America. Unitary authorities like the Scottish Borders Council are equivalent to provinces such as Como in Italy, but they are more centralised and focused on the largest towns within their areas. Even district councils in England, such as Stratford-on-Avon District Council, cover populations similar to the Scottish Borders Council – over 120,000 – and have become highly centralised.
There are two consistent themes which underpin the multiple reorganisations of local authorities in the UK over the last 5 decades. The first is the doctrine that larger (in population terms) is better when comes to managing the provision of local public services as well as carrying out the statutory responsibilities imposed by central government on local authorities. This reflects an assumption that scale is more important than heterogeneity in managing local government. That is not the basis on which local government operates in Italy and elsewhere. The second theme is that central government will not tolerate local government units that are large enough to challenge it. The ideal is units that are large enough to do things efficiently but small enough to do what they are told.
Such complaints may mean little to those for whom London, Manchester or Glasgow are their whole world. That ignorance is reflected in the policies formulated and implemented by people with no experience of or sympathy for the substantial portion of the population who live outside large cities or towns and whose quality of life has been eroded by the mindless pursuit of “efficiencies” in providing local public services. Much as I value the quality of life in our part of the Scottish Borders, there is no doubt that our public services in Bellagio are far better than those in Scotland. In day-to-day terms it is like comparing chalk and cheese.
Freedman attaches large weight in explaining the extension of central controls over local government to Mrs Thatcher’s dislike of metropolitan authorities as competing centres of power. This is an odd claim bearing in mind his general emphasis on systemic rather than personal explanations. I was writing about the issue at the time, he was not. The decisive changes took place in the 1970s with (i) the reorganizations of local government referred to earlier, and (ii) a change in how central government grants were allocated, which established redistributive mechanisms that gradually eliminated local control over local spending.
These were changes implemented by Labour and Conservative governments. No matter what Mrs Thatcher thought of Ken Livingstone et al, there was a consensus in Westminster on exercising greater control over local government. The reason was that policymakers saw local government not primarily as a provider of local services but as an agent of central government, exercising powers and spending money in ways that focused on redistribution. Under all governments over the last 50 years, any notion of encouraging fiscal or operational autonomy for local governments has been firmly suppressed.
The Scottish and Welsh Governments have followed that approach. The Scottish Government, in part because its powers are greater, has been particularly unfriendly to local authorities, which are all unitary authorities. Even though the Scottish Government is the product of legislation that was claimed to devolve powers away from Westminster, the reality is that the gap between local populations and the providers of public services has become much wider over the last two decades.
This experience highlights the general problem of devolution in the UK. It comes over as an urban-focused argument for employing second or, more often, third rate politicians whose main role is PR and whose administrative capacity is often negligible. A related issue is that too many devolved governments in the UK are one party states. Local administrations are elected for reasons that have almost nothing to do with their performance in administering local services.
Of course, this is a chicken and egg problem. If local governments have no autonomy, why would any voter bother to vote according to who they think will manage local services best? On the other side, why would anyone agree to give more power to a Scottish Government that appears to be incapable of managing local services and implementing policies in an even vaguely competent manner?
To return to Italy. Few Italians have any trust in the competence of the national government in Rome. Italian bureaucracy, driven by legislation adopted in Rome, is notoriously obscure and baroque. The stories about Commissario Brunetti in Venice by Donna Leon are not just fiction but slightly exaggerated descriptions of the daily life of the bureaucracy. Even so, the North of Italy is richer than the UK and most people outside large cities have a high quality of life, which is underpinned by a reasonable provision of local public services. Political parties are not noticeably visible in campaigns for commune posts, except in the largest communes. There are notorious problems in some large cities, but Milan and Turin are not Naples and they function relatively well. Political and administrative culture matter a lot.
Of course, Northern Italy is not a nirvana. Many locals complain about things, though the idea of living in Scotland – rather than just visiting – is not well received! Inevitably, having 4 tiers of government from Rome down to Bellagio involves additional costs, though these are not particularly large. Most of the functions must be managed by someone at some tier of government. As percentages of GDP, public spending and taxes are higher in Italy than in the UK, but the UK is heading quite fast in the Italian direction.
The UK is the outlier when considering the management of local public services. Even a (formerly) highly centralised country like France has a tier of municipalities with functions and powers that are equivalent to the communes in Italy. This is where the central problem of decentralisation in the UK lies. How can a tier of municipalities be restored when most of the focus over the last 50 years has been on centralisation and, in practical terms, the elimination of local control over all public activities? Without such a step all talk of decentralisation is simply window dressing.
The problem with the forms of devolution espoused by Freedman and others is that the authorities thereby created have few useful functions. Consider the example of the Tees Valley Combined Authority (TVCA) which was widely publicised by the previous government because of its self-promoting Mayor, Ben Houchen. It covers 5 unitary authorities with a population of about 700,000. Its responsibilities are claimed to be “strategic” but are, in practice, limited to transport and economic development.
Comparing this with Italy, the TCVA is equivalent to regional authorities, smaller than most, but between Basilicata (0.6 million) and Umbria (0.9 million) in population. The competences of TCVA are much more limited than those of Italian regions. Crucially they play a central role in the provision of healthcare services and overseeing many other state functions. Overall, Italian regions are much closer to US states than to UK strategic authorities.
However, we should note that even in Italy there is a tug-of-war between centralisation and decentralisation. Under pressure from the Lega Nord (representing the richest parts of Italy) the current government has passed legislation to devolve substantial fiscal autonomy to regions. Less than 10 years ago the Democratic Left government proposed a constitutional amendment that would have abolished provinces and transferred most of their functions to the regions. This was rejected in a referendum. On the other hand, there has never been a serious proposal to abolish communes and transfer their power to higher tier authorities.[4] There may be disputes about how responsibilities should be divided between central, regional and provincial governments, but the core role of the lowest tier is not a matter of dispute.
Real decentralisation in the UK requires two changes and it is hard to see either happening except in a crisis. The first is to create the equivalent of municipalities by, at a minimum, separating the provision of local services from unitary authorities, thus reversing the mergers of local authorities in England and Wales. The second is to split up the National Health Service and transfer responsibility for managing healthcare services to regional authorities. This is not as difficult as it might seem since there are already entities such as health boards (Scotland), local health boards (Wales), and integrated care boards (England) which might provide the basis for forming regional authorities.
Ultimately the details are less important than the general principles. Without question the UK needs to decentralise the way in which its government functions. But devolution as currently envisaged is simply not the right answer. To make a genuine difference, any decentralisation strategy must transfer responsibility for local public services to much smaller lower tier authorities equivalent to communes in Italy or municipalities in many other European countries. Such entities should obtain most of their revenues from a local property tax and service charges. That will be a difficult proposition to enact and implement, but without that change every other option will be tinkering at the edges.
Samuel Johnson wrote “When a man is tired of London, he is tired of life, for there is in London all that life can afford”. I am sure that one could find a similar statement about Paris. However, certainly no Italian would say that of Rome and few Germans would say that of Berlin. Even in the 18th century, many in England profoundly disagreed with Johnson and even more do today. That is the challenge posed by decentralisation. How can we get rid of the assumption that Nanny in London always knows best? In truth, it knows little and understands less. We should encourage our Italian side and accept that a large dose of local heterogeneity and autonomy is a strength rather than a defect.
[1] Sam Freedman – Failed State, London: Macmillan, 2024.
[2] See: ‘Fiscal federalism in the UK’, Oxford Review of Economic Policy, Volume 3, May 1987.
For empirical data on fiscal decentralization across OECD countries, see: ‘Economic aspects of decentralised government: structure, functions and finance’, Economic Policy, No 13, October 1991 (with Stephen Smith).
[3] A brief sketch for those unfamiliar with the terminology of local government structure in the UK. Prior to 1974 the lowest tier of statutory local authority in England and Wales outside metropolitan areas consisted of town, borough or rural district councils. These were primarily responsible for providing what I refer to as local public services. These were merged into larger district councils. The tier above them consisted of the historic county councils whose main responsibility was education but by extension they acquired more general responsibilities for children’s and other social services as well as various statutory responsibilities. To complicate matters there were some historic county boroughs.
In 1974 smaller counties were merged into larger units – e.g. the new county of Cumbria was formed from the old counties of Carlisle, Cumberland, and Westmorland plus some areas that were previously in Lancashire and the West Riding of Yorkshire. At both tiers the goal of mergers was to create authorities that were thought to be large enough (in population terms) to realise economies of scale so as to discharge their responsibilities efficiently. That belief in administrative economies of scale subsequently underpinned central government pressure to form unitary authorities which combine district and county level responsibilities, thus combining two tiers of local authority into a single tier.
From 1974 the arrangements in London and other metropolitan areas were conceptually similar with metropolitan boroughs instead of districts and metropolitan counties. The latter were abolished in 1986 so that metropolitan boroughs were converted into unitary authorities.
Arrangements in Scotland were more complicated. Historically, outside the 4 large cities, large towns had burgh councils while everywhere else had county councils. Both were effectively unitary authorities. These were reorganised in 1975 into a two-tier system of district and regional councils. The regions were hugely different in size and were dominated by Strathclyde which accounted for nearly 50% of Scotland’s population. Finally, for now, in 1996 the regional authorities were either split up or transformed into single tier unitary authorities.
[4] To be clear, central and regional governments provide incentives and put pressure on communes to merge to form larger and, it is hoped, more efficient units. That is how the current commune of Bellagio was formed, but that is a matter of forming small units from micro units, not removing the lower tier of decentralised government.
The Birmingham City Council case was greatly exacerbated by disastrous financial management by the council. Other councils have dealt with the consequences of this ruling rather better. Remember too this was not the consequence of central direction but a (probably unintended) consequence of poorly drafted legislation - what is work of equal worth?
Part of the problem is that large authorities think they are too big to fail. That is one of my objections to the usual ideas of devolution - think of the behaviour of the GLA and the Mayor of London with respect to Transport for London - first lower fares and then ask to be bailed out.
Some large Italian communes are equally bad. A number of metropolitan areas are classified as metropolitan cities which function like provinces. But there is usually one large commune for the central city - e.g. the communes of Naples and Milan. Milan is relatively well run, whereas Naples is another story with endless political drama involving the mayor, unions, the Mafia (colloquially), etc. The centre has little or no power over communes. This is exercised by either regions or provinces. I have never heard of the kind of central direction of communes that is a regular occurrence in the UK.
Your example is a good one. Many Italians from the North regard Rome (and the South) as something they have to bear. That is why the Lega Nord has been so strong in the regions of the North, though the current Fratelli government is reducing its support. On the other hand, it is important to remember that regional differences in Italy are very large. A lot of the resentment in the North focused on what were regarded as corrupt and inefficient mechanisms to level up in the British terminology.
Your point about disease control is a reflection of the fact that staff in municipalities are locals. They know their neighbours. Voters see them daily at the local store or bar so there is immediate feedback on what works and what doesn't. For some people this is all rather intrusive - villages in which everyone knows everyone else's business. The balance is difficult but, in my experience, having a local authority where most staff work 60 km away from the areas they are supposed to serve is far worse. Anonymity is all very well but in practice it means that public services are mostly blind.