A new geography of Britain? Analects for a non-autobiography.

Although it would be neither quick nor easy to achieve, public policy should try to emulate in the North of England the evolution of the South East 'mega-city region' by finding ways of encouraging non-core northern towns to develop a 'new economy' base, first as 'information satellites' within

AUTOBIOGRAPHY is a literary form best left to waning politicians, or to ghost writers labouring on behalf of those short-lived celebs who achieve instant fame for being famous, and lose it soon after. Any self-respecting academic or planning practitioner should resist the temptation, with a small but significant exception for those enlivening their sunset years with contributions to planning history. But there does come an occasional point in any academic career where it can prove useful, for the reader as for the writer, to describe an intellectual journey: a chronicle of terrain recently travelled, and of a dimly-perceived landscape that is just opening up ahead. That's the point at which this ageing academic-practitioner finds himself in autumn 2006. So, no further apology for the occasional personal pronoun.

The terrain behind, stretching back three long and eventful years, extends over a large part of Europe. In May this year we published the results of a major research project: POLYNET. (1) With an astonishingly generous grant of

2.4 million from the European Commission and our own ODPM/DCLG (Office of the Deputy Prime Minister/ Department for Communities and Local Government), eight research teams from seven European countries have worked together to try to understand an urban phenomenon we call the polycentric 'mega-city region'. We started, as we concluded, believing that it is the fundamental urban reality of the early 21st century.

First recognised in Asia, in such regions as the Pearl River Delta and Yangtze River Delta of China, it has its European counterparts in such areas as the Dutch Randstad, the Rhine-Ruhr and Rhine-Main regions of Germany, and the region that stretches from Basel to Zurich in northern Switzerland. And, most significant because it is largest and most complex of all, South East England: not the curious L-shaped region recognised in government policy or official statistics, but a vast area stretching some 100 miles from London, as far out as Bournemouth, Swindon, Northampton, Peterborough and Colchester, and numbering nearly 19 million people--one-third of the population of England, on about one-sixth of its area.

The essence of such mega-regions is that they are polycentric: they consist of a whole series of cities and towns and their surrounding spheres of influence--no less than 51, in the case of South East England--which in significant ways are functionally interconnected. That is the clue to their success and their continued growth.

It is no accident at all that the Pearl and Yangtze Deltas are the great workshops of the entire early 21st-century world, in exactly the same way as Lancashire was 200 years ago in the first Industrial Revolution. It is likewise no accident that the San Francisco Bay Area--recently recognised in American research as one of ten such regions in the United States, together housing 197 million people or nearly 70 per cent of the entire population (2)--retains its unique position as the outstanding technologically-innovative region in the world. All are examples of complex networked production, dependent fundamentally on the rapid generation and exchange of information in an economy that is based on knowledge.

They are successful because they permit the principle of clustering--incessantly discussed in the academic literature of innovative production--to occur almost without limit, through a principle that Dutch planners described 30 years ago: concentrated deconcentration. They grow and thrive through relatively faster growth--in terms of population and production--of their smaller and more peripheral members, while maintaining sufficient local concentration, coupled with excellent infrastructure for personal and electronic exchange of information between all of them.

In the mega-city regions of China, as in late 18th-century Lancashire, this principle is harnessed to the production of goods. In Silicon Valley, it is harnessed to research and development that increasingly results in the production of immaterial goods: software and the information it contains. In the mega-city regions of Europe, it is increasingly and exclusively harnessed to the generation and exchange of knowledge in the advanced producer service (APS) industries that are the economic drivers of the post-industrial Western economies: finance and business services, law and accountancy, advertising and public relations, print and electronic media, and design services like architecture and civil engineering. The key to understanding these regions is to probe the flows of information within and between these APS firms and the places where they locate.

But this is easier said than done. The amount of information is so vast as to be beyond measurement. Even if sampled, it flows in ways that are difficult to capture. Phone calls and e-mails are subject to data protection and commercial confidentiality. It is impossible to capture the information that flows in a meeting without being there, or intercepting it. Most fundamental, even if it were possible to capture a fraction of the information, there is the critical question of its quality. The old market stall adage 'Never mind the quality, feel the weight' is as misleading here as in its original context.

In the POLYNET study, we sought to grapple with these problems in at least three different ways.

First, we drew on the expertise and accumulated wisdom of Peter Taylor and his colleagues in the Globalization and World Cities team at the University of Loughborough, who had faced the same challenge in developing their global city hierarchy. (3) They use internal structures of large multi-locational service firms, like banks and accountants, to serve as some kind of indirect proxy for information flows. Second, we sought directly to measure the most important flows, by asking a sample of CEOs (chief executive officers) and senior managers in APS firms to record their most significant phone and e-mail exchanges and their business trips. This, despite elaborate preparation, yielded thin results. Third, we did succeed in interviewing substantial samples of these same CEOs and managers, allowing us to build up a valuable qualitative picture of the way they themselves saw these relationships.

The results were surprising: at least, they surprised us, many of us battle-hardened research veterans. A region like the Randstad Holland could appear polycentric, and yet prove far less polycentric in the way it actually functioned, with the APS information flows strongly concentrated on just one city, Amsterdam; likewise with Frankfurt in Rhine-Main, or Zurich in relation to northern Switzerland. Conversely, South East England, which at first glance seemed quite monocentric--with half the people and half the employment concentrated in the London city-region--could emerge as quite polycentric in terms of information flows, with places like Reading and Cambridge and Milton Keynes showing a degree of independence from the capital.

Indeed, as we in the UK team reflected on the phenomenon, we found a central paradox: the more successful London has become as a global city of the first rank--and the 20th anniversary of the financial services 'Big Bang' has just allowed us all to reflect on how phenomenally successful it has become--the more it can export some of that success out into the 50 other cities and towns, many of them quite small, that constitute its hinterland, eventually allowing at least some of them to function in a semi-independent fashion in the modern knowledge economy. In other words: the greater the global success of the leading city in such a region, the greater its power to irrigate the other places in that region.

At the end of the POLYNET study, all that Euro-money spent, we're not even half way in understanding how precisely this operates; we have merely the outlines of an understanding. But it is enough to get on with. Thanks to a windfall gain in the form of a big research prize, (4) I'm going to be able to test it through grants to young researchers starting on their PhDs. We'll extend our research from the South East--the golden corner of the English and the UK economy--to the rest.

The basic hypothesis

For the South East, the basic starting hypothesis reads like this. Those other places--most of them--were always service-oriented: they were county towns (Oxford, Reading, Winchester, Maidstone, Chelmsford, Cambridge, Bedford, Aylesbury, Northampton) and/or university towns; they therefore developed a strong middle-class population with high levels of general education and special qualifications. Then, this character was further intensified as they developed as longer-distance commuter centres for the London knowledge-based economy in the period 1951-1981, ironically encouraged by green belt controls which restricted development closer to London.

Then, from about 1970, they began to attract decentralised back-office activity from London. In addition, some places attracted European headquarters offices of global corporations. Over time, progressively, they became information-generating centres in their own tight, albeit with strong continuing links to and from London. And this aided the formation of many small subsidiary functions, like legal and accountancy services, thus producing multiplier effects in the local economies.

We would therefore expect to find that over time these 50 centres have become first less and then more independent of London: starting as independent market towns with a strong service base and local agriculturally-based industries, they first became more dependent on commuting to London, but then (after about 1971) less so as they developed their own bases as centres for advanced producer services--as evidenced by a strong growth of these services, by increasing self-containment, and by increasing non-London commuting ties. Further, we would expect such patterns to start to develop in the adjacent East Midlands and South West regions.

But the big research puzzle concerns the rest of Britain. Here, too, we can immediately see what appear to be mega-city regions: in the West Midlands, North West England, Yorkshire, the North East, South Wales, and Central Scotland. Indeed, Patrick Geddes recognised them as such in his pioneering Cities in Evolution, 90 years ago; (5) the Northern Way initiative has rediscovered them and elevated them into an instrument of policy.

But they have a very different genesis and seem to be developing in a very different way. They first developed as networked systems of production in the first Industrial Revolution. Essentially a central city (Manchester, Leeds, Sheffield) acted as commercial core for a surrounding core of manufacturing towns (Liverpool is a partial exception). But then, in the great de-industrialisation of 1970-1990, the manufacturing base (textiles, engineering, iron and steel) collapsed.

In the commercial core cities this has been slowly and partially replaced by an alternative 'new economy' base of advanced producer services plus consumer services, including a strong public service element (higher education, health, media and culture) but also private services (commerce, tourism). But, with few exceptions, the surrounding industrial towns have been unable to follow suit. As a result, the Midlands and North display what the French urban analyst Pierre Veltz calls an archipelago economy: (6) islands of economic light separated by vast seas of economic darkness.

The evidence from a previous study we made (7) suggests that while the 'core cities'--Birmingham, Bristol, Leeds, Liverpool Manchester, Newcastle upon Tyne, Nottingham, Sheffield--held their own in the urban hierarchy, many of these other towns slipped between 1965 and 1998. If the hypothesis is right, it will be evidenced by divergence in the economic structures of the core cities and the other towns between 1971 and 2001, coupled with weak commuting ties to the core cities. In other words, we will fail to find the critical evidence that these regions are replicating the virtuous circle of deconcentration and reconcentration that proved to be so critical for the South East.

There is some preliminary evidence for this pessimistic view in Tables 1 and 2. They contrast the recent records for the core cities and for a large sample of northern non-core cities in the mega-city regions around them, basically old one-industry or two-industry towns that have suffered drastic de-industrialisation--including one, Blackpool, whose basic industry was and is tourism. Table 1 presents selected indices of deprivation from the 2001 Census; Table 2 aggregates the data to generate two basic sets of measures:

* Regeneration need--evidence of social deprivation arising from economic causes (loss of employment in basic industries; failure to compensate by developing 'new sector' jobs): Here, there are a great variety of social data available from the 2001 Census and conveniently assembled in the State of the English Cities 8 database. However, the key deprivation indicators have been synthesised in the Indices of Multiple Deprivation developed by ODPM/DCLG and last updated in 2004 (and published as The English Indices of Deprivation 2004 (9)).

* Regeneration achievement--evidence of economic regeneration already occurring, which could potentially help to reduce the resulting social deprivation without the need for external (exogenous) intervention: Essentially, what we need to analyse here is the lack of regeneration achievement: in other words, the failure so far of a city or town to find ways of meeting its regeneration need. Here, we use two sources: first, our earlier analysis of shifts in the English and Welsh urban hierarchy; and second, the State of the English Cities database, which has useful dynamic indices of changes in urban economies over the period 1995-2005. (10)

In Table 2, the core cities display great variations. Three--Liverpool, Manchester and Nottingham--are among the seven most deprived local authority areas in England overall. Birmingham stands at 15th rank and Newcastle at 20th place, but the three remaining core cities--Sheffield, Bristol and Leeds--are evidently much less deprived, at 60th, 67th and 68th place, respectively. But what is most notable about these cities is their relatively strong record on the regeneration achievement indices, particularly in GVA (gross value added) per FTE (full-time equivalent) employee and (Birmingham and Manchester excepted) in unemployment reduction.

Longer-term, most maintained their rank in the urban hierarchy over the 1965-1998 period; only two, Liverpool and Sheffield, dropped one place, while one, Bristol, advanced one notch. This confirms the general impression of strong urban economic renaissance in these cities, led by their central business cores, in the last decade, symbolised by the 'crane count' of physical regeneration which is so evident in them.

The northern non-core cities likewise display considerable variations--and might have displayed even more, ff several had not been eliminated because of data aggregation problems. But nine out of 14 display high overall indices of regeneration need, with rankings between nine and 40 out of 354, and--even more significantly for the basic hypothesis--their regeneration achievement indices are lower than those of the core cities:

* Most maintained their position in the urban hierarchy over 1965-1998, but three--Middlesbrough, Doncaster and Hull--lost a position, while two--Preston and Wakefield, both administrative county towns--gained one notch.

* All but two record GVA/VFE growth of less than 20 per cent (although several--denoted by an asterisk--are aggregated).

* Employment growth is also weak, and only on the reduction of unemployment do they show strong results, evidently because of high initial unemployment rates.

Among the seaside resorts, Blackpool emerges as the 24th most deprived local authority area in England; it has slipped in the urban hierarchy, and it has doubled its unemployment rates, in contra-distinction to almost every other location in the table. There is a striking contrast here with Brighton and Bournemouth, not shown in the table, which have shown strong improvements on these indices.

Emulating the South East's evolution of a 'new economy' base

The conclusion is that public policy should try to emulate in the North the evolution of the South East urban system over the last half-century, finding ways of encouraging these towns, like their southern equivalents, to develop a 'new economy' base. This could include encouraging them to develop first as commuter towns for the core cities, on the South East England pattern, with a local economic base in consumer services, before they build an independent producer service base--first as 'information satellites' within a regional hierarchy, and then evolving into semi-independent units within a functionally polycentric system.

Policy should therefore establish a similar infrastructure of strong public services in the medium-sized towns of the North, building a middle-class population and an appropriate labour force for the knowledge economy, and it should enhance transport links so as to encourage them to develop as commuter towns for the core cities before they evolve as significant employment centres in their own right. On this basis, the development of a successful core city like Manchester or Leeds might in time also be followed by the evolution of a similar polycentric region. This, of course, is the essence of the Northern Way city-region strategy. Over time, a place like Rochdale might become a place like Bedford; Blackpool might become a northern Brighton.

Such a policy would be far from easy to achieve: half a century ago, the towns of the South East had far more going for them than their northern equivalents have today, in the form of a social and cultural legacy that went back deep into the geography of pre-industrial England. The towns around the northern core cities are unlikely to develop similarly, because their labour forces developed to perform a manufacturing role that has now disappeared, and because deep cultural assumptions are often hostile to education.

So it would be neither quick nor easy: it might take a generation to achieve. But it may be the only policy on offer. Otherwise, the stark alternative might be to recognise that such places are a legacy of the first Industrial Revolution--a short and anomalous period in history--whose time has come and gone. But such an assumption may be more than any policy-maker could bear to contemplate.

Notes

(1) P. Hall and K. Pain: The Polycentric Metropolis: Learning from Mega-City Regions in Europe. Earthscan, London, 2006. Further results are reported in a special issue--'Polycentric Development across Europe'--of Built Environment, 2006, Vol. 31 (3), and in a forthcoming special issue of Regional Studies, 2007

(2) R.E. Long and D. Dhavale: 'America's megapolitan areas'. Land Lines: Newsletter of the Lincoln Institute of Land Policy, 2005, Vol. 17 (3), Jul., pp.1-4

(3) P.J. Taylor: World City Network: A Global Urban Analysis. Routledge, London, 2004

(4) The Balzan Foundation's 2005 International Prize for research on the social and cultural history of cities since the beginning of the 16th century

(5) P. Geddes: Cities in Evolution. Williams and Norgate, London, 1915, reprinted in R. LeGates and F. Stout (Eds): Early Urban Planning 1870-1940. Vol. 4. Routledge, London, 1988

(6) P. Veltz: Mondalisation Villes et Territoires: L'economie d'archipel Presses Universitaires de France, Paris, 2000. Third Edition

(7) P. Hall, S. Marshall and M. Lowe: 'The changing urban hierarchy in England and Wales, 1913-1998'. Regional Studies, 2001, Vol. 35, pp.775-807

(8) State of the English Cities. Office of the Deputy Prime Minister, London. Mar. 2006. Vols 1 & 2

(9) The English Indices of Deprivation 2004. Office of the Deputy Prime Minister, London, 2004

(10) These do not refer to cities and towns as generally understood, but to 'primary urban areas' (PUAs) including neighhouring local authority areas in some cases; one critical measure (GVA per FTE worker) is available only for EUROSTAT NUTS 3 regions, which are much larger than local authority areas and in some cases aggregate many such areas, making it impossible to distinguish between them

Sir Peter Hall is Professor of Planning and Regeneration in the Bartlett School of Planning, University College London, and President of the TCPA. The views expressed here are his own. Tables 1 and 2 were compiled as part of a tidier analysis submitted by ReBlackpool URC to the Casino Advisory Panel, which may be viewed on the CAP website.

Table 1 Core cities and non-core cities--measures of deprivation

                     Households        General     Unemployed,
                       without         health        percent
                       a car,        'not good',
                      per cent         percent

Core cities
  Newcastle
    upon Tyne           45.2            11.8           4.7
  Leeds                 34.5             9.8           3.3
  Sheffield             35.7            11.3           4.2
  Liverpool             48.3            13.8           6.0
  Manchester            47.8            12.5           5.0
  Birmingham            38.5            10.9           5.7
  Nottingham            44.9            11.0           5.3
  Bristol               28.8             9.2           3.1
Northern
non-core cities
  Sunderland            39.9            12.9           4.8
  Middlesbrough         41.0            11.7           6.2
  Barnsley              32.2            14.1           3.9
  Doncaster             30.7            12.0           4.2
  Bradford              32.5            10.1           4.4
  Wakefield             30.6            12.3           3.5
  Hull                  43.8            11.2           6.2
  Bolton                30.3            11.0           3.5
  Rochdale              33.5            11.2           3.9
  Wigan                 27.5            12.0           3.2
  Warrington            20.9             9.1           2.9
  Burnley               34.1            11.9           3.1
  Preston               31.4            10.3           3.4
  Blackburn
    with Darwen         33.5            11.1           4.1
Traditional
seaside resorts
  Blackpool             37.3            13.9           4.2

                         No          Households    Overcrowding
                   qualifications,    without       indicator,
                      per cent         central     per cent of
                                      heating,      households
                                      per cent
Core cities
  Newcastle
    upon Tyne           32.6             4.2           7.8
  Leeds                 30.9            20.7           7.8
  Sheffield             32.0             8.2           6.3
  Liverpool             37.8            27.0           7.6
  Manchester            34.0             9.0           11.2
  Birmingham            37.1            20.5           9.6
  Nottingham            33.9             6.9           9.0
  Bristol               26.1             8.8           7.8
Northern
non-core cities
  Sunderland            36.9             2.6           5.7
  Middlesbrough         36.8             7.0           5.9
  Barnsley              41.1             3.5           3.9
  Doncaster             38.1             6.3           3.9
  Bradford              35.1            22.9           8.3
  Wakefield             39.1             6.6           4.4
  Hull                  41.2            21.3           6.4
  Bolton                33.3            13.6           6.0
  Rochdale              36.1             8.0           7.0
  Wigan                 35.3             6.4           4.2
  Warrington            26.9             9.4           4.3
  Burnley               36.0            18.1           5.3
  Preston               31.1            15.1           6.7
  Blackburn
    with Darwen         37.2            14.7           6.8
Traditional
seaside resorts
  Blackpool             37.8            19.2           7.3

Source: Census 2001

Table 2 Core cities and non-core cities--indices of regeneration
need and regeneration achievement

                             Indices of need

                       Rank of  Rank of      Rank
                       average  extent     of local
                        score            concentration

Core cities
  Newcastle upon Tyne     20       24           5
  Leeds                   68       64          24
  Sheffield               60       51          30
  Liverpool                1        5           2
  Manchester               2        4           3
  Birmingham              15       14          15
  Nottingham               7        9           9
  Bristol                 67       71          34

Northern non-core cities
  Sunderland              22       23          37
  Middlesbrough           10       11           4
  Barnsley                28       28          40
  Doncaster               40       32          43
  Bradford                30       31          11
  Wakefield               54       56          61
  Hull                     9       12           6
  Bolton                  50       46          36
  Rochdale                25       29          12
  Wigan                   53       52          48
  Warrington             147      114          78
  Burnley                 37       40          21
  Preston                 59       48          32
  Blackburn with
    Darwen                34       25          25

Traditional seaside resorts
  Blackpool               24       35          10

                                    Indices of achievement

                          Urban     GVA/FTE,   Percentage   Percentage
                        hierarchy  percentage  change in    change in
                          rank       change    percentage   percentage
                         change    1995-2001   employment  unemployment
                        1965-1998               1996-2003    1998-2002

Core cities
  Newcastle upon Tyne       0         18.9        7.3         -26.5
  Leeds                     0         29.3        0.1         -12.8
  Sheffield                -1         25.1       10.8         -38.3
  Liverpool                -1         25.6        8.6         -32.5
  Manchester                0         34.5        3.9          -9.1
  Birmingham                0         34.7        1.5          -8.3
  Nottingham                0         25.7        1.5         -40.4
  Bristol                   1         25.8        4.1         -32.5

Northern non-core cities
  Sunderland                0         10.6       10.6         -25.1
  Middlesbrough            -1         17.8        0.3         -12.3
  Barnsley                  0         13.7 *      4.4         -42.9
  Doncaster                 0         13.7 *      7.4         -32.2
  Bradford                  0         25.2        3.9         -20.3
  Wakefield                -1         26.9       13.0         -31.9
  Hull                     -1         17.9        8.7         -39.3
  Bolton                    0         11.9 *      4.0           0.4
  Rochdale                  0         11.9 *     -2.6           5.0
  Wigan                     0         11.9 *     10.1          -6.1
  Warrington                0         14.8        0.1          69.6
  Burnley                   0         18.4 *      9.6         -31.3
  Preston                   1         18.4 *      4.3          -5.1
  Blackburn with
    Darwen                  0         13.8        5.1         -22.7

Traditional seaside resorts
  Blackpool                -1          9.7        1.5         107.2

* Aggregated data

Sources: The English Indices of Deprivation 2004; (9) P. Hall, S.
Marshall and M. Lowe: 'The changing urban hierarchy in England and
Wales, 1913-1998' (7); State of the English Cities (8) database