Review of Scotland's Cities - The Analysis
2 CITY DEMOGRAPHICS
2.1 DECLINE AND SHIFT
The historical norm of economic development has been uneven geographies of population change. Sometimes this has reflected marked variations in fertility and death rates between areas, but within a nation, disparities are more likely to reflect differences in net migration into and out of particular towns, villages and cities. In this sense population and household change within places are a measure of their popularity as places to work and live.
There are widespread assumptions made about how cities are faring. The purpose of this section of the report is to provide evidence to inform debates. Examination of the evidence leads to two broad conclusions:
- patterns of change vary substantially across our cities, there is no common template of decline or growth;
- since the middle of the 1990's cities have been better at retaining people.
This section examines past evidence as well as projections for the next decade.
2.1.1 City Change: the Bigger, Longer Picture
There have been major commonalties in the population trajectories of advanced economy cities over the last century. Scotland is one of the most urbanised countries in the world and was one of the first to urbanise. Indeed the growth of Glasgow after 1800, its decline after 1950 and its relative stabilisation and economic growth over the last decade make it a major exemplar of the kinds of city change that have accompanied industrialisation, then de-industrialisation and more recently post-industrial development.
Different stages of city change have been apparent across the advanced economies. Until the early decades of the 20th century the fast rise of city employment and parallel rural depopulation (absolute centralisation) meant that not only did city populations grow, but the share of national population contained within them increased (relative centralisation). However in the first half of the 20th century the pattern changed. Core city employment and population continued to increase (for instance Glasgow's population peaked at 1.1 million in 1951) but the development of suburban transport meant that the suburban areas grew even more rapidly than the core cities (relative decentralisation). This meant a declining share of national population in the city cores even though these cores were still expanding and, often, bursting at the seams.
This period of relative decentralisation was then, after World War II, typically followed by the loss of jobs and population from the city cores. Not only was their share of national population falling but their absolute totals of people and jobs began to fall and the city to decline (absolute decentralisation). As personal mobility increased, more affluent and above average income family households often spread to the suburbs. There they sought both larger homes and refuge from the difficult environmental conditions of core cities. As this population decline meant abandonment of homes and streets and facilities, the landscape of decay and decline fuelled further suburbanisation and left in the core, disproportionately, those who could not afford to leave.
These trends were reinforced by the further sharp reduction of the older economic base of cities, through the de-industrialising decades of the 1970's and 80's, and the reduction of public investment in services and infrastructure. Housing and planning policies, for instance through relatively generous tax breaks for home owners (invariably purchasing in the suburbs), reinforced these shifts, at least until the middle of the 1970's, and so this synthesis of decline, decay and disadvantage became the story of our cities. The broad thrust of that tale, at least until the end of the 1970's is relevant to Scotland but it is too generalised to provide a more reasoned interpretation of how Scots have chosen, or rejected, their cities over the last quarter century. We need a more detailed understanding of the pattern of change and its policy implications, and above all we have to avoid an unthinking extrapolation of the past. We have to identify what is likely to be different in the next decade, the next quarter century.
The big drivers of change have shifted. The old economic base has largely left our cities. New sectors are now growing faster than the old are declining, and these issues are examined in detail in the next chapter. The costs of commuting, not just fares and fuel, but also the opportunity cost of time, are increasing as are the complexities of managing household activities for two adult households with career jobs, and there is a much deeper recognition of the social and environmental costs of dispersed suburban development. If it was the growth of the family that fuelled suburbanisation then the new demographics ahead imply some quite different possibilities. These different urban futures could be frustrated if our cities have acquired negative social and environmental attributes that essentially force suburban living on particular household income and age groups. Additionally, since the 1970's there has also been both 'urban' renewal and neighbourhood regeneration policies which have impacted patterns of urban change.
The European story of the last two decades, and indeed much of the USA, has been of recovering core cities. This has occurred in some but not all cities. In some places rates of change have resulted in relative centralisation, but more typically the experience has been that city change has been positive but at less rapid rates than the national average, so that city growth with relative decentralisation is the most common pattern now prevailing. Not only do city trends differ from place to place, but localised patterns of city population and economic change have become more diffuse and fragmented. Tales from the city have become diverse and complicated, and at the same time encouraging. However we have to examine in more detail the distinctive Scottish experience.
Tales for the city have become diverse and complicated, and at the same time encouraging...
2.1.2 City Population Change in the Wider Scottish Context
The changes in our cities, and the surrounding city-regions, have to be seen in the context of national developments in population. And there are three particularly important contextual issues for Scotland's cities.
- Scotland's population has been slowly declining since the peak recorded in the 1971 census, and it is in this regard that Scotland is unusual within Western Europe;
- Household size, in common with Western Europe, has fallen consistently over the last half century and, in consequence, household numbers have grown despite population decline; fewer families and more single person households, both young and old, are key structural shifts; and
- Scotland's population is ageing.
With these three main trends in mind, it is important to assess how change has been manifested in the cities and to establish how effective the cities and the city-regions have been in competing for the residential demands made by Scots.
2.1.3 Population and Household Changes
Most commonly commentators draw attention to changes in city population. However with rapid reductions in household size accompanying economic growth and social change it is equally important to examine trends in household numbers as well as total population. For instance with rising incomes and reducing household size it is possible to have a significant reduction in the population of a city but no substantial decline in wellbeing as smaller household units may use all of the vacated dwelling stock. From a policy perspective it is household rather than population numbers which primarily drive overall demand and need for housing and certain other key services as well as the number of domestic tax units. We have to understand what is happening to both totals.
Estimated population and households totals, at the start of each decade, for Scotland and the four largest cities are shown in Table 2.1 for the period 1961 to 2001. There are two totals recorded for 2001. 2001 (A) is the forecast estimate of GRO based on the 1991 Census rolled forward with births, deaths and estimated migration to 2000 and projected forward to 2001, and 2001 (B) is the estimate based on the 2001 Census (and announced in September 2002). These two totals for 2001 are rather different and they inevitably mean a heavily qualified interpretation of the demographies of Scotland over the last two decades. The Registrar General, in announcing the 2001 Census based estimates, suggested that there were two likely causes of the disparities between 2001 (A) and 2001 (B). First, it had been feared that the introduction of the Community Charge would have led to under-recording of population in 1991, and 1991 Census estimates were adjusted upwards to allow for that factor. It now transpires that the upward adjustments were probably too generous and that the published census figures had been an over-estimate of 1991 population. Secondly, estimates based on the Census, for inter-censal years, are influenced by birth and death registrations and estimated mobility rates. It is now believed that the measures used to estimate migration have generally under-stated outward mobility from Scotland.
TABLE 2.1: Population and Households, Scotland and Four Cities, 1961-91
These difficulties, which apply in the rest of the UK as well, mean that the 1991 figures, and subsequent estimates based upon them have some limitations. One obvious strategy for analysis is simply to contrast actual Census outcome figures for the two periods 1961-81 and 1981-2001 (avoiding the 1991 mis-estimates). Contrasts for these periods are provided (Table 2.2). However two decades is a long period and there are other non-census variables, such as employment totals, new construction and house price changes, which suggest a change in city fortunes after the mid-1990's. It is therefore important to look at growth trends from 1996-2001, and future estimates, emphasising trend rather than scale issues. Although the Registrar General intends to produce revised data series for each year from 1982 to the present, he has indicated that revisions are not expected to have a great effect on the year to year changes in the "old" series, but may steepen or lessen the trend over time.
TABLE 2.2: City Population Change Rates Based on Census Outcomes
The significant decline of population, in the cities, can be seen for the 1961 to 1981 period (Table 2.2). Only Aberdeen (where previously more rapid population loss was turned around after the beginnings of North Sea oil exploration in the late 1960's) fell at a rate less rapid than the falling Scottish average. Between 1961 and 1981 the share of city population in the Scottish total fell from 39 to 31%. Much of this was due to the unparalleled decline in Glasgow, population there falling by a third in two decades. It is always important to remember these figures: the absolute and relative decline of Glasgow has been on a scale unmatched in Scotland, at the extreme end of city decline experiences in Europe and amongst the worst of advanced economy cities.
In all the cities, population loss was largely attributable to households leaving the city for the surrounding city-region. At times in the past, emigration overseas and to the rest of the UK have been more important than at the present, but moves out of our cities were and are essentially local movers. Movement surveys emphasise different factors in different places and at different times but they also highlight that the shift to the suburbs of the city-region was selective both by income and family type, with middle and upper income households with children moving out.
The declining economic base of Glasgow undoubtedly contributed to population decline. But considerable decentralisation was policy rather than market led. In the period from the 1950's to the 1980's at least 150,000 people were induced to leave Glasgow for New Towns and overspill areas in a deliberate attempt to decongest the city. Indeed the desired size for the city's population, as envisaged by planners at the end of the 1940's, was broadly in the same range as the 2000 population. Research has shown that public sector induced migration from Glasgow was skill selective and so were the suburbanising flows of middle income families and young, childless couples. The latter groups moved out because the city contained relatively few home-ownership options as public authorities owned the bulk of housing land and used it for municipal housing. These historical observations simply make the point that past policy has had negative as well as positive impacts on our cities and that big planning decisions have not always taken account of consumer preferences. The important issue is whether our planning and research systems are now sensitive to consumer preferences.
The household numbers recorded in the 1961 and 1981 Censuses, (Table 2.1), indicate increasing numbers in the cities over the 60's and 70's. From the 1960's onwards the demography of reducing household sizes was already off-setting some of the effects of population decline.
The two decades 1981-2001 (when aggregated together) show some substantial differences from the previous period. Although Scottish population continued to fall both Edinburgh and Aberdeen recorded population growth. Whilst Glasgow's decline rate had reduced significantly it still remained problematic and Dundee's rate of decline had accelerated. Again the 'four city' share of population had fallen, but at a much slower rate, falling from 31% to 28% of the Scottish total. In overall terms relative city decline was slowing.
2.2 A NEW DYNAMIC?
These twenty year figures mask too much and it is important to understand trend changes within the period, especially over the last five years. We need to establish or question whether there is a new demographic dynamic for the cities.
If the absolute scale of city population or household totals is disregarded it is still possible to establish recent directional trends which are 'untainted' by 1991 over-estimates/under-estimates. Table 2.3 sets out 1996 to 2001 population indices derived from estimated annual totals (projected from a 1991 base). As long as migration rate errors (and this is a strong assumption) are similar across the cities and Scotland then these indices still indicate trends relative to 1996 and each other. The indices suggest that Edinburgh's population was still growing, Glasgow was declining at a much slower rate than previously and less rapidly than Aberdeen (now declining rather than growing) with Dundee incurring the sharpest decline rates.
TABLE 2.3: Indices of Population Change 1996-2001, Cities and Scotland (Projections Based on 1991 Census)
For the 1996 to 2001 period household numbers, (Table 2.4), show a marginal fall in Dundee but increases in the three other main cities. Largely similar trends were projected ahead to 2016 (from 2000), with both Edinburgh and Glasgow increasing household numbers ahead of the Scottish index (of 112, 2001-2016), Aberdeen increasing and Dundee falling.
TABLE 2.4: Four Cities: Indices of Changes in Household Numbers 1996-2001 and Projected Changes 2001-2016 (1991 Census Base)
These trends and future projections may be relatively unaffected (as directional indicators) by the revised series to be derived from the estimate totals of the 2001 Census. They may allow us to be more optimistic about the cities, though only cautiously. But they do, however, serve as a reminder not to extrapolate the gross change 1981-2001 into the future. The main points of 'directional' change in the recent past have been:
- For all cities and for both past and projected changes, household growth has been less rapid in the core city than in the rest of the city-region
- Household growth numbers have been above the Scottish average in the core cities of Aberdeen and Edinburgh and they have been positive in Glasgow
- For the decade ahead it is projected that both Edinburgh and Glasgow will have household growth above the Scottish average, pressures will become less pronounced in Aberdeen city and Dundee will experience small declines
Population and household projections are essentially the outcomes of extrapolating present behaviours into the future. They are likely to be wrong, especially at more local levels, if either the path of economic change alters or households begin to choose different residential options. Or indeed they may fail to anticipate the positive effects of large regeneration programmes. That said, it is illuminating to examine the geography of forecast rates of change of household numbers across Scotland in the decade ahead.
All of the authorities with household growth forecasts significantly above the Scottish average, with the exception of East Renfrewshire, lie in an east coast belt that runs down through Aberdeenshire, missing out Dundee and Fife, to Stirling and through the Lothians. Local income estimates also mark these growth localities out as areas of higher incomes. In contrast the six areas of least rapid change, are Inverclyde (where household numbers are set to fall), West Dunbartonshire, East Ayrshire and Dumfries and Galloway (or the western fringes of the Glasgow centred housing market) and Dundee City and Aberdeen City. These patterns reflect the operation of a number of different geographies, which suggest the emergence of important city and urban change agendas in addition to the continuing pressures for recovery from Glasgow. These are:
- the strengthening of growth management and infrastructure provision agendas within a broad region of growth on the east of the country;
- the emergence of severe absolute decline in the outer, western areas of the Clydeside conurbation and the South-West, where the old economic base is still collapsing but the asset base for new service industries is less apparent;
- core city quality and choice issues in Dundee and, to a much lesser extent, Aberdeen.
There is however a case for considering how new demographics, of household size and age, will transform overall change into different housing and neighbourhood requirements. The ageing of the overall population in the decades ahead is well established. But household structure is set to continue to change, and significantly. The proportion of Scottish households containing any children has been falling significantly over the last two decades and in 2003 only a quarter of Scottish households will contain children (18% in two adult households and 7% in single parent households). The three-quarters of households that are all adult households are split between 34% one person households and 41% with two or more adults. By 2013 it is forecast that only a fifth of households (8% single parent and 12% two or more adult families) will contain children. Of the four-fifths 'all adult households', 38% will be single person and 42% multiple adults.
This is a significant change in the transmission of demographics into housing needs and demands, especially at more local, city scales. Present forecasts for the cities for the period 1998 to 2012 suggest that:
- Edinburgh (34%), Glasgow (30%) and Aberdeen (28%) will experience single person growth numbers relatively close to or just below the Scottish average (34%) but Dundee will lag (17%);
- Both Glasgow (37%) and Edinburgh (39%) will encounter single parent family growth faster than the Scottish average (33%), in contrast to Dundee (16%) and Aberdeen (4%);
- All of the cities will experience reduction rates for households with children and two or more adults significantly above the Scottish average (-34%) with the exception of Edinburgh (-31%);
- It is also predicted that only Edinburgh (16%), out of the four main cities, will increase two adult households more rapidly than the Scottish average (15%).
Once again it is important to note that recalculated totals based on the 2001 Census may amend this pattern.
With a current population of around 55,000 the City of Inverness is only a third the size of Dundee - the smallest of the other Scottish cities. As a regional centre, Inverness is however performing well; its growing population reflecting the city's ability to attract migrants.
The population of Inverness has grown rapidly, albeit from a low base, since the 1970's. In the last thirty years, the population of the Inverness Area, which includes the city and outlying urban settlements, has grown 34% from 41,000 to 65,500. Most of the growth appears to have been due to in-migration of families and early retirees due to increasing employment opportunities and the perceived high quality of life. The immediate hinterland has also been growing, against a backdrop of a static Highland population.
More modest growth is projected for the future, largely fuelled by in-migration and the continued growth in the local economy. The Inverness Area is projected to grow by around 4 to 5% in the next ten years, against a backdrop of a slowly declining Scottish population.
In essence, the projections set out above highlight the assumed ability of Edinburgh both to retain a higher share of households with children and two person (often high earning) households. The challenge for the other cities is in part to become more competitive, vis-à-vis their surrounding city-regions, in attracting families with children and retaining two adult households. These issues will be explored further below in relation to city housing issues but there is clearly an important set of issues in these figures for cities; will they be more competitive as housing locations as demographic change fashions more childless households? Looking at Scotland as a whole over the next decade the obvious challenge for the cities is to compete more effectively for the:
- 150,000 extra single person households
- the 90,000 extra adult only households
- the 30,000 extra single parent households; and
- to be least adversely affected by the reduction in two-adult families by 105,000.
The less well recognised challenge may however to be to raise income, employment and population potentials for Scotland so that all of these household growth estimates are under-estimates. Simply meeting these household projections is a prediction for a Scotland growing slowly in household numbers but declining in significance within the UK and Europe. Changing these trends will depend, to a significant extent on how the economies of our cities develop.