9.1. Urban Dynamics – Growth and Stagnation in Cities
It has been estimated that the year 2007 will be a momentous point in world history when, for the first time, urban dwellers will outnumber those living in rural areas (Urbanization, 2003). Alongside this rise in urban population, there has also been an increase in the number and size of very large cities, bringing both benefits and problems of urban growth. Figure 9.1 shows the population of London over 200 years from 1800 to 2000. The city was already large at the start of the nineteenth century with more than one million inhabitants. Over the next hundred years, throughout the Victorian era, London's population grew steadily to reach seven million people by the early 1900s, before peaking around eight million in the 1940s and gently declining back to seven million in the remaining sixty years up to 2000 (although the Home Counties surrounding London continued to grow).
Figure 9.1. Growth of London
City growth and stagnation is a dynamic phenomenon well suited to modelling and simulation. Why do cities grow? What factors limit their growth and economic success? Why is the standard of living in some cities much higher than others? Why do cities that were once successful fall into periods of stagnation and decline? How do you turn around a failing city? These kind of questions faced Jay Forrester in the creation of the Urban Dynamics model. The story of the model's development is told in 'The prophet of unintended consequences' an article about Forrester's life and work by Lawrence Fisher (2005, p 8):
When John F. Collins, a former mayor of Boston, took a temporary appointment at MIT as a visiting professor, he was assigned the office next door to Professor Forrester's. It was 1968, riots had broken out in cities across America, and the two instructors naturally fell into conversations about solving the stagnation and unemployment that plagued many cities. 'Collins was very much a man of action', Professor Forrester recalls. 'I suggested enlisting researchers – not urban studies students, but people who knew the real urban world – for a half day a week for as longas it would taketo extractadynamic pictureof the problem. Collins's immediate answer was: "They'll be here Wednesday afternoon"'. With Mr Collins's clout, the two quickly assembled a team of high-level advisers from politics and business to research the dynamics of urban poverty. After four months, Professor Forrester had the basis for a new book, Urban Dynamics, with a startling assertion: the harder a policymaker tried to relieve poverty, the more that poverty would increase.
(Reprinted with permission from strategy+business, the award-winning management quarterly published by Booz Allen Hamilton.)
9.1.1. Urban Model Conceptualisation
We will return to the policy implications later. Now I want to concentrate on the model's conceptualisation that captures a fundamental interaction between urban development and the ageing of urban infrastructure. The dynamic hypothesis recognises that under favourable conditions, the interplay between the parts of a new urban area cause it to develop. As the area grows and its land area fills, the processes of ageing cause stagnation. As the urban area moves from the growth phase to the equilibrium phase, the population mix and the economic activity change. Unless there is continuing renewal, the filling of the land converts the area from one marked by innovation and growth to one characterised by ageing housing and decline. This verbal hypothesis is shown as causal loops in Figure 9.2. Growth and stagnation arise from reinforcing and balancing loops. The reinforcing growth process (R) is economic expansion of the city in which the existence of new economic infrastructure encourages further expansion of infrastructure and so on. This process continues until the available land area is filled, as represented by loops B1 and B2. Meanwhile new economic infrastructure gradually ages to become old infrastructure in loop B3. Ageing infrastructure combined with rising population eventually lead to city stagnation by making the city less attractive thereby reducing net inward migration in loop B4.
Figure 9.2. Preliminary dynamic hypothesis for urban growth and stagnation
This intriguing hypothesis, inspired by meetings between the urban advisers and the modeller, was fleshed out to yield the stock and flow diagram in Figure 9.3. In particular, the concept of economic infrastructure took on a specific and concrete meaning in terms of nine stock accumulations or levels grouped into three subsystems. Forrester (1975, pp. 178–179) describes these sectors in the following way:
Figure 9.3. Stocks and flows in urban dynamics model
Source: Forrester, 1975 Collected Papers of Jay W. Forrester, Chapter 11, Figure 1 from: System Analysis for Urban Planning, Pegasus Communications, Waltham, MA. Reproduced by permission of Jay W. Forrester.
Across the top the industrial sector contains commercial buildings in three categories distinguished primarily by age. Across the centre are residential buildings in three categories, also distinguished by age and condition. Across the bottom are three economic categories of population.
Even though the information linkages are not shown in this figure it nevertheless begins to show possible reasons for urban decline. Continuing Forrester's interpretation:
The age of a building tends to determine the character of its occupants. A new commercial building is occupied by a healthy successful commercial organisation that uses relatively more managers and skilled workers than unskilled workers. As the building ages, it tends to house a progressively less successful enterprise with lower employment skills. In addition to the changing employment mix as the industrial building ages, there is a tendency for total employment per unit of floor space to decline. On the other hand, as residential buildings age, there is a tendency for occupancy to increase as well as to shift to a lower economic category of population. One then perceives a condition in which the aging of buildings in an urban area simultaneously reduces the opportunities for employment and increases the population. The average income of the community and standard ofliving decline.
(Forrester, 1975 Collected Papers of Jay W. Forrester, Chapter 11, from: System Analysis for Urban Planning, Pegasus Communications, Waltham, MA. Reproduced by permission of Jay W. Forrester.)
Figure 9.4. Information links to underemployed arrivals
Figure 9.4 shows the same nine system levels and one of the 22 flow rates. This single flow rate depends on the factors or conditions in a city that urban experts believe are likely to attract underemployed people from other regions of the country, thereby causing inward migration. For authenticity, the abbreviated variable names from the original model are shown. However, full names are given for all variables that directly affect inward migration through the Attractiveness Mobility Multiplier (AMM). The dotted lines are the information linkages from the system levels to control the one flow rate – here the arrival of underemployed population into the urban area. The various levels of the system combine to create a composite attractiveness that determines the inflow rate to the area. If the area is more attractive than those from which people might come, a net inward population flow occurs. If the area is less attractive, an outward flow dominates. Five components of attractiveness are shown in the figure, which are now described – starting at the top of the diagram and working downwards. In the upper right corner is the underemployed job multiplier (UJM), which relates the population to the available jobs and represents the income-earning attractiveness of the area. The circle UAMM (Underemployed Attractiveness from Mobility Multiplier) generates the attractiveness created by upward economic mobility. In other words, an area with high upward economic mobility is more attractive than one offering no hope of advancement. The circle UHM (Underemployed Housing Multiplier) relates the underemployed population to the available housing. The area becomes more attractive as housing becomes more available. UHPM (Underemployed Housing Programme Multiplier) represents the attractiveness of a low-cost housing programme if one exists. In the lower right corner, PEM (Public Expenditure Multiplier) is the influence on attractiveness of public expenditure per capita. Rising per capita expenditure means better public services, better schools and higher welfare budgets.
Source: Forrester, 1975 Collected Papers of Jay W. Forrester, Chapter 11, Figure 2 from: System Analysis for Urban Planning, Pegasus Communications, Waltham, MA. Reproduced by permission of Jay W. Forrester.
The concept of attractiveness is fundamental to the population flows. All the characteristics that make an area attractive – these five and many more – combine to influence migration. An attractive area draws people. But almost every component of attractiveness is driven down by an increase in population. If there is an excess of housing, the area is attractive, but a rising population crowds the housing. If there is an excess of jobs, the area is attractive, but the incoming flow of people fills those jobs. In other words, migration continues until the attractiveness of the area falls and becomes equal to that of all other places from which people might come.
9.1.2. Policy Implications of Urban Dynamics
Urban Dynamics challenged conventional wisdom that urban problems are caused by economic and social forces beyond the control of cities. On the contrary cities' own policies determine its economic success and the quality of life of its citizens as illustrated by the base run of the model in Figure 9.5, which is a reproduction of an original simulation made in 1968. Notice the healthy development of the city over the first 100 years when labour, managers, worker housing, premium housing and new enterprises grow strongly. Meanwhile, declining industry and underemployed housing remain small. Over the next 50 years social and economic conditions in the city deteriorate. Declining industry rises and new enterprise shrinks. The number of underemployed rises swiftly, peaking in year 120 and premium housing ages to become underemployed housing. The city gradually settles into an equilibrium characterised by chronic underemployment, run-down housing and reduced enterprise.
Figure 9.5. Base run of the urban dynamics model
Source: Forrester, 1975 Collected Papers of Jay W. Forrester, Chapter 11, Figure 3 from: System Analysis for Urban Planning, Pegasus Communications, Waltham, MA. Reproduced by permission of Jay W. Forrester.
The model's message to policymakers was counterintuitive. Poverty cannot be solved by welfare spending – at least not in the long-run. Simulations suggested that an increase in low-income housing exacerbates the decline of the central city by attracting low-skill workers. Similarly, government sponsored job programmes for the underemployed were found to be ineffective because they too attract low-skill workers and trap them in declining businesses and inadequate housing. Instead, to help cities return to economic health, spending should be allocated to incentives that stimulate new businesses and that speed the demolition of declining business structures and derelict housing. Another surprise result from simulations was that external financial aid to the city, to supplement local spending, eventually forces up local taxes. The reason is that the natural tax base gradually erodes as more and more underemployed people arrive in the city seeking government handouts.
These controversial recommendations sparked much political and academic debate at the time. Business Week reported that 'Urban Dynamics is an unsettling, complex and ground-breaking new book about our cities and the decay that afflicts them'. In addition, Fortune magazine noted that 'Urban Dynamics has become the subject of heated debate even prior to publication. It has captured the imagination of politicians and managers in the public domain, and it has been furiously attacked by social scientists'. It remains a classic model in the system dynamics literature and a potent reminder of the ability of models to influence the thinking of political and business leaders.