Flying Carpets and Urban Development Trajectories
Flying Carpets and Urban Development Trajectories
Piet Rietveld 0
0 P. Rietveld (
1 ) Department of Spatial Economics, VU University Amsterdam , Amsterdam , Netherlands
Generalised costs of transport and communication have decreased in many domains during the past centuries. Some cities have strongly benefitted from these developments. The more recent development is that passenger transport has become the leading driving force, as opposed to freight transport in earlier logistical revolutions. Information and Communication Technology (ICT) is a substitute for some forms of physical transport but certainly not for all; cities continue to depend on high quality transport networks. An underestimated component of transport networks is vertical transport in high rise buildings. This played a key role in urban developments in the 20th century and it may be expected to continue its role in the next century. Probably, the 'flying carpet' notion is most clearly visible in the vertical movements in high-rise buildings.
Flying carpet; Logistics; Agglomeration advantages; Vertical transport; Accessibility
From the very beginning transport has been one of the key factors underlying urban
growth and development. Leading cities in world history such as Rome, Venice,
London and New York owe their success in a considerable degree to technical and
logistic innovations in the transport domain. As we will discuss, there is a close
relationship between the dynamics of transport and logistics on the one hand and the
dynamics of cities, on the other.
The key question to be addressed in this article is to what extent the ongoing process
of improvement of transport and communications will lead to the further growth of
cities, or whether just the opposite may be expected to occur. For example, Clark
(1958) wrote an article with the title ‘Transport: maker and breaker of cities’,
suggesting that transport may also have very adverse effects on cities. A similar claim is
Piet Rietveld deceased on November, 1, 2013, at the age of 60 years.
sometimes made for improvements in transport technology, for instance, applications of
advanced information and telecommunication technology (ICT) such as the internet
and social media.
We will show that developments in transport and communication have indeed
led to turbulent developments of cities, but in most cases the cities have
continued to grow, although the relative positions of cities may have changed.
During the last decades we have observed an ongoing decrease in transport
costs (in particular, for long run travel) and the emergence of modern
telecommunication and the internet. This may be interpreted as a ‘flying carpet’
phenomenon: movements can take place smoothly and at low costs, giving
more freedom to citizens and firms to choose where they like to go. In this
paper we will discuss how cities will be affected by this development.
This paper is organized as follows. Network developments and urban
development trajectorieswill address network developments and urban development
from a long term perspective. Inter-urban transport and urban development; a
history of logistical revolutions brings the two domains together by looking at
the joint development of transport technology and cities. Internal transport
within cities and urban development addresses the implications of intra-urban
transport for urban development. Agglomeration advantages, transport and
communications will discuss, in more detail, the key concept of agglomeration
advantages and to what extent it relates to costs of transport and
communication. Agglomeration advantages and transport networks gives a more detailed
account of which specific transport networks are important for the realisation of
agglomeration advantages. Special attention is given to vertical transport in
high-rise buildings, an underdeveloped element in the analysis of urban
development. Conclusion concludes.
Network Developments and Urban Development Trajectories
Transport and communication networks during the past centuries have gone through
substantial cycles. A nice illustration can be found in Levinson (2005) for the USA.
The 19 and 20th century have witnessed the growth and decline of at least three major
networks: canals, rail and telegraph. Given the high costs of investments it sometimes
took 50 years or more to build such networks, but after consolidation these networks
soon started to shrink, apparently because emerging competing networks were
generating high benefits.
Gruebler and Nakicenovic (1991) have undertaken a similar analysis for a
broader range of countries and found that this typical inverted U-shaped pattern
of development not only applies to network length, but also the network use. The
effective use of canals is incredibly smaller than it was 150 years ago, and the
telegraph system has vanished. The decline of rail is not as drastic as with canals
and telegraph, but there has been a huge decline. Only recently the construction of
high speed- and light rail has led to a renaissance of rail in a limited number of
The above figure gives an important but one-sided image of long term trends
in transport. Another story is told by Table 1 based on Crafts and Venables
Table 1 Real costs of maritime transport (1910=100)
(2001), which shows that in other domains continuity tended to prevail. The
table shows that for maritime transport real costs decreased with about 83 %
between 1770 and 1990. Impressive efficiency improvements took place in
particular during the period that the steamships entered the arena (between
1870 and 1910), but also in most other periods.
As mentioned before, the key issue to be addressed in this paper is to what
extent these long term cycles and trends express themselves in urban
development trajectories. Do we see definite patterns where cities continue to grow, or
may the opposite occur? The overall pattern for urban growth we find in the
literature is represented in Fig. 2. Growth takes place in various stages,
alternated with periods of decline, but in most cases the rather drastic lifecycle
pattern shown for specific transport networks in Fig. 1 does not apply. The
future development is uncertain, and hence at some stage a bifurcation may
The periods of upturn and downswing of cities usually have to do with lifecycles of
sectors where sectors that were vital during one stage of development turned into
decline, leading to deterioration in the cities that specialized in this sector. There may
well be cases where this leads to the collapse of the city (think of ghost towns, rapidly
developed after the finding of a natural resource, and also rapidly left after the depletion
of the resource), but as indicated in Brown (1974), this mechanism will no longer work
when urban size has reached higher levels (he mentions a number of 250,000
inhabitants). In that case various mechanisms (industrial diversification, invested social
overhead capital, political power associated with larger cities) will stabilize the
development of cities, leading to a so-called urban size ratchet providing a stabilizing force
and reducing downward tendencies.
Within the context of the present paper it is relevant to find out to what extent
transport conditions—in particular, new developments in transport technology—
had a specific effect on the growth or decline of specific cities. It appears that
some cities with a strong transport based position in the past like London,
Amsterdam, and Venice continued to keep important positions in the urban
networks of Europe, although their economic orientation may have shifted far
away from transport. On the other hand, there is a good number of towns that
were important hubs in the past (specific port cities in the Roman empire, and a
good number of Hanseatic towns) that nowadays are just insignificant spots on the
maps due to physical phenomena, such as, silting up of water connections and
because of shifts in trade flows.
Fig 1 U.S. Networks as a proportion of maximum extent (source: Levinson 2005)
Another line of research on urban growth trajectories strongly supporting the
ratchet notion has been developed by Davis and Weinstein (2002). They find for
Japan that when cities receive a shock such as intensive bombings in the
second world war, cities quickly return to their original position in the urban
hierarchy. In a similar fashion Brakman et al. (2004) find that bombings of
German cities during the second world war did not substantially change the
positions of cities in the German urban hierarchy. This underlines that
stochastic shocks in urban systems do not necessarily affect longer run growth rates of
cities. This does not mean, of course, that all cities grow at the same rate, but
it demonstrates that urban growth is not just a random walk; the locational
fundamentals of a city appear to have a long term effect. At the same time the
specific sectoral compostion of cities will have its effect. For example, Cheshire
(1990) shows that port cities in Europe performed badly in terms of
employment, most probably since during the 1970s and 1980s containerization reduced
the labour intensity of freight transport. More recently economic growth in port
cities is no longer smaller than that of other cities (Koks 2012). Thus, port
cities experienced a dip during several decades but seem to have coped
effectively with the challenges of the container revolution.
We conclude that where in specific transport networks there have been clear
upswings and downswings during the past centuries, downswings in urban
developments have not been as dramatic. When cities grow beyond a certain
level, the ratchet mechanism tends to work as a stabilizer preventing strong
downturns in cities. So, cities that depended on a strong transport function in
the past and that lost this position for some reason may still be vital when they
succeed in shifting their economic orientation. Meanwhile, it is clear that
depending on several factors such as the sectoral composition of a city and
the overall demographic developments in a country, cities will grow at varying
rates. The next section gives a more in depth account of the role of transport
during the various phases of economic development.
During the past millennium, four major periods of rapid and fundamental changes in
the infrastructure system took place that had major impacts on the development of the
urban system. Andersson (1986) described them as four logistical revolutions. He states
that the great structural changes of production, location, trade, culture and institutions
are triggered by slow but steady changes in the logistical network. For an update see
Karlsson et al. (2009).
The first logistical revolution took place between 1000 and 1500 was based on
improvements in particular in sea transportation on the one hand and the creation of a
basic infrastructure for trade on the other hand. The development of larger vessels, new
sea routes and new ports helped to integrate trade between city states along the coast of
the Mediterranean and the Hanseatic cities along the coasts of Northern Europe. The
emerging city system formed the basis for this increased trade. Owing to larger vessels,
transport costs and trade barriers were reduced. Large economic trade centers in this
period were Venice, Florence, Bruges, and Lubeck. The population size of those cities
grew until around 100,000 inhabitants. Keywords of this era are emergence of cities,
merchants, and specialization of production.
The second logistical revolution (1500–1800) was again based on maritime transport
and on institutional innovations. The Portuguese and later the Dutch were able to build
ocean-crossing merchant ships, while an efficient banking and credit system was
developed. Both transport costs and trade barriers were further reduced by those
developments. Initially, Antwerp, and somewhat later Amsterdam, were the central
nodes, but all the large capital cities in Europe were engaged in international trade
(Braudel 1994). By the end of this era, London became the commercial and financial
centre of the world (Karlsson et al. 2009). London had about 800,000 inhabitants at that
time. Keywords of this era are international trade, central banking system, science, and
The third logistical revolution (1800–2000) is linked to the industrial revolution.
Driven by technological innovations, the infrastructure network expanded from sea
transport towards land transport; first canals and later on rail, and finally road (see
Fig. 1). The new and relatively large scale transportation systems made it possible to
establish vertical integration between the raw materials in Northern America and the
large scale manufacturing industries in the port cities of Western Europe. Division of
labor, the onset of motorization and mechanization, mass production, standardized
products, and multinational companies were keywords of this era. Large new industrial
cities developed: Manchester, Liverpool, Detroit, Chicago and later on cities all over
the world. The size of cities grew from Manchester 0.5 million inhabitants to over
10 million inhabitants in Chinese industrial cities Fig 2.
The current fourth era of logistical revolution manifests itself in
knowledgeintensive and high-tech urban regions and corridors. It is based on a rapid extension
1 Inter-urban transport and urban development; a history of logistical revolutions and Internal transport within
cities and urban development are partly based on Rietveld and Bruinsma (2012).
Fig. 2 Illustrative long-term trends in urban development
of motorways and air connections and on an increased capacity and speed of
information processing and transmission (Karlsson et al. 2009). Keywords of this era are
knowledge workers, ICT, computers, R&D investments, out-sourcing,
just-in-timesystems and customized production. Main factors behind this development have been
the liberalization of international trade, international capital flows, and foreign direct
investments. It is in particular in the fourth phase that the travel of persons has become
the dominant factor as opposed to the transport of goods, which leads to an increasing
emphasis on the need for transport at high speeds.
Karlsson et al. (2009) conclude that each revolution has been fuelled by successive
technological revolutions, in particular in advances that have cut the costs of
transportation of goods, people, and information. Moreover they state that the improvements
have reduced transportation and communication costs and time, which have brought
different parts of the world closer to each other.
Most of the literature on this subject in the regional science domain is based on the
third logistical revolution. This also holds true for the new economic geography that
focuses on the role of agglomeration advantages in the location of the manufacturing
sector. A main result on the role of transport costs has been formulated by Krugman
(1991) and is represented in Fig. 3. When transport costs are high (on the right side of
the curve) agglomeration advantages are not strong enough to attract manufacturing to
the core region and the manufacturing sector is just spread across the core and
peripheral regions. As transport costs go down (see Table 1), a concentration tendency
will occur towards the core region. The advantage of the core region is that it has a high
internal market so that production is cheaper here than in the periphery. However, when
transport costs decrease further, manufacturing will move to the periphery, since the
share of transport costs in total production costs will decrease and the periphery is
cheaper for some of the other inputs. This is the well known urban rural manufacturing
shift that first took place within countries and later on occurred at the international
level. An essential lesson is that reductions of transport costs can have a concentrating
effect during the earlier stages of development and a deconcentration effect during the
An important question is of course what happens with the economic core when it
loses a substantial part of its manufacturing sector. The answer is that here we arrive at
the limits of this model. The increasing role of the service sector in urban and national
economies is not incorporated here. And it is exactly the shift from a manufacturing
driven economy to a (both business and personal) service- and knowledge-driven
economy—although this also holds, to some extent, for the other transitions mentioned
above—that marks the transition from the third to the fourth logistical revolution. The
mechanisms of the latter will be discussed in more detail in Agglomeration advantages,
transport and communications. But first we will shift the focus from the interurban to
the intra-urban level.
Internal Transport Within Cities and Urban Development
In the preceding section we focused on the effects of changes in interurban transport on
urban developments. In the present section we will now concentrate on the relevance of
internal transport within metropolitan areas. In his interesting contribution, Clark
(1958) indicates how changes in transport costs have always played a role in urban
dynamics. In the nineteenth century, one of the bottlenecks observed for the larger cities
was that transport of the increasing number of workers to the work places became
problematic. Thus, cities remained compact and limited in size. The emergence of
tramways in the second part of the nineteenth century finally enabled a process of
spatial expansion of cities (Hall 1994), and this was further reinforced by the entry of
the electric tram in the beginning of the 20th century, as well as by underground
systems. The coming of bus and car led to a further growth of urban agglomerations,
but now this was mainly an increase of surface area, not of population size. Gradients of
density and rents started to level off. An extreme example is Los Angeles where the
center-supporting role of rail was broken down and a large area with about
homogeneous densities of workplaces and residences emerged.
Clark expected in 1958 a similar development for the large European
agglomerations; not only sub-urbanization of residences, but also of work. However, this
expectation did not come true. Examples of cities where there are still strong concentrations
of work in the centers and where public transport plays a dominant role in commuting
are London and New York. Although indeed a certain trend can be observed of
deconcentration of work places towards the city fringes, this tendency remained limited in
many cases. In addition, it appears that in those cases that jobs moved towards the
fringes of the cities, this often led again to concentrations of employment. For this
phenomenon, Garreau (1988) introduced the concept of edge cities. Researchers such
as Henderson and Mitra (1996), Lucas and Rossi-Hansberg (2002) and Glaeser and
Kahn (1993) developed new models for this purpose, in which explicit attention is paid
to agglomeration advantages for firms. The center is ultimately restricted in its growth
by high transport costs from the residential areas located further away. The formation of
sub-centers allows the further growth of urban economies. The underlying trade-off
concerns the advantages of a reduction of total transport costs of employed persons and
the lower productivity that follows from the spread of economic activities in the region.
Agglomeration Advantages, Transport and Communications
Three mechanisms play a central role in agglomeration advantages: sharing, matching,
and learning (Duranton and Puga 2004). The essence of sharing is that indivisible
facilities with high fixed costs face high thresholds in terms of consumer demand
before they become affordable for private or public suppliers. Typical examples are
infrastructures such as airports, sea ports, highways, and railways. Sharing may also
relate to amenities that make the life of citizens more pleasant, such as, shopping
centres, museums, and hospitals. Further, sharing is relevant for human capital
formation via schooling. Clearly, the sharing mechanism applies to both manufacturing and
service sector activities. Matching is central in the case of heterogeneous products,
services and workers. To improve the quality of matches, producers and consumers
may go into intensive search activities and the costs of search clearly are lower when
there are many potential matching partners in the vicinity. Finally, learning has to do
with knowledge spill-overs between firms or persons that lead to an increase of
In a low-tech world, sharing, matching, and learning just depend on density of
households or firms within a limited spatial range. As transport technology improves
the spatial range of these mechanisms will increase. As a rule of thumb one may say
that when transport becomes faster, people increase travel distances in a proportional
way, so that total travel time remains more or less constant (see for example
Mokhtarian, 2004; van Wee et al. 2006). The increase in travel speed thus leads to a
larger number of potential sharing, matching and learning partners.
A major implication is that the traditional way of addressing urbanization economies
by looking at population density in a certain demarcated urban area is superficial, since
it ignores the costs of interaction and also the possibility of spatial interactions with
destinations outside urban areas. A recent trend in research in this domain is that
accessibility is used as a central concept (Melo et al. 2009), where accessibility is
defined as the potential for spatial interaction that incorporates both population size of
destinations and distances. One step further would be the explicit introduction of travel
times in such approaches. There is indeed a strong need for making the notion of
transport costs or travel times more explicit in the analysis of agglomeration
advantages. Exploitation of agglomeration advantages cannot only be achieved by making
cities larger or denser, but also by improving their transport systems.
The problem with many transport networks is that they tend to function rather well
between nodes, but congestion is high within nodes and at the interfaces between links
and nodes. City fringes are typically the place where in the real world the bottlenecks
are most prominent. Thus, metropolitan areas are usually the places with lower travel
speeds compared with other locations. Nevertheless, the accessibility of places within
metropolitan areas is higher than elsewhere, since the low performance in terms of
speed is more or less compensated by high densities. An important risk is that
congestion may increase in the course of time when transport investments would lag
behind, in particular since the value of time tends to increase with income, implying
that the time cost of transport follows an upward trend (Gunn 2001; Rietveld and
Vickerman 2004). An important consequence is that where transport costs tended to
have a downward trend in the long run, this is no longer evident in the future when
congestion problems are not overcome. Transport costs have declined strongly in
freight transport, and also in long distance passenger transport (air and high speed
rail), but this decline is not evident for short distance passenger transport in many cities.
A possible way-out may of course be the use of information and communication
technology (ICT). The last decade has shown an exceptional increase in the use of
internet, advanced mobile phones, and social media (Waters 2012). They have
dramatically reduced the cost of communication and improved its quality, so that the ‘flying
carpet’ metaphor is also particularly appealing here. No doubt these developments will
have affected the mechanisms of sharing, matching, and learning that are the main
formative forces of cities. Digital data banks showing vacant dwellings, or profiles of
job seekers have strongly affected the search processes on essential urban markets such
as the housing and labor market. Would this lead to major changes in the forces
stimulating the formation and continued existence of cities?
An important observation is that although the last decade has witnessed an extremely
rapid development in ICT applications, this is not the first time that large shocks take
place in communication networks. In a similar fashion telegraph and classical telephone
networks have meant revolutionary developments in the speed and costs of
communication. These developments have strongly improved the efficiency of command and
control activities in organizations and, hence, they have increased possibilities to run
international organizations in an efficient manner. This may well have reinforced the
position of a limited number of large cities as attractive locations for these large
organizations (Scott 2001). Another effect has been that better communication
improved the possibilities of specializing, serving larger spatial markets, and outsourcing,
thus leading to a sorting of activities where some cities tend to focus on command and
control, and R&D, whereas other places specialize in manufacturing and particular
services. Examples of sectors where ICT has a substantial effect on the location of
activities are call centres and the printing industry. And also the retail sector is
increasingly facing the effect of internet shops (Forman et al. 2009). A major result
is that improved communication indeed leads to changes in location patterns of firms,
and thus in urban systems, but the way in which this occurs is complex and cities may
be affected in rather diverse manners, some strongly losing, and others gaining in
This result probably also holds for the internet and social networks that have become
so popular in the third millennium. The statements by Cairncross (2001) that distance is
‘dead’ or Friedman (2006) that the world is flat make a strong case that the forces that
shaped cities have changed in a fundamental manner. Waters (2012) gives a nice review
of the huge developments in this dynamic domain, but he emphasizes that the spatial
effects are easily overestimated. For example, Barthelemy (2010) finds that geography,
and hence distance, still strongly matters in social media networks, since a substantial
part of interactions strongly depend on physical distances between the partners with a
gravity parameter of minus 2. Although social networks supported by the ICT driven
social media no doubt have the potential to broaden the networks of people, an
important part just supports and maybe reinforces the social interactions that take
place via movements in the physical world. Another example is the role of languages
and national preferences in the internet. Waters (2012) indicates that these play a large
role. For example, while MySpace has pretty much disappeared, its rise was
concentrated on the west coast of the United States, while Facebook dominated on the east
coast. Thus, there are still clear signs that distance related factors are relevant.
People and firms communicate on many things. Part of the communication can
easily take place via the internet or social media, and another part less so. In particular
the types of contact where trust, enjoyment and emotions are crucial still seem to
depend strongly on face to face contacts and this makes trips to meeting places
unavoidable. Telecommunication and physical contact are imperfect substitutes. And
also for the enjoyment of high quality amenities in a city, the internet cannot provide a
substitute. The buzz that is typical for urban contexts (Storper and Venables 2004;
Asheim and Coenen 2007; Arribas-Bel et al. 2013) and that attracts people to urban
environments can only be partly substituted by the internet.
Agglomeration Advantages and Transport Networks
Agglomeration advantages imply that certain highly accessible locations become very
attractive as a location for firms (and for workers since the high productivities lead to
high wages). Thus, we may expect fierce competition for land and hence high land
prices around these locations (see for example Fujita 1989, and Lucas and Rossi
Hansberg 2002). The high land prices affect land use on these places, so that it pays
to make it more intensive: higher labour/land and capital/land ratios. One of the ways to
achieve this is the construction of high-rise buildings.
Since Otis invented the safe elevator in the 19th century, offices have become higher
and higher. Sky-scrapers became the land marks of large metropolitan areas, such as,
Chicago and New York since the 1890s (Barr 2010). Much of the discussion on
changes in urban form has focused on horizontal transport (see Internal transport
within cities and urban development), but it has often been overlooked that vertical
transport must have been of equal importance. Without the development of safe and
efficient elevator systems urban densities would have remained much lower. Thus, the
metaphor of the ‘flying carpet’ applies undoubtedly to the use of the elevator in
This of course depends on the capacity of elevators. The use of elevators is free and
with sufficient capacity they are usually fast, which complies with the ‘flying carpet’
notion. However, elevator systems may also be crowded and that makes congestion
theories potentially relevant. However, pricing is not introduced in this context. A key
reason why, in general, the congestion problems in large office buildings are limited is
that the congestion problem is internalized. The owner of the building who also uses it
has an incentive to optimize the capacity of elevators given the demand of his
employees. And when an office building is rented the owner will notice that the user’s
willingness to pay will be higher with a non congested elevator system.
There is another element in high-rise buildings that deserves mention. Not only do
high-rise buildings contribute to the possibility to experience agglomeration advantages
in a metropolitan area since it allows higher overall densities in cities, but they also may
yield high agglomeration advantages within the buildings themselves. This holds for
both residents and firms, but this holds true in the first place for internal scale
advantages within a firm. For a firm the efficiency will be higher when the various
divisions are not located in separate buildings - implying the use of slow and
uncomfortable horizontal transport - but when related activities are concentrated in one large
building. Further, large office buildings hosting more than one firm may yield similar
advantages. In particular, when the high rise building hosts clients or firms among
which cooperation may take place in joint projects. Large office buildings thus provide
an attractive basis for matching, but also for learning and sharing. Thus, one would
expect that high (= large) office buildings will yield higher rents than low office
buildings. This is indeed what is found in Koster et al. (2012): high-rise buildings lead
to rents that are substantially higher than those of low buildings. This is an important
field of research, since it would directly demonstrate the contribution of the elevator to
urbanization advantages. On the other hand, it is not so easy to disentangle the
agglomeration effect from various other effects. For example, the view effect (higher
floors may be more attractive and thus a higher price has to be paid) and also the status
effect (having office in a really high rise building may have an additional value) have to
be controlled for. And also the quality effect may play a role since the higher the
building the more it pays to give it a superb but expensive entry. This is certainly a
promising field of research on the flying carpet theme.
It is not possible, however, to completely separate economies owing to vertical
transport in high-rise buildings from the state of horizontal transport. High-rise
buildings lead to large flows of people entering these buildings and this imposes substantial
requirements on horizontal transport. One direction is that sufficient parking facilities
should be available at short distance implying need for multi-storey car parks. Another
direction is that high-capacity public transport should be available at short distance.
This calls for a broader view on urban transport networks and their contribution to
agglomeration advantages. As shown in Table 2, several ranges of distances can be
distinguished where agglomeration advantages play a role. For example, for the
shortest horizontal distance of approx. 500 m walking is the obvious choice, as is
evident from many experiences. The relevance of this range can be inferred from the
fact that universities with the many encounters between staff and students typically aim
at this range of distance instead of having buildings scattered around a city. These are
also the distances one may have at smaller office areas. Larger office areas and
industrial parks may lead to longer distances up to say 2–3 kms. At the level of medium
sized cities distances may be as long as 20 kms and in large metropolitan areas
distances may go up to 100 km. Interurban transport will be beyond this level. For
each of these ranges typical most appropriate transport modes present themselves (see
Table 2). A more refined analysis of the relevant spatial ranges of matching, sharing,
and learning would be most welcome to find out in which main transport mode
investments are most needed.
A major step forward in the literature is provided by Rice et al. (2006) who go
beyond distances and into travel times. They carry out an analysis in terms of travel
time bands. They find that in the UK a doubling of the working population within a
range of 30 min travel time increases productivity by 3.5 %. A doubling of the working
population between 30 and 60 min leads to a productivity increase of 1.5 %. Between
60 and 80 min the increase is 0.5 % and beyond 80 min no increase is notable. A
specification of agglomeration advantages in this manner is of much help to assess the
potential productivity effects of improvements in transport infrastructure. The pattern
found by Rice et al. (2006) suggests that with the given transport technologies the main
productivity enhancing effects of infrastructure take place in ranges up to some 20 kms,
and that beyond 100 kms the agglomeration effects are very small.
We conclude that in order to reap the potential agglomeration benefits within
metropolitan areas it is essential to have adequate and tailor-made networks for distinct
classes of passenger transport. At the same time it is in metropolitan areas that the costs
of upgrading and new construction are highest, so that poses a considerable policy
challenge. ‘Flying carpets’ are not yet free…
Transport networks have experienced growth and decline for particular transport modes
during the last centuries. Cities too have their downturns and upswings, but for most
cities less dramatic cycles have been observed. From a long term perspective, each
phase of development in the logistical system during the last millennium was linked to
a specific type of city that successfully made use of the logistical opportunities. That
does not mean to say that cities that prospered in earlier phases completely disappeared.
A tendency can be observed of an urban ratchet effect: once cities have achieved a
certain level, they rarely entirely collapse given the variety of activities present that
make it less vulnerable and also given the sunk costs already invested in social
overhead capital. A major difference between the present stage and the preceding
stages of urban development is that the developments are less dominated by freight
transport; passenger transport has become often the leading part in this theme.
The main result of the long run decline in transport costs during the last century has
been urban sprawl. Nevertheless the position of cities has not weakened. The reason is
that metropolitan areas have just increased in spatial range, but they remain important
attractors of business investments. More recently, a tendency can be discerned that
central cities have become strong attractors of employment and population again. A
main reason—probably more in Europe than in the USA—is the increasing valuation
of the amenities, including commercial activities (in particular business services and
personal services) and cultural heritage for households and firms.
Communication technology has developed extremely rapidly during the last
decades. This has no doubt had a strong effect on the organization of production and the
choice of its location; nevertheless cities remained remarkably robust under these
changes. An important factor is no doubt that the substitution between physical
transport and digital communication is smaller than is often thought. Face-to-face
contacts remain essential for many types of interactions between people. And the best
way to enjoy urban amenities is still to just be there.
Transport infrastructure is relevant for a wide range of distances in cities. The trend
that transport becomes cheaper and faster is most probably difficult to maintain within
metropolitan areas. For the effective use of agglomeration advantages a properly
functioning transport system is a sine qua non. Of special relevance is the vertical
dimension in transport. The use of elevators in buildings starting just before the
beginning of the 20th century will continue to have an impact during the 21th century,
if only since more and more cities and countries make ever higher buildings in order to
gain prestige. So the ‘flying carpet’ metaphor of low cost rapid transport and
communication is most telling for transport in the vertical dimension. This explains also the
emergence of modern mega-cities. For horizontal transport, substantial efforts will be
needed to safeguard accessibility within metropolitan areas.
Open Access This article is distributed under the terms of the Creative Commons Attribution License which
permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source
Andersson , A. E. ( 1986 ). The four logistical revolutions . Papers in Regional Science, 59 ( 1 ), 1 - 12 .
Arribas-Bel , D. , Kourtit , K. , & Nijkamp , P. ( 2013 ). The socio-cultural sources of urban buzz . Environment & Planning C (forthcoming).
Asheim , B. , & Coenen , L. ( 2007 ). Face-to-face, buzz, and knowledge bases: sociospatial implications for learning, innovation, and innovation policy . Environment and Planning C: Government and Policy , 25 , 655 - 670 .
Barr , J. ( 2010 ). Skyscrapers and the Skyline: Manhattan . Real Estate Economics , 38 ( 3 ), 567 - 597 .
Barthelemy , M. ( 2010 ). Spatial Networks . Physics Reports , 499 ( 1-3 ), 1 - 101 .
Brakman , S. , Garretsen , H. , & Schramm , M. ( 2004 ). The strategic bombing of German cities during World War II and its impact on city growth . Journal of Economic Geography , 4 , 201 - 218 .
Braudel , F. ( 1994 ). A History of Civilizations . New York: A. Lane .
Brown , D. M. ( 1974 ). Introduction to Urban Economics . New York : Academic.
Cairncross , F. ( 2001 ). The death of distance: How the communications revolution is changing our lives . Boston: Harvard Business Press.
Cheshire , P. ( 1990 ). Explaining the recent performance of the European Community major urban regions . Urban Studies , 27 , 311 - 333 .
Clark , C. ( 1958 ). Transport-Maker and Breaker of Cities . Town Planning Review , 28 ( 4 ), 237 - 250 .
Crafts , N. , & Venables , A. ( 2001 ). Globalization in History: a Geographical Perspective . Cambridge : Paper prepared for the NBER conference on 'Globalization in Historical Perspective' , NBER conference on 'Globalization in Historical Perspective', National Bureau of Economic Research.
Davis , D. , & Weinstein , D. ( 2002 ). Bones, Bombs, and Break Points : The Geography of Economic Activity . American Economic Review , 92 , 1269 - 128 .
Duranton , G. , & Puga , D. ( 2004 ). Micro-foundations of urban agglomeration economies . Handbook of Regional and Urban Economics , 4 , 2063 - 2117 .
Forman , C. , Ghose , A. , & Goldfarb , A. ( 2009 ). Competition Between Local and Electronic Markets: How the Benefit of Buying Online Depends on Where You Live . Management Science, 55 , 47 - 57 .
Friedman , T. L. ( 2006 ). The world is flat: A brief history of the twenty-first century . London: Penguin.
Fujita , M. 1989 . Urban economic theory: Land use and city size . Cambridge Univ Pr.
Garreau , D. ( 1988 ). Edge City . New York : Anchor Books.
Glaeser , E. ( 1999 ). The future of urban research: non-market interactions . Boston: Mimeo, Harvard University.
Glaeser , E. , & Kahn , M. E. ( 1993 ), Sprawl and urban growth. In Handbook of urban and regional economics . Amsterdam: North Holland.
Gunn , H. F. ( 2001 ). Spatial and temporal transferability of relationships between travel demand, trip cost and travel time . Transportation Research , 37 (E), 163 - 189 .
Hall , P. ( 1994 ). Squaring the circle: can we resolve the Clarkian paradox? Environment and Planning A, 579 - 594 .
Henderson , V. , & Mitra , A. ( 1996 ). The new urban landscape developers and edge cities . Regional Science and Urban Economics , 26 , 613 - 643 .
Karlsson , C. , Andersson , B. , & Stough , R. ( 2009 ). Entrepreneurship and development: local processes and global patterns . CESIS Electronic Working Paper Series 160.
Koks , E. ( 2012 ). To what extent do harbors influence the economic performance in a region? MSc Thesis , VU University, Amsterdam.
Koster , H.R.A. , Van Ommeren , J.N. , Rietveld , P. ( 2012 ). Is the sky the limit? An analysis of high-rise office buildings . Mimeo.
Krugman , P. ( 1991 ). Geography and Trade . Leuven: Leuven University Press.
Levinson , D. ( 2005 ). The Evolution of Transport Networks . In D. Hensher (Ed.), Handbook 6: Transport Strategy, Policy and Institutions (pp. 175 - 188 ). Oxford: Elsevier. Chapter 11.
Lucas , R. E. , & Rossi-Hansberg , E. ( 2002 ). On the internal structure of cities . Econometrica , 70 ( 4 ), 1445 - 1476 .
Melo , P. C. , Graham , D. J. , & Noland , R. B. ( 2009 ). A meta-analysis of estimates of urban agglomeration economies . Regional Science and Urban Economics , 39 ( 3 ), 332 - 342 .
Mokhtarian , P.L. , & Chen , C. ( 2004 ). TTB or not TTB, that is the question: a review and analysis of the empirical literature on traveltime (and money) budgets , Transportation Research Part A: Policy and Practice , 643 - 675 .
Rice , P. , Venables , A. J. , & Patacchini , E. ( 2006 ). Spatial determinants of productivity: Analysis for the regions of Great Britain . Regional Science and Urban Economics , 36 ( 6 ), 727 - 752 .
Rietveld , P. , & Bruinsma , F. ( 1998 ). Is transport infrastructure effective? Transport infrastructure and accessibility: impacts on the space economy . Berlin: Springer Verlag.
Rietveld , P. , & Bruinsma , F. R. ( 2012 ). Transport and urban development . Amsterdam: VU University.
Rietveld , P. , & Vickerman , R. ( 2004 ). Transport in regional science: The death of distance is premature . Papers in Regional Science, 83 ( 1 ), 229 - 248 .
Scott , A. J. ( 2001 ). Global city-regions: trends , theory, policy. USA: Oxford University Press.
van Wee , B. , Rietveld , P. , & Meurs , H. ( 2006 ). Is average daily travel time expenditure constant? In search of explanations for an increase in average travel time . Journal of Transport Geography , 14 , 109 - 122 .
Waters , N. ( 2012 ). Social networks and regional science . In M. M. Fischer & P. Nijkamp (Eds.), Handbook in Regional Science. Berlin: Springer.