GaWC Research Bulletin 341

GaWC logo
  Gateways into GaWC


From Connectivity to Metropolis Power: Measuring National City Networks with METROX Methodology - The Case Of Poland

M. Bańczyk*


A search for hierarchy does not need to be always misguided. Even the global city network reveals some hierarchical tendencies. On the national scale the hierarchy is even more pronounced while network qualities of the national city system are also observable and influential. The METROX methodology aims at analysis and presentation of the system of cities on a national scale. It is based on GaWC business connectivity and limited to the area of one country treated as universum. Within this spatial limit, METROX enables the measurement of the hypothetical Metropolis Power of cities proposed as the main vital force and motive power responsible for cities ability to produce excellence. The paper introduces METROX theoretical basis, its mechanism and the results of the pilot application in Poland and compares the outcome to the findings of GaWC/CASS research.


The basic and overwhelming reason for investing time and energy in development of any new methodology should not be but answering questions so far unanswered or at least – answering better the questions addressed so far. Such was the basic purpose underlying the creation of METROX or a new connectivity-based method for analyzing and presenting relative positions of cities within one national system. That it is organized in a ranking – hierarchical by definition – does not necessarily eliminate its network focus. Network relations are as fundamental for the METROX approach as they are for the GaWC main research lines [and we clearly hold these main lines even against GaWC-present critics – see Smith and Doel RB 221]. This however, does not necessarily mean that hierarchies have to be media hypes nor that all searches for the hierarchy need to be misguided. We simply wanted to find and/or improve clarity of answers to the following questions:

  • which cities of a given country are interconnected and how strong
  • which cities of a given country/state (e.g. Poland) are closer to the world city profile and which are not
  • how is the power of cities distributed within the country
  • where are the highest chances for living a fulfilled life
  • how does the national network compare to the global
  • where to set up the first Ferrari Showroom or Gucci Boutique
  • which cities we should keep a close look on in the next few years, on which – not necessarily or how appearances from outside can cheat even the serious researchers

Apart from these, we intended to come up with a cohesive method of looking into a system of cities that is not necessarily part of a greater world network but performs as a network on its own. Our interests were primarily focused on possibilities to monitor a medium-sized economy which a system of cities can be assumed. The consequence of such a choice is a narrower yet deeper vision: we cover just one country instead of the globe, on the other hand – we are able to consider for instance TWO more dimensions to a city than just pure network connectivity. Establishing the relative positions of cities within such 3-dimensional analytical space and measuring the strength of links between them could lead us in turn to conclusions not only on cities themselves. We aimed at a method of analysis of states by their cities.

Whether we like it or not (or – to be precise and honest – despite the fact that we DON’T like it), we are still trapped in a world marked by the obsolete concept of states. As long as it remains so, we need to deal with it. Measuring national city systems as represented by METROX is not an attempt of hijacking the development of cities back to the tyranny of the state. It is rather a sad yet necessary procedure of bringing some city-related reasonable contents to the state-related classification and comparisons. The basic logic here is pretty straight forward: if anything like national state exists in the observable economic reality at all, its level and development perspectives depend in largest part on the position, connectivity and potential of its cities. Or simply stated: the strength of cities and their number in a given state speak well of its overall position. That many strong cities is better than few weak cities is probably well known. What we claim is a bit beyond that common wisdom: we venture to assume that many very strong cities is better than very many strong cities or even – that few very strong cities is better than many strong and very many so-so strong cities. Which is just about exactly against the contemporary social dogma of balanced development and the very idea of socioeconomic convergence. This way or another, the same basic question pops up: what does it actually mean a strong city? And this is how we arrive at the key challenge: to propose a concept of metropolis power.


In general, METROX is a methodological product of network approach to city research as pioneered and further developed by GaWC. This is exemplified by several concepts brought up and promoted by prominent researchers at GaWC [Taylor RB177, RB229, RB300] and that were adopted as fundamental in METROX. These includes several key statements, such as:

  • there is no such thing as a city, there’s only a network of cities
  • certain types of professional services (APS) are critical to city network research
  • exchange, not – accumulation, is the main modus operandi of cities
  • cities are more important than states

In addition to these similarities, METROX clearly acknowledges both rationale and application of connectivity as a prime quantitative articulation of the city network. This is to say, METROX methodology is not only based on the GaWC-induced network city theory but also owes great deal to and openly stems from world city network connectivity research as proposed and conducted by Taylor et others. At the same time, METROX was created to demonstrate and incorporate into computation the few significant differences in both vision and method that separates the pure globalist approach from a standpoint that we claim to represent and that for our own use we call global metropolist approach.


In the early days of GaWC, the actual interpretation of connectivity-based map of global cities appeared to convey the idea that the whole exercise more or less adequately reveals processes critically affecting the allocation and structure of the overall creation of progress and wealth in the world. Such an approach seemed to be evident behind the first and the second connectivity research editions while the essays spoke of no less than “new metageography of the world”.  It also seemed to be clearly reaffirming the standpoint of frequently quoted classic works on the concept of “world city”, most notably Friedmann’s, and Sassen’s,  which implies not only that certain cities reached their new heights but also that global cities become dominant civilization-forging power in the contemporary world. The recent connectivity reaearch comments are way more restrained and even openly reduce the assumed importance of connectivity-based world city network. According to this much more narrow and reserved view, connectivity is supposed to measure no more than “one process of city development: servicing of global capital”.

The approach adopted for METROX ranking is by all means closer to the original, hence – multidirectional and generally significant – interpretation of the world city network. In this regard, we stick strictly to the thesis from the previous paragraph: cities are more important than states hence they perpetuate and lead the march for better and more fulfilled life, whether corporate, institutional or private. Therefore, it’s crucial (and commonly expected) to measure not only ‘one process’ yet the very essence of the citiness of cities which is also said to be the major ‘motive power’ and value generator of the world. This statement is a starting point from which we have departed into a chain of questions and conclusions. We have landed upon a standpoint that differs from GaWC connectivity research position in the three following points:

1) The citiness of cities we see wider than just connectivity of business processes yet we uphold the predominant importance of the network connectivity. In a shortcut: we call a class of top world cities GLOBAL METROPOLISES, we claim there is a basic monadic METROPOLIS POWER that can be observed but only in partial manifestations in very many aspects. METROX intents to collect these factors and retrace (recalculate) the basic motive power underneath them. Of 5 types of cities [Beaverstock, Faulconbridge, Hoyler RB 322], we claim 4 (‘political city’ excluded) to stem from one and homogenic foundation of excellence being both source and consequence of the basic motive power we coin ‘metropolis power’, METROX methodology aims at gauging it as it appears in real life – in fragmented form – and put back together.

METROPOLIS POWER is as monadic or even more monadic and simple to imagine a parameter than GaWC connectivity. Only – its more complex to compute since besides the main factor – connectivity – it incorporates two more dimensions: a) power and profile of economy and b) scope of life options. However, as will be shown below, we had to modify even the connectivity formula itself.

2) There has always been a question whether APS articulate or constitute the network of world cities. The last standpoint of P.J. Taylor presented during our 2006 session was not as sharp as of now [Taylor RB300 revised]. If so, we clearly differ. We consider APS markers, NOT: makers of the network. Although this issue is so important and mind-provoking that it calls for a separate RB (which will follow soon), we have decided to provide short set of arguments to support the “APS-as-network-markers”.

The real network is made by the global flows of 3 most important resources: people, capital and concepts as well as by the global flows of solutions (expertise, technologies, products) made of these global resources. These flows are not limited to the APS nor are they limited to the corporate sector as such. APS are crucial but not because only they create so much, rather because they service all those who really create that much, if not all. Basically, even the service sector is not the best platform to trace the network flows – some services of that kind are not necessarily outsourced (in-house designers), there are also services of little value added of material character (repairs). That’s why we rather concentrate on those industries that produce high, immaterial value added which is actually very close to the concept of creative industries. Regardless from the exact classification and arguable scope of creative industries, we tend to believe that the role of APS, say: a global ad agency, in the world city network of flows boils down NOT to the exchange of processes between the very offices of this ad agency in the world but to the exchange of processes between the offices of all these global or international companies this agency serves but we can’t reasonably count, trace and locate due to their number. This is why APS offices are good proxy for the process flows not of their industry but of all other industries they service.

3) It would not make much sense to measure the positions of cities in the NATIONAL system only through GLOBAL connections. We claim that being connected or – being a node of a network – improves critically city position even if it is a national network. This is so of course, only if the national economy system is large enough to provide scale for any network at all. Denmark, for instance, despite high level of economy is nothing but one metro area of Copenhagen. Poland, on the other hand, certainly provides scale for some network of national cities. Following such appreciation of networkness per se (not only global networkness), we extrapolated its function from global to national level. In other words, we scale global connectivity down to national connectivity in which we treat national level as universum. As mentioned above, it requires some calibrations of the original GaWC connectivity formula.

In original form it was meant for the scale of the world and as such did not provide for situation in which there is only ONE HQ of an important APS and there are no offices otherwise. hence, the entire country is serviced from this sole HQ. This is hardly imaginable in global scale but certainly true for the scale of a country, even medium size country like Poland (population-wise: 6th largest of Europe, 32nd of the world, 18th economy of the world). The one-central-office situation does not apply to the entire scope of businesses but there are APS niches when entire country can be practically run from one central office in the most suitable and it certainly SPEAKS of the role and importance of this city as a node in the flow of processes in the country. Simply said: the other locations appear as ‘movable’, each time dispatched from the HQ to the serviced city. Having acknowledged that, we have decided to modify connectivity as to incorporate points also for the sole location HQs. We have included also minor changes on the actual types of APS that correspond to the shift form global to national perspective, for instance: we counted separately BTL and ATL agencies, we added business insurance companies but very carefully treated banks as normal, regular bank establishments servicing individual customers can’t be taken for important network marker as opposed to investment banking offices.

In general, these modifications combined with data collection problems and the necessity to incorporate two more dimensions, has made the whole exercise originally scheduled as GaWC connectivity country scale version a fully blown research that has taken about 4 months and some 28 versions of the computing formula until we reached the satisfactory complexity-adequacy-clarity balance and versatility of the calculation engine. In the present form, it is basicably applicable to any country featuring: 1) national system of cities, 2) national statistics providing critical data on city or at least large urban are level.


Once upon a time we have claimed that a phenomenon of a city is composed three different yet mutually interfering and interlinked types of interaction between the city and the world we eventually referred to as levels of urban impact [for more see: Banczyk, Achrem, Mroz, Smiejkowski, RB 240]. Following Taylor’s distinction between cityness and townness, we have attributed purely network/exchange quality to the first and called it a network level. It is well described by the global network of flows between otherwise non-territorial nodes. We have also attributed purely central place/agglomeration quality to the latter (townness) and called it a regional level. It is well described by the concentric areas of resources gravity to the central point inside. Eventually, we proposed the third level, representing an interaction of a city within itself claiming it different from ‘city’s own hinterland’ or a very small region. This type we called municipal or internal level. The core basic mechanism of multiplied value creation by so defined multi-level city or metropolis we have defined as optimization of three crucial resources: capital, human and ideas that are absorbed through the regional level and from the network, are then processed from simple resource form into the high-end solutions (products, services, professionals) and finally they are exchanged for value through the network level which induces growth of accumulated value on the municipal level. With some simplifications and readjustments, these are basically the theoretical standpoint we still maintain which makes it a natural methodological basis for METROX ranking.

The ranking is organized into three main constituents or parameters assuming individual values that will be used to described the overall position of a city in the 3-dimensional analytical space. On a certain level of approximation it is similar to the tree levels of urban impact described above, however with regard to regional and municipal interaction these approximations are highly conventional.  For this reason a much more concise and methodologically stable approach is to use the division into three constituent data categories. Each category is aggregated to a single parameter, and finally, these parameters are aggregated to the level of synthetic Metropolis Power index.

The first constituent (A) relates closely to the network level. We have classified it as: business connectivity and concentration of processes. It was originally planned to be nothing but GaWC connectivity applied to the universum of one state (national city network) but methodological proceedings proved that on the national scale the GaWC  formula needs very few but rather essential adjustments. Research-wise this constituent is the most demanding as the interactions that constitute connectivity or – flows pf processes between agents (companies) in the nodes (cities) take place very fast and in many different directions. We took into account the following types of services (in brackets – the number of companies in each type):

A1.          Business services - legal (37)
A2.          Business services - consulting (13)
A3.          Business services - advertizing ATL (10)
A4.          Business services – PR companies (9)
A5.          Business services - banks (20)
A6.          Business services – global real estate (8)
A7.          Business services – IT services (30)
A8.          Business services - insurance (14)
A9.          Business services – media houses (11)

The second constituent (B) was supposed to represent regional level or – in other words – the gravitational force attracting the key resources to the central place and processing them and hence - responsible for the regional impact. While we have maintained strong affinity of this parameter to the regional level as described, it is needed to say that the eventual set of available indicators we have selected to portray this constituent made us reconsider its profile. It is still connected strongly to the regional/gravitational model of value creation yet the most appropriate term here would rather be: economic power and profile. The following data constitute this category (source in brackets)

B1           List of 500, Rzeczpospolita 2007, total sales income (OM)
B2           GDP per capita for year 2005, PLN (GUS – Head Statistic Office)
B3           Number of employed in companies as a % of total population (2006, GUS)
B4           Investment in market services as % of total investment, (2006, GUS)
B5           Investment per 1 inhabitant (2006, GUS)
B6           Average salary (2006, GUS)
B7           Image of an „economic tiger” (dedicated research)
B8           Offices of media companies  (Pressbook database)
B9           Patents and copyrights filings (2006, Patents office and GUS)
B10         New companies with foreign capital (2003-2007, COIG, after Forbes)
B11         Student gravity (% of  total number of students in Poland, 2006, GUS)

The third constituent (C) as previously stated should in general measure the municipal or internal level which is the level and intensity of processes building an interaction of city within itself. This is basically where quality of both – public space and living both emerges and manifests. As pointed in our original concept though, the municipal level has relatively strong ties not to the regional but to the network level. Nevertheless, what must be pointed out here is that we do not aim at measuring the overall ‘quality of living’ conditions here. ‘Condition’ is specifically wrong term in this case because above all we claim metropolis to be a generator of life fulfilled and not: life as smooth and easy-going as possible. Living in a village can be very easy-going but to most daring and skilled individuals who present lust for life would find it good from far but far from good. This is why the main focus of our interest here is the richness and quality of options for fulfilling needs from rather higher layers of Maslow’s pyramid: cultural, recreational, aspiration-related. These options inspire and motivate humans rather than meeting the necessary yet ordinary conditions of living. This is why we do not consider traffic jams or access to drinking water here. One will suffer the traffic provided there is any magnet in the city to suffer it. It there isn’t, the lack of a traffic jam won’t make it up.  Having said that, we can introduce the name of this constituent as Quality and scope of life, yet we will point out frequently how it differs from mere living conditions. This constituent is based on the theory that the prime consequence of process density in a metropolis is the multitude of options. In the aspect of life quality it means the number of alternative, real and accessible options for a rewarding, fulfilling and inspiring realization of the moment which in common language would be simply stated as having a great time. This category constitutes the following data:

C1           Number of cultural options – movie theaters – shows in 2006 r. per 1000 inhabitants
C2           Number of cultural options – plays and concerts in 2006.
C3           Number of cultural options – exhibitions in art galleries in 2006.
C4           Lifestyle quality institutions (concert halls, operas, theaters)
C5           Modern commercial areas (multifunctional shopping malls) per ‘000 population, 2007
C6           Number of matches in highest leagues (soccer, basketball, volleyball, speedway) in season
C7           Number of hotels, 5* (3 pts) and 4* (1 pt) []
C8           Emigration balance (2004-2006, GUS, incl. suburban areas)
C9           Tertiary education according to High school ranking of WPROST magazine   
C10         Airport connection quality (on the basis of total passengers in 2007)

Within each category the results were computed by decimal standardization and applying the quotient scale. Subsequently, the average decimal for each city was calculated and presented as percentage of the leader’s score (1). The final score of the cities for each constituent of metropolis power as well as for the aggregate results is presented in very similar form to that of GaWC relative connectivity (i.e. the results of consecutive cities in the ranking are referred to the lead city)

As stated before, the typical engine of a metropolis brings together all three levels: metropolis draws resources from the region, processes them within municipium, then releases a small amount of the processed resources back into the region, Most of the resources focus within the metropolis building up a critical mass needed for a new solution. The new solution is then sent into the metronetwork. At the same time the metropolis receives other new solutions from the network. Metropolis accumulates the global ideas, that is new solutions – both own and those acquired from the metronetwork. Regional and municipal resources are processes and multiplied via network exchange.

In order for the network exchange mechanism to be shown in figures on a national scale we need to significantly modify Taylorean connectivity. The pure connectivity formula allocates score not the cities hosting just corporate HQs but to the cities hosting corporate HQs of companies with possibly many other corporate locations (branches, divisions, offices). This assumption makes perfect sense in a global scale, however if instead of the globe we take the universum of even a medium-sized country, then it becomes clear that at least some APS services are very well distributed from just ONE location being a HQ and a single national office at the same time. In pure connectivity formula each such a case equals to zero, i.e. it does not enhance city’s connectivity. Technically – the city indeed is NOT institutionally connected to anything, it is just a source of radiation, or – a generator of temporary connections – to companies in other cities that are all run from one location. But this is exactly what we intend to mode. The key variable measured while modeling a city network is not the STRUCTURE but the PROCESS FLOW. The structure of institutions (APS locations) is just an indirect platform of research from which to conclude on the direct one: processes. Which speaks in favor of incorporation of single-office-APS functions into the vehicle of measuring process flows between cities on national scale.

This notion leads to small but significant adjustment of the GaWC connectivity formula. It technically means weighing the network-only calculated results by the index of concentration of HQs and offices within a given city that is in turn computed as a simple sum of points per location. Mathematically speaking, the outcome of such multiplication is second-degree product of network connectivity times concentration index.

Now, we may enter the final stage of computation which is: to describe the basic city strength  or a motive power behind the metropolis. Hereby, it appears fundamental to reaffirm our standpoint on the basics: we claim that the universal approach to the phenomenon of world city network and contemporary city-driven world can go beyond connectivity and still retain its universalism and monadic profile. We claim that various data observable in the top-class world cities are just various symptoms of the same primary force that also breeds connectivity and that we coin metropolis power. This is the main motive power of cities. All we intend to do either through computation of connectivity or Metrox aggregate index is to retrace the symptoms of this basic metropolis power. We do not sum or level out the selection of assets, we restore the motive power that creates and processes the assets.

Since metropolis power does not constitute an average of resources, but instead proves to be a causative phenomenon, it is highly productive. Already this calls for a multiplication formula the outcome of which would mathematically be a product. aggregation of constituents A, B and C. Hence should be based on the mechanism of combining regional and municipal resources and exposing them to outside influences through connectivity. Due to somewhat overlapping selection of data and  imperfect interpretational boundaries of categories B and C, (demonstrating Quality and scope of life and Economic power and profile) these two constituents cannot be explicitly and sharply attributed to municipal and regional levels. In order to avoid such a uncertainty, it would be better to use a region-municipal complex of data as the representation of wide variety of both ability to attract regional resources or develop internal ones. This suggest in terms of logics an alternative which corresponds with the mathematical formula of a sum The final value though is produced and exposed onto the network again by the network influence (Constituent A). Altogether, the resulting scheme would imply summing the constituents B and C and to multiplying the sum by the constituent A (Business connectivity and concentration of processes).

Such a result, however, would be mathematically a third degree multiplication because the adjustment of the connectivity formula has already applied process concentration as a multiplier. Further multiplication would make the formula too ‘spiky’. On the other hand there is no serious alternative to such modified construction of the connectivity formula if we want to take into account the density of business processes incurred by single-location APS companies that are quite frequent the case of medium-sized countries. For this reason, in order to avoid the overmultiplication and retain the national scale connectivity adjustment we need to reduce degree of multiplication only at the stage of final aggregation of all constituents into the ultimate index of METROX metropolis power (Mx). This can be done by using a square root of A (Business connectivity and concentration of processes) in the aggregation formula instead of the straight A value.

Hence, the final formula for Mx (METROX) index is as follows:

Mx = √A x (B+C)


Mx – aggregated relative METROX Metropolis Power Index
A – relative value of Constituent A Business connectivity and concentration of processes corresponding strictly with the network level of urban impact
B – relative value of Constituent B Economic power and profile corresponding roughly with the regional level of urban impact
C – relative value of Constituent C Quality and scope of life corresponding roughly with the municipal level of urban impact
The construction of METROX index clearly suggests a product of the sum of municipal and regional resources times the square root of national-scale-modified connectivity as the best mathematical representation of the hypothetical monadic Metropolis Power.


The first research carried over within METROX methodology as described above took place in Poland between March and July 2008. It comprised data mostly from the secondary sources (predominantly – national statistics or GUS) with just one exception (constituent B – image of economic tiger, dedicated research, n=800).  The data reflect the period 2005-2008. The title of the main sponsor of the whole exercise was assumed by NORDEA bank.

Initial selection of cities for the research was based on very simple criterion of Poland’s administrative system so that to obtain comparable and cohesive data.  We investigated 18 main cities being capitals of the NUTS-3 administrative regions (voivodships). There are 3 special cases in this selection. The dual capitals of Kujawsko-Pomorskie region: Torun and Bydgoszcz were considered separately as there is still too little evidence to account for a duopolis here.  Gdansk together with the neighbouring cities of Gdynia and Sopot on the other hand were analyzed as one functional entity named Tricity as it reflects better the real life. Finally, we applied an aggregated approach to the Silesian capital of Katowice and all the cities and towns within the Silesia conurbation. It resembles Ruhrgebiet to some extent though it’s smaller and more concise plus on top of it – it is recently undergoing attempts of formal integration under the name of Metropolia Silesia.

Eventually, we have calculated results for each of the constituents (A, B, C) and presented them on the same scale (Charts 1-3)

Chart 1: The hierarchy of Poland’s regional capitals by NORDEA METROX Business connectivity and concentration of processes (constituent A). Source: NORDEA METROX.2008.

If there was rather little surprise that came with the actual ranks of the cities (little yet still – there was some, like the borderline situation of Lodz), no one had predicted so steep differences especially between the leader and the other double-digit contenders. Network-wise and – more significantly 9 in terms of concentration of process flows through single HQs, nothing in Poland comes even distantly close to Warsaw. One should point out that this discrepancy between Warsaw and other cities is even (though not very much) bigger than predicted by the relation of global connectivity [Taylor and other authors RB 333 – and other results of GaWC/CASS 2008]. Bigger than expected differences occur also in both other individual categories and – eventually – in the aggregate.

Chart 2: The hierarchy of Poland’s regional capitals by NORDEA METROX Economic power and profile (constituent B). Source: NORDEA METROX.2008.

Since most of the data representing economic power and profile were normalized by population (brought to per capita terms), the distance of challenger cities to Warsaw is not as striking as in the previous chart. The economic position number 2 in Poland is commonly attributed to Poznan, yet neither its distance to the leading Warsaw was believed to be so wide, nor its advantage over Krakow was believed so narrow. The use of profiling data which in general means measuring proximity to the STRUCTURE (and not: gross volume) of a high-value-added service-based economy of a world city has also some impact on this chart. In terms of simple GDP per capita, the city of Poznan outperforms significantly any other except Warsaw but its advantage does not translate into better economy structure. This will be addressed further in the final part of this section. Once again, the overall differences between upper and lower parts of the table are steeper than a common wisdom would suggest.

Last chart presenting results of an individual Constituent (C – Quality and scope of life) reveals also some surprises as traditional wisdom reserves Kraków as the best place to live. It does not quite seem so when the quality and number of options are concerned. The strongest Metropolis takes it all. Again.

Chart 3: The hierarchy of Poland’s regional capitals by NORDEA METROX Quality and scope of life (constituent C). Source: NORDEA METROX.2008.

Having computed and briefly analyzed the three constituents of the total Metropolis Power index, we can advance to the grand final (Chart 4)

Chart 4: The hierarchy of Poland’s regional capitals by NORDEA METROX.2008 Metropolis Power Index. Source: NORDEA METROX.2008.

This final picture shows a general hierarchy that probably could have been expected. However it also shows gaps between the positions of cities that were quite unexpected as it does not correspond with any of the single city performance indicators officially published. The only exception is GaWC connectivity.


As briefly mentioned above, we claim that cities demonstrate a specific vital force or motive power we name Metropolis Power that manifests in certain forms which analyzed individually can be and have often been misinterpreted. We believe that such a monadic Metropolis Power is a common factor behind vast majority of the variables describing city position and its development perspectives. This approach is marked by a relevant resemblance to the standpoint of some cognitive researchers on intelligence: there are many specific types of intelligence observable in Wachsler III psychometric tests but they are all related to ONE common general intelligence factor accounting for majority of differences between targets. So it seems with the cities: there can be discrepancies here or there but eventually the highest form of human settlements known to date, or a Global Metropolis, takes it all: it produces world excellence and offers highest chances for a fulfilled life. Fulfilled – NOT: unproblematic and easygoing or safe. High chances for lifetime achievements mean also high risk of lifetime failures and so it must be. High dynamism and variety of real options, or in other words – high density or processes ocurring per each cubic meter of a metropolis also means that life gets at times inconvenient. But so does for instance deer hunting in the areas known to be populated by deer. These areas are also known to be frequently crowded by hunters which turns inconvenient for each of them, because of the competition, risks of being shot and so on. All these risks and inconvenience can be avoided if one goes hunting to, say, Times Square: no other hunters competing and cops for safety everywhere. The only problem is – there is no deer to be scored there. And in deer hunting one is supposed to kill a deer no matter what effort instead killing time effortlessly.

Metropolis is able to produce best chances for a successful achievement hunting frequently referred to as fulfilled life, because it maximizes chances and possibilities by multiplying the resources and abilities. By Metropolis Power we define city’s own ability to create such an environment for fulfilled life, i.e. to attract best human, capital and ideas, process these resources, transform the outcome into solutions and exchange these solutions with other cities. In other words, Metropolis Power determines city excellence expectancy. It describes a theoretical chance that a city creates the best solution in a given area, for example, the first most modern car in a given class in Poland, the most cutting edge data transfer technology, the brightest star of contemporary music or the first global solution in Poland in an area that was poorly developed to date.  It shows the chance for many commercial and scientific breakthroughs for a city, including the most fashionable gadgets. Of course, such a hypothetical excellence expectancy does not fully describe the perceived life image. The determination of the inhabitants or a single remarkable individual can surely withstand the force sucking away capital, ideas and talents in a totally different direction. In this way exceptions, enclaves, areas of discontinuity appear on the civilization map. However, the end of their importance and very existence is most often just a question of time. In the long run, it is all depending on Metropolis Power [compare with the of multiple city classifications in Duarte, Ultramari RB 334], the most vital force of our civilization of the early XXI century.

METROX methodology, as metioned before, aims at retracing and recalculating the very integral level of this power from its fragmented observable effects scattered here and there in the analytical space. The results portray the national system of cities in Poland in a new way. There is no other report methodology offering similar sharpness of vision on gaps between the leading (and Poland’s only) metropolis of Warsaw, its five challengers (possibly being metropolises in the making) Krakow, Poznan, Wroclaw, Silesia and Tricity or Trojmiasto, the borderline case of Lodz, and the following cities with no signs of metropolis formation. METROX ranking provokes questions but also brings clearer answers to the existing ones, such as: why Warsaw concentrates about 70% of Poland’s total demand for new office space. Or why it advances fast year after year without having any reasonable land transportation connection like highway in any direction. Or why it attracts most of the young professionals, artists and other talents despite the long established image of Poznan as economic and financial center or Krakow as Poland’s capital of cultural offer. No single factor analysis explains these facts. No ranking other than by METROX Metropolis Power has demonstrated so far quantitative way of looking into the city hierarchy on national scale taking into consideration the newtorkness as the most important constituent.  The only exception, as pointed out previously, to this is of course GaWC connectivity.

The results of 2008 GaWC/CASS global urban study cover 5 Polish cities: Warsaw, Krakow, Poznan, Wroclaw and Lodz. It can be noticed that in comparison to METROX, two cities are missing: Silesia and Tricity. The most likely reason is that neither of these two is actually a single city in administrative sense, both are rather functional polycentric metropolitan areas consisting of three (Tricity) or many more (Silesia) cities and towns. Krakow assumes 24,34% of Warsaw’s score in GaWC/CASS 2008 [Taylor and other authors 2009 – correspondence with author] and 21,23% in METROX.2008 [Pierwszy Ranking NORDEA METROX 2009]. The results for Poznan are respectively: GaWC 16,94%, METROX 17,04% while for Wroclaw the score is: GaWC 15,49%, METROX 15,40% and for Lodz it is: GaWC 3,92%, METROX 8,52% This shows a very high correlation between the two optics, yet only in the real top positions (specifically: positions 2, 3 and 4, Warsaw’s score is incomparable between GaWC and METROX as relative ranking’s leader scores are incomparable by definition) while already in case of Lodz the discrepancy is more than 100%

It may suggest that GaWC global connectivity predicts well the city hierarchy on national scale as measured by Metropolis Power but only - for the very TOP cities or such cities with remarkable evidence of metropolis formation. It should be also pointed out that this hypothetical GaWC-METROX correlation works only on the METROX aggregate level, or between GaWC connectivity and METROX  Metropolis Power Index, and not – between GaWC global network connectivity and METROX network connectivity (Constituent A Business connectivity and concentration of processes, Chart 1.) This, in turn, may suggest that national piece of the global network is not exactly followed by the very national network. Which brings is to a remarkable conclusion, that a national city system features a global-connectivity-related hierarchy with strong country-specific network tendencies.

At the end, it seems interesting to portray these national network tendencies in some more details than just results of city scores as in Chart 1. The most important and interesting part of the network is its networkness or the pattern and intensity of the individual connections between the nodes. The map of such interconnections for Poland’s top six cities as per METROX ranking is presented below (Chart 5.)

Chart 5: Poland’s city network by business connectivity versus APS share and GDP size. Source: Author’s own calculation based on GUS and METROX.2008.

The vertical axis represents the city GDP share (percentage) of total value added generated in the city by high-professional services. With some minor approximation these services are the same as GaWC ‘regular’ APS (Advanced Producer Services) or services considered crucial in business connectivity calculations. Apart from the network-creative function, these services show by far the highest productivity of labor which makes them a good descriptive parameter of the modern structure of service economy. The horizontal axis shows simple GDP per capita in the city in real terms (PLN). The circles are sized accordingly to the total city GDP and the thickness of lines predictably reflects upon the strength of each business connection between pair of cities as computed purely on GaWC connectivity formula.

This picture can be interpreted as small but significant post scriptum to the main issues addressed in this article. We can clearly see that the leader (Warsaw) enjoys the highest connectivity represented by strongest connections with all other cities and the highest overall economic output represented by GDP per capita AND the most efficient structure of this output represented by the highest share of highly productive APS in GDP. The other positions though are not so straight-forward to interpret. Poznan appears the oddest case here: it makes part of Poland’s strongest single business connection (POZ-WAR) but despite this, shows the least favorable structure of GDP generation (low share of APS) AND despite this drawback is still able to produce second-highest GDP per capita topping over all other non-capital cities with better economy structure (higher APS share). This could well suggest that business connectivity has crucial role in determining the economic output measured by GDP but basing such a conclusion on the single, curious case of Poznan is certainly premature. Nevertheless, relationships between these variables is intriguing and calls for further research on wider scale.

At the end, it should be pointed out that neither the NORDEA METROX.2008 ranking nor its underlying METROX methodology has any goal in reviewing Polish, or any other countries’ cities’ local governments. Its goal is to portray the map and position of cities within national networks because through such positioning we claim to say more about the given country than any classic macroeconomic analysis would. Once again, there is no such a thing as state economy, there’s only a network of cities that counts. But the world IS organized in states so if we have to deal with the still existing tyranny of national states as entities, let us refer to the states through city systems. METROX is one of the possible forms of analyzing city system on the national state level. 

Not every city should be or try to become a metropolis. The world certainly needs power generators and oases of a sweet laid-back nothingness. People both as private persons and in companies strive for a chance for the best life, even at the cost of risk, hurry, traffic jams, floods of occurrences. The best, not cheapest, nor just above the average. This is how ultimate well-being is made. It can be done only in metropolises - in points where resources are concentrated and where international exchange perpetuates production engine for solutions, national or global excellence. Such cities drag the whole countries behind them. Biology is one of the hidden yet important sources for METROX approach. It proves without doubt that there is no such thing in nature as balanced growth. Homogenous super-balanced structure is only possible for simple organisms, which often tend to be quite dumb, such as: sea sponges, algae, etc. Higher organisms require higher level of specialization or organs, tissues and cells.

Metropolization described as the process of turning cities into metropolises has a crucial social [Jalowiecki 2009], economic [Gorzelak 2009], spatial [Gzell 2009], scientific [Olechnicka and Ploszaj 2009] or even existential [Banczyk 2009] impact on our lives. The vertical gap  between the level of cities in one country are often bigger than the respective horizontal gaps between the countries. This phenomenon is a topic for another research and article as it gathers speed especially in our times of vanishing political borders and technological change. In this respect, the rise of global metropolis is certainly signum temporis. Yet there is also an aspect of Metropolis Power that seems to win more than just a flavor of the century prize. It seems universal and eternal. After all, the quest for metropolis has been present throughout the whole millennia or in other words every era had its Rome. So it continues nowadays and there is rather little chance it will ever change. Why? Because metropolis means a multiplied life, an escape from death.


Bańczyk, M., Achrem, T., Mroz J., and Smiejkowski, D., Pushing the Taylor Materialist Theses Forward: Concepts for the Nuclear Theory of Metropolis, GaWC RB 242 (A)

Bańczyk, M., Rankingi metropolii – jaka metodologia?, in: Czy metropolia jest jeszcze miastem, Wydawnictwo SCHOLAR, Warszawa 2009

Beaverstock, J.V., Faulconbridge, J.R., and Hoyler, M., Capitalism and the City, GaWC RB 322

Duarte, F., and Ultramari, C.,  Inflexiones Urbanas y Ciudades Globales, GaWC RB 334 (Z)

Derudder,B., Taylor, P.J., Ni,P., De Vos, A., Hoyler, M., Hanssens, H., Bassens, D., Huang, J., Witlox, F., and Yang, X., Pathways of Growth and Decline: Connectivity Changes in the World City Network, 2000-2008, GaWC RB 310

Gorzelak, G., Metropolizacja a globalizacja, in: Czy metropolia jest jeszcze miastem, Wydawnictwo SCHOLAR, Warszawa 2009

Gzell, S., Architektura metropolii, in: Czy metropolia jest jeszcze miastem, Wydawnictwo SCHOLAR, Warszawa 2009

Jalowiecki, B., Spoleczne oblicza metropolii, in: Czy metropolia jest jeszcze miastem, Wydawnictwo SCHOLAR, Warszawa 2009

Jeney, L.,  Sectoral Background of Urban-Rural Economic Development Inequalities in Visegrad Countries, GaWC RB 337 (A)

Olechnicka, A., Ploszaj, A., Polska nauka w sieci? Przestrzen nauki i innowacji. Raport z badan, UW, Warszawa 2009

Parnreiter, C.  Global City Formation and the Transnationalization of Urban Spaces: Conceptual Considerations and Empirical Findings, GaWC RB 336

Smith, R.G., and Doel, M.A., Questioning the Theoretical Basis of Current Global-City Research: Structures, Networks, and Actor-Network' GaWC RB 221

Taylor, P.J., Cities within Spaces of Flows: Theses for a Materialist Understanding of the External Relations of Cities,  GaWC RB 177 (A)

Taylor, P.J., Ni, P., Derudder, B., Hoyler, M., Huang, J., Lu, F., Pain, K., Witlox, F., Yang, X., Bassens, D. & Shen, W. (2009a). Measuring the world city network: new developments and results. GaWC RB 300 (revised)

Taylor, PJ., Cities and States: An Elemental Reinterpretation of their Origins and Relations, GaWC RB 229 (Z)

Taylor, P.J., World City Networks: Measurement, Social Organization, Global Governance, and Structural Change, GaWC RB 333 (A)

Pierwszy Ranking Metropolitalnosci Miast Polskich NORDEA METROX.2008, Instytut Konkurencyjnej Ekonomii Regionow ( July 2008, Poznan, POLAND, published as: Ranking Miesiecznika METROPOLIA (Ranking by METROPOLIA Monthly, 9-10 (September-October), Poznan 2008


* Marek Bańczyk, founder and CEO of the Institute for Competitive Economy of Regions (IKER), Ph.D. dissertation defended at Poznan University of Economics, Poznan, Poland, email:

Edited and posted on the web on 23rd April 2010