Urban Competitiveness and US Metropolitan Centres

Abstract

In this paper, the urban competitiveness of 23 major US cities is examined. The methodology allows the obtaining of results that are not available to other methodologies. Several determinants of urban competitiveness are identified that are statistically verifiable and it is possible to show how both these determinants and the competitiveness of 23 US urban economies have changed during the past two decades. The results are presented in a manner that will be of use to urban decision-makers and planners. This study follows up on two earlier studies of this topic.

The study of urban competitiveness has taken off during the past two decades (see the Appendix, Table A2). At the outset, most researchers were committed either to Michael Porter’s (1990) focus on the competitiveness of nations or to Paul Krugman’s (1993) micro-economic focus on the firm. One of the authors remembers approaching a US foundation that was advertising itself as the primary supporter of research on competitiveness, with a proposal for a study of urban competitiveness—only to be told that cities had nothing to do with competitiveness. Shortly thereafter, the OECD had a conference on “Cities and the New Global Economy” (OECD, 1994) and this journal published a review issue on “Competitive Cities” (Urban Studies, 1999). Then Eurocities (2002) declared that Europe was a network of cities. Cities had emerged as the primary locus of the reality of competitiveness.

This focus on urban competitiveness is due to at least three factors. First, there has been a reduction in the policy capability of national governments as a consequence of, among other things, the emergence of supranational entities such as the European Union, a plethora of multinational and bi-national trade agreements, and the fiscal difficulties of most governments. Secondly, the fact that sub-national entities such as states, provinces and regions are torn between the demands of rural/agricultural and urban economic interests means that urban economies cannot rely on sub-national governments to provide funding for infrastructure projects and the policies that they require to enhance their competitiveness, given the competing demands of rural areas. Thirdly, in all industrialised countries, and many others, we have seen the reinvigoration of activist mayors and other municipal leaders who are aware of the necessity for them to take actions if their urban economy is to realise the future to which its residents aspire. In actuality, we can now say that the competitiveness of nations is little more than the composite of the competitiveness of their urban economies and that, for the firm, it is the actions of the local leaders that underlie and support its competitiveness. Cities such as Amsterdam, Copenhagen and Barcelona have made the argument that a nation could not be competitive in the absence of economic strength of its principal city. One researcher has written that

with the fading of the role of nation-states a new condition of international competition will be focused on the large cities (Matthiessen, 1990, p. 46).

Even in Porter’s famous ‘diamond’, three of the four points are contingent upon, or heavily affected by, policies enacted, and assets created, by local governments (Porter, 1990, p. 72).

More generally, competitiveness is important to local authorities because of the exposure their economies have to distant markets and competitors due to the cost- and distance-reducing consequences of technological change. Globalisation is also opening markets throughout the world to flows of goods and factors of production—labour, capital, technology and raw materials. Cities must be competitive to attract and to retain the talented and educated labour force that the industries of the coming years will require. Cities that ignore competitiveness enhancement will be destined to a future of marginalisation and stagnation.

From the outset, there have been different approaches taken to the process of describing and determining how competitive individual cities or urban economies (hereafter, cities) actually are, in relation to each other, and to the determination of the most effective policies that city leaders can implement. We do not argue that one approach is preferable to any other, but each does offer different insights to those who have to make decisions about urban economic strategic policies and initiatives.

In this paper, we will use a methodology that we developed for an earlier article in this journal (Kresl and Singh, 1999) to give an empirical analysis that will identify fundamental, statistically verified determinants of urban competitiveness. This approach generates results that are more objective than those obtained using either of the two other methodologies discussed here. Since this is the third such study we have done using this methodology, we are also able to show how the relative competitiveness of the 23 US MSAs (metropolitan statistical areas) has changed over the period 1977 to 2002, how the individual determinants have become important or have dropped off the list, and how some have changed from being negative to positive in their impact. Finally, while we offer a ranking of the 23 US MSAs, the more important and interesting possibility is that of demonstrating the areas in which each MSA has empirically verified relative strength or weakness.¹ Hence, the results we offer are designed to be of maximum benefit to urban leaders and strategic planners when charting the future course of the economy of their city. This methodology captures change from all sources and allows one to examine change over several years to an extent that is not possible with other methodologies.

Methodological Approaches to Urban Competitiveness

There are essentially four distinct methodological approaches to the analysis of urban competitiveness. Each has its own attractiveness and shortcomings.

In one approach, cities are rated in accordance with several variables that are asserted to be of importance. This is often done with the analyst making an assumption as to what economic specialisation or structure will be of most importance to a city in the contemporary economic environment. Recently, the most popular assumption has been that the competitive city must be a city of high-tech or research-intensive production (Maskell and Törnquist, 2001; and Lever, 2002). The cities are then ranked in accordance with a set of variables that logic and theory suggest ought to be determinants of a successful city with the preferred specialisation. The creative quality of the workforce is considered by some to be the essential element (Florida, 2002; Florida and Tingali, 2004; and Gertler et al., 2002). Others, such as the Competitiveness Institute (n.d.), argue that competitiveness is little more than clusters. However, experience informs us, on the one hand, that many cities can be competitive as centres of logistics, or culture and recreation, or administration, or niche manufacturing and, on the other hand, that clusters are effective in some industries or locations or in conjunction with specific policies adopted by local authorities, but not in others (Rantisi, 2004; Johansson and Paulsson, 2009; and Cumbers and MacKinnon, 2004). While offering very useful guidance to cities in certain situations, this approach is less useful as a general model of policy analysis and prescription.

A second approach has been that of benchmarking in which a set of cities is ranked in accordance with a large number of variables, without an assertion as to which specialisation is the preferred one. In this approach, it is argued that all of the variables are contributory to a city’s competitiveness without priority necessarily being given to one subset of them. The most ambitious of these benchmarking studies is that by Ni Pengfei of the Chinese Academy of Social Science (Ni and Kresl, 2010). Ni uses over 40 variables for 500 cities throughout the world, so it is possible to combine subsets of the variables that highlight some specific aspect of the competitiveness, or lack thereof, of any of the cities. An individual city can, of course, be examined in relation to a set of other cities that are similar or of interest.

One of the difficulties in benchmarking is the validity of some of the variables included. Not all convey what they are assumed to convey. In the US, one of the prime examples of this is the use of patents registered. These data are readily available, so there is a temptation to include them. However, in the US it is impossible, without examination of thousands of individual patent registrations, to know how the patent registered relates to the locus of research activity which it is purported to measure. A large firm has the option of registering the patent in the city in which the research was conducted that generated the patent (perhaps the result of a contract issued to a researcher based at a university), or in the city in which the research division is located (perhaps for taxation and regulatory reasons), or in the city in which the head office of the firm is situated.² This is, of course, just one of the variables often used. Furthermore, there is no way to demonstrate convincingly either which of the variables actually are determinants of urban competitiveness or how important any one of them is. Much benchmarking is therefore based on assertion rather than on objective analysis.

A third approach is that of a structural analysis. Negrey and Zickel (1994) classify cities according to positive or negative changes in population and manufacturing employment and put cities in one of six types, such as deindustrialising, innovation or new services centres. Markusen (1996) uses 4–6 characteristics to place cities in one of three types—hub-and-spoke, satellite industrial and state-anchored districts. Finally, Pollard and Storper (1996) identify three activity types of cities—intellectual capital, innovation-based and variety-based industries, according to their technology and manufacturing characteristics. It is anticipated that, by offering these urban economy types to city authorities, these decision-makers will be able to chart the most effective course. However, the signals given to the local authorities are very general and do not give specific strengths or weaknesses or most effective strategic planning guidance. In short, they tend to be descriptive rather than analytical and prescriptive.

The fourth approach is two studies of the competitiveness of a large number of US cities, we have published on previous occasions, utilising a rather different methodology. The first study was done for an OECD conference on globalisation and urban economies (Kresl and Singh, 1994) and the second was published in this journal (Kresl and Singh, 1999). The first step in this methodology is that of selecting a small set (three) of variables that could serve as general indicators of urban competitiveness, evaluating or ranking the cities included in the study in accordance with this measure of urban competitiveness. In the second step, a regression analysis is conducted and the resulting coefficients obtained enable one to gain a set of other variables, the determinants, which explain that ranking.

In the first Kresl–Singh study, we noted that urban competitiveness is a composite of economic and strategic determinants. The latter include governmental effectiveness, urban strategy, public–private-sector co-operation and institutional flexibility. We sent to mayors’ offices to gain information about these aspects and strategic plans were obtained. Unfortunately, the data developed from this exercise could not be put in a form that would make them comparable with, or as reliable as, the data for the economic determinants, so that and subsequent work has been limited to the economic determinants. These economic variables were, through the regression analysis, thereby verified as being statistically significant determinants of urban competitiveness. The third step is that of ranking the cities in the study in accordance with the variables that have been revealed to be determinants of urban competitiveness. It was then anticipated that city leaders, planners and decision-makers could use this analysis to gain an understanding of the actual strengths and weaknesses of their city or urban economy in relation to its competitors. The period studied for the 1994 paper was 1977–87 and for the second (1999) it was 1987–92. The analysis that is reported here is for the period 1997–2002. Due to the lag in availability of these data, we were not able to extend the research to a date later than 2002. Given the changes that we will show have taken place during the three sub-periods of 1977–2002, it is certain that additional changes will be found when a study is made using later data.

The Research

In the first two studies, the variables we selected to be indicators were the growth over a 5- or 10-year period of: manufacturing value added, retail sales and a set of professional services. Retail sales indicate the degree to which the city is experiencing growth in population and/or personal income and is considered by non-residents to be an attractive place to come for culture, recreation, shopping and, in general, an urban experience. Professional services are required if the city is to undergo a process of transition to an economy that will be suitable for the coming decades—designers, engineers, financial services, consultants and so forth—and they also indicate the extent to which the city has moved from being a traditional centre of manufacturing. Finally, manufacturing value added was used because during the 1980s and 1990s the revival of manufacturing and its transition from traditional to high-technology production, with higher value added, was one of the key elements in a competitive economy. The validity of these three variables as indicators of urban competitiveness was confirmed by discriminant analysis, in which the hit ratio, giving the percentage of ‘grouped’ cases that is explained correctly, was a very acceptable 80.0 per cent.

These three variables do not explain the ranking of MSAs, but rather are used to generate the ranking. They are selected because they are, as stated earlier, logically congruent with economic competitiveness and because of the strength of their discriminant loadings. This is the final step in the methodology in which statistical verfication rather than assertion is used. Causation of urban competitiveness will be given by the determinant variables, discussed later.

In the present study, we have used the same methodology, but we have reconsidered the use of manufacturing value added as an indicator of urban competitiveness and have decided to replace it, for two reasons. First, the revival of manufacturing is no longer as central to urban competitiveness as it was in earlier decades for almost all cities. The cities in our study were dominated by their strength in 19th-century industry; today, manufacturing is one of several strategic options for them. Secondly, in recognition of the fact that manufacturing is only one of the several activities that an urban economy could choose to have as one of its principal strategic options, it should be placed on the same level as other options, such as becoming a centre of information communication technology or bio-pharmaceutical activity, of logistics, of recreation and culture, or of administration. (For the messiness of this transition to a new economy, see: van Winden et al., 2007). Thus, we now treat manufacturing as a strategic option rather than as an indicator of competitiveness.

In its place, we have decided to use the growth in payroll per employee. Our aversion to using income or employment as general indicators of urban competitiveness is that neither captures accurately what is needed. Measures of income include retirement income, transfers and other items that do not relate to income derived from productive activity. Employment can be declining in a city in which a traditional labour-intensive industrial activity is no longer competitive, but in which a new high-tech ‘activity of the future’ is growing but not utilising sufficient labour to offset the decline in the other sector; or employment could be increasing or constant depending on the strengths of the two elements.

The growth of payroll per employee variable captures wages and salaries from all productive activity, per worker, and its rise over a period of time will give one indication of the degree to which the city or urban economy is experiencing higher productivity and can be considered to be competitive relative to other similar entities. In the environment of today, it is not conceivable that union pressures are forcing up salaries in absence of increases in productivity, and often not even then. Thus, the equation used in this study³ for the measurement of urban competitiveness is

Urban competitiveness = % Δ payroll per employee + % Δ retail sales + % Δ professional services

The period used for the growth of each of the indicators was 1997–2002. Using this equation, the ranking of 23 large US metropolitan statistical areas is presented in Table 1. The data are for MSAs rather than for cities since competitiveness attaches to the urban economic region rather than just to the city itself. The city and the surrounding area that makes up the MSA are symbiotic in that one could not exist effectively without the other. Earlier, the MSA was thought to be the city and its commuter ring of suburbs, but this relationship has evolved into one that is more complex than jobs in the city and residences in the suburbs and commuting is usually done in both directions. Each component of the MSA specialises to a degree in certain functions and together they combine to give, say, New York, Chicago or Seattle its degree of relative competitiveness. We have included 23 US MSAs with a population in excess of 1 million inhabitants, but have excluded Washington, DC, since the presence of the federal government makes its competitiveness not at all comparable with non-capital cities.

Table 1.

Urban competitiveness ranking of 23 US MSAs, 1997–2002

Ranking	MSA	Score
1	Miami	4.86262
2	San Diego	4.69515
3	Phoenix	4.25141
4	Kansas City	4.09651
5	Atlanta	4.08037
6	New York	4.02490
7	Dallas	3.91852
8	Houston	3.86829
9	Seattle	3.85999
10	Minneapolis	3.82088
11	Denver	3.80705
12	Tampa	3.79711
13	Boston	3.78234
14	Pittsburgh	3.75574
15	Los Angeles	3.72517
16	Chicago	3.70830
17	St Louis	3.70813
18	Philadelphia	3.67111
19	San Francisco	3.62440
20	Cincinnati	3.61277
21	Detroit	3.55411
22	Milwaukee	3.37481
23	Cleveland	3.30274

A few things should be noted about this ranking. First, there are some clear surprises in the placement of many of the MSAs. Favourites of some, such as Boston and San Francisco, do not fare well, while others such as Kansas City and Pittsburgh do unexpectedly well. In the case of San Francisco, this is because the MSA data do not include San Jose–Sunnyvale–Santa Clara, which means that Silicone Valley is excluded. This may not meet the requirements of some researchers, but the result for San Francisco, per se, is of interest if one wants to focus on that urban economy and its own strengths and weaknesses. Boston is a city that is highly regarded as a city of learning, or a ‘learning region’, but the linkage between this activity and overall urban competitiveness may be far more tenuous than one would assume. Additionally, a highly successful research sector may not bring benefits to the majority of the population, or have a dominating impact on the urban economy. Surprisingly successful cities, such as Kansas City and Pittsburgh, may be so because they are emerging from a period of time in which their economy was troubled and have been effective in responding to the challenges of that earlier period. They should be looked at for the keys to their even moderate success. For example, Pittsburgh has successfully managed a transition from steel production to electronic instruments and medical technology (Cohon, 2009; and Bloomberg, 2009), both of which are based on its excellent institutions of higher learning.

Secondly, this approach to evaluating cities according to their relative competitiveness stresses movement over time; that is, successful achievement of percentage growth in the three indicators—retail sales, professional services and payroll per employee. It accepts that this can be achieved via any of a number of paths or strategies and simply values improving the general economic situation of the residents of that urban economy.

Thirdly, the ranking does not privilege the economies that are favoured by most of those who advocate policies to enhance the competitiveness of an individual city or of cities in general—typically prescribing some aspect of high-technology production, learning, creativity, the information-communication sector, bio-pharmaceuticals, nanotechnology and so forth. Rather, it accepts the notion that the end-result of a competitive city should be that of realising the aspirations of the residents of that city—the particular mix of employment, income, leisure time, degree of income inequality and social exclusion, cultural and recreational facilities, and general urban amenities to which they aspire. (This is in conformity with the definition of urban competitiveness of the Global Urban Competitiveness Project, n.d.). The competitive city can be competitive as a centre of specialised manufacturing, logistics, culture and education, health care, specialised services and so forth, some of which have a solid linkage to innovation and creative thinking, but would not be celebrated as such by many consultants in this field (see Markusen and Schrock, 2006).

The Rise and Decline of a City’s Competitiveness

One of the enduring questions of strategic planners is whether a city is dominated by its geographical or regional location. That is, do all cities in a region rise or fall because of region-specific characteristics? This was analysed in our 1999 study and we present results here from our more recent analysis. The changes in position for each of the cities for 1992–1997 to 1997–2002 is presented in Table 2, with the cities grouped into five US regions: the Industrial Triangle, the Pacific Coast, the North East, the South and the Centre.

Table 2.

Changes in competitiveness, major US metropolitan areas, by region, between 1992–97 and 1997–2002

The Industrial Triangle (Pittsburgh–Milwaukee–St Louis)
Chicago	+5
Detroit	−1
Cincinnati	−6
Cleveland	−7
Milwaukee	−3
Pittsburgh	+9
St Louis	+5
The Pacific Coast
Los Angeles	−6
San Diego	+9
San Francisco	−17
Seattle	−6
The North East
Boston	−7
New York	−2
Philadelphia	−3
The South
Atlanta	+2
Miami	0
Tampa	−4
The Centre
Dallas–Ft Worth	+3
Denver	+1
Houston	+10
Kansas City	+13
Minneapolis–St Paul	+3
Phoenix	+2

Figure 1 shows the average gain or loss in the competitiveness ranking for the MSAs in each region. The results differ somewhat from those of the earlier 1999 paper. In that paper, during 1977–87 to 1987–92, MSAs in the industrial triangle gained 8 positions, on average, where in this study, for 1987–92 to 1997–2002, they are essentially unchanged; and the centre rose by 4 positions, while here the gain is 5.5 positions. It was also the case that MSAs in the Pacific Coast, North East and South lost respectively, 6.5, 11 and 4.5 positions in the earlier study; whereas, in the present study, MSAs in these three regions are still losers—but by 6, 5 and 1.5 positions respectively. MSAs on the two coasts and in the South continue to be lacking in competitiveness, the Centre still surges and the Industrial Triangle is holding its own. The regional advantages and disadvantages have become less extreme for the 1987–92 to 1997–2002 period than they were for 1977–82 to 1987–92. Of course, the earlier period was marked by a major shock (the petroleum price increases) with important differential regional impacts and the recovery from it, whereas the later period was relatively tranquil.

Figure 1.

Regional winners and losers: changes in competitiveness ranking, between 1987–92 and 1997–2002.

Determinants of Urban Competitiveness

Rankings do give some cities bragging rights, but are not all that interesting analytically. However, once we have this ranking we can then move to the more important part of the analysis, that of ascertaining the specific determinants of urban competitiveness; that is, the answer to the question: ‘Why is city x more competitive than city y’? A regression analysis was conducted, with the results given in Table 3. These variables have been demonstrated statistically to be determinants of urban competitiveness. The t-ratios are given for each variable and exceed 2.0 in most cases; as shown by the p-values, six variables are significant at the 0.05 level of confidence, one at the 0.10 level. (For analysis of variance see the Appendix.) Finally, the coefficient of determination is 0.833, very high for cross-section analysis.

Table 3.

The determinants of urban competitiveness: urban competitiveness ranking = −3.199 + 1.139x¹ + 0.000085x² + 0.0040x³ + 0.028x⁴ + 0.002x⁵ + 0.040x⁶ − 0.002725x⁷ + 0.003x⁸

Predictor	T-ratio	P-value
Constant	−1.61	0.130
MVA02/97	3.63	0.003
Hospitals 98	3.32	0.005
% BA BS 25+	3.21	0.006
Labour force/FIRE	2.84	0.013
Culture	2.66	0.019
100-F100	1.73	0.107
Labour Force/RC	0.913	0.377
Transport	2.49	0.026
R ²	0.833
Adjusted R²	0.738

Notes: x¹ = growth in manufacturing value added, 1997–2002; x² = hospital beds in 1998; x³ = percentage of the 25 and older population with a BA or BS degree; x⁴ = labour force/finance, insurance and real estate employment; x⁵ = the number of cultural institutions; x⁶ = 100 minus the percentage of firms with 100 or more employees; x⁷ = transport infrastructure; and x⁸ = labour force/university and government research centres.

Sources: Gaquin and DeBrandt (2007); Savageau (2007).

The signs for all of the determinant variables are positive. “Labour force/finance, insurance and real estate” and “labour force/research centres” are seemingly perverse as the results indicate that neither the FIRE component of the labour force nor research centres in relation to the labour force are positive for competitiveness. This does not come as a surprise because, in the first study, in 1994, a variable that was similar to Labour force/FIRE, Engineering, administrative, research and management (EARM), had the same impact on competitiveness. That study was for the period 1977–87. It was concluded at the time that this was emblematic of the widespread understanding that the US economy was, if anything, overmanaged. Periodic reports of cutbacks in administrative staff by large firms suggest that a similar situation is found in subsequent periods. The other negatively related determinant, labour force/RC, reflects that, while research may be done in one urban economy, the production of goods and services, as was noted with regard to Boston, takes place in another. The current results also indicate that the most competitive cities are not those in which the economy is dominated by large firms employing hundreds or thousands of workers, but rather by an industrial structure that is dominated by smaller firms (the percentage of firms with fewer than 100 employees), the much-lauded ‘start-up’ and ‘spin-off’ firms that are typically focused on some niche activity both in traditional sectors, such as steel production, or in the newer high-tech sectors (Corona et al., 2006, ch. 2). Large firms in the US have been reducing their workforce for many years in the effort to cut costs and to meet the challenges from goods produced elsewhere. It is also interesting to note that, in our first (1994) study, being located in the South was a determinant, whereas in the current study this was not a factor of significance—the Pearson coefficient of correlation between it and the competitiveness ranking was only 0.491. Finally, in the second (1999) study conducted, using data for the period 1977–92, the percentage of the 25 or older workforce who had a university degree had a negative sign, but in our study this indicator of the education of the labour force has become a significant and positive factor. This is reflective of the transition of the US economy from basic manufacturing to niche manufacturing and high-level services.

The transport infrastructure has become important for urban competitiveness, whereas it was important only for the relatively skilled EARM (engineering, accounting, research and management) component of the labour force in the 1999 study. The city’s endowment in cultural institutions has been a determinant of urban competitiveness in each of the three studies, partly because it attracts visitors to the city and partly because it is important in attracting and retaining educated/skilled workers (Americans for the Arts, n.d.; Markusen and King, 2003). Even if these workers are too occupied to participate in cultural activities, they demand them for their children. We also take the richness in a city’s endowment of cultural institutions as a proxy for urban amenities, which would be too complex to capture more directly. Health care, in the form of hospital beds per 100 000 residents, has emerged as a significant determinant for the first time.

Finally, the growth in manufacturing value added is shown to be a determinant of urban competitiveness. This variable indicates that the manufacturing sector is expanding or that it is moving from low to high value added activities, presumably related to increasingly technology-intensive production.

We have been able to do a regression analysis of the determinants of one of the indicators of competitiveness so as to obtain additional second-tier determinants of urban competitiveness—the percentage of the population, 25 years of age and older, who have attained a university education. This regression exercise gives city leaders additional information as to which policy initiatives are likely to have a positive impact on the city’s competitiveness. The results of this analysis are presented in Table 4. These results indicate to us that to achieve a high percentage of residents with a university education, the city must ensure that these individuals will be assured of personal safety through a low level of crime activity. City leaders must also work to ensure adequate opportunities for leisure activities, including recreational structures and cultural events. Finally, the transport system must satisfy the needs of the educated workforce. The less civilian employment (1/civilian employment) determinant represents a scale indicator and it tells us that large urban economies with their large civilian employment do not have a competitive advantage over smaller ones when it comes to attracting an educated workforce—quite the reverse is true.

Table 4.

The determinants of %BA BS 25+: %BA BS25+ = −44.613 + 0.027x⁹ + 0.651x¹⁰ + 0.169x¹¹ + 1.408x¹²

Predictor	T-ratio	P-value
Constant	−2.829	0.001
Crime ranking	2.802	0.012
Leisure ranking	3.971	0.001
Transport ranking	4.034	0.001
Less civilian employment	2.575	0.019
R ²	0.605
Adjusted R²	0.517

Notes: x⁹ = ranking in crime; x¹⁰ = ranking in leisure; x¹¹ = ranking in transport; and x¹² = less civilian employment.

How Urban Leaders Can Use This Analysis

For these results to be of use to decision-makers and planners in the individual MSAs, all of the determinants must be presented in a form that highlights the specific competitive strengths and weaknesses of that MSA. We do this in Table 5. Here, we present two sets of determinants: the primary determinants that explain the urban competitiveness ranking; and the secondary determinants that explain the educational attainment of the population of that MSA. Two explanatory comments are required. The value for Labour force/FIRE indicates that for most MSAs a higher share of the labour force being in finance, insurance and real estate does not enhance competitiveness. The products of this sector may not in most cases be extra-regional traded services and may do little to increase economic growth. The positive impact of Labour force/research centres suggests that research centres develop new products and new technologies but the use of these competitiveness-enhancing results of research may be used in production activities elsewhere. Hence, when San Diego has the highest value in Labour force/FIRE, this means that its economic structure is not heavily weighted in the FIRE service sector activity, whereas Philadelphia, with its low value, is. Similarly, while Detroit has the highest value in Labour force/research centres this means that Detroit is not a research town, while Pittsburgh, with its low value, is. Secondly, there are two transport determinants, one is the per centile position of the MSA among the 354 MSAs and the other is the MSA’s ranking in that same grouping. While the correlation coefficient between the two is 0.991, each does slightly better in regression analysis either as a primary or secondary determinant respectively.

Table 5.

Metropolitan area rankings by explanatory variable, 1997–2002

	Primary determinants									Secondary determinants
	ΔMVA	HOSP	Educ	LF/FIRE	Cult	F<100	Trans	LF/RC	Crime	Leisure	Tran R	CivEmp	Rank
Miami	2	17	19	5	21	1	22	13	23	10	22	7	1
San Diego	1	23	8	1	20	6	18	6	6	7	18	16	2
Phoenix	20	22	16	12	23	9	21	5	19	10	21	14	3
Kansas City	10	5	11	16	9	19	19	4	1	22	19	22	4
Atlanta	6	2	6	10	17	19	9	15	11	14	9	8	5
New York	21	1	7	21	1	2	3	20	5	1	3	1	6
Dallas–Ft Worth	7	11	10	11	14	16	7	2	21	17	7	5	7
Houston	19	12	14	2	16	13	20	14	2	17	20	9	8
Seattle	5	21	5	4	15	5	5	22	15	4	5	13	9
Minneapolis–St Paul	11	8	4	18	6	17	13	16	7	12	13	12	10
Denver	23	18	3	13	12	3	6	8	12	14	6	17	11
Tampa	4	16	23	15	22	7	23	3	22	16	23	19	12
Boston	17	9	2	22	3	11	8	17	3	6	8	6	13
Pittsburgh	9	7	21	8	7	14	10	23	4	20	10	18	14
Los Angeles	8	20	15	6	4	8	4	7	13	2	4	2	15
Chicago	13	6	9	14	5	18	2	10	18	2	2	3	16
St Louis	14	3	17	9	19	19	12	9	20	19	12	15	17
Philadelphia	3	4	12	23	10	10	11	18	10	8	11	4	18
San Francisco	22	13	1	19	2	3	1	21	8	4	1	10	19
Cincinnati	15	10	18	7	11	23	14	11	17	23	14	21	20
Detroit	16	14	22	3	18	12	15	1	16	8	15	11	21
Milwaukee	12	19	13	20	8	22	16	19	14	21	16	23	22
Cleveland	18	15	20	17	13	15	17	12	9	13	17	20	23

Notes: Primary determinants: ΔMVA = increase in MVA, 1997–2002; HOSP = hospital beds/100 000; F<100 = percentage firms with fewer that 100 employees; Educ = percentage of 25+ population with university degree; Cult = ranking of 354 MSAs; LF/FIRE = finance, insurance, real estate empoyment; LF/RC = labour force/recearch centres; Trans = ranking of 354 MSAs.

Secondary determinants: Crime = ranking of 354 MSAs; Leisure = percentile of 354 MSAs; Trans R = percentile of 354 MSAs; CivEmp = less civilian employment, 2000.

With these caveats, what understanding can an MSA leader gain from Table 5? It has been argued that, for effective strategic planning, decision-makers must understand how their MSA stands in relation to others that might stand in competition with it (Kresl, 2007, ch. 2). In isolation, something the MSA has put in place may make leaders feel they have gained some competitive advantage when, in reality, what they have done just keeps the MSA in the same competitive position since other MSAs have undertaken the same initiative, such as a convention centre. Clearly, a full understanding of this dynamic can be gained only from intensive study of the specific situation, but the general understanding that can be gained from Table 5 can also be of use. For example, top-ranked Miami has strengths in the growth in MVA, in the fact that it is not overdominated by the FIRE sector, and in the large percentage of firms that have under 100 employees; however, it has clear weaknesses in its cultural and transport infrastructures, in the education of its labour force and its high crime rate. Clearly, there are specific initiatives MSA leaders in Miami could undertake to enhance their MSA’s competitive strengths and to diminish its competitive weaknesses. This sort of analysis can be done for each of the MSAs in this study. While MSAs ranked at the bottom, such as Detroit, Milwaukee and Cleveland, can improve their situation by taking action on almost any or all of the determinants, closer on-site analysis would allow one to design a strategy that could be relatively successful by focusing on a small number of these determinants where improvement would generate the maximum enhancement of competitiveness. It is the MSAs in the middle, from Dallas–Ft Worth to Chicago, for which the relative strengths and weaknesses could be used most effectively to fashion a strategic approach for competitiveness enhancement. Some of the weaknesses will be relatively easy to fix, whereas others will be more intransigent—calling for a triage sort of approach to action. Each of the MSA’s strengths will be challenged by another MSA, so having strength in a particular determinant should not be an excuse for self-congratulation and passivity in this area.

Fundamentally, the response of city leaders to the information in Table 5 should not be that of focusing on the ranking, trying to move up a step or two, but rather to use the rankings for each determinant to make tangible, objective improvements in specific areas of relative strength and weakness. The position of the MSA in the rankings table will take care of itself.

How the Determinants of Urban Competitiveness Have Changed over Time

Finally, since these three studies of urban competitiveness have been done over three decades, we can note the changes there have been in the explanatory determinants. The determinants are presented in appropriate groupings in Table 6. Four appeared in all three studies and eight were found in only one period. Some of the determinants—specifically, fiscal, regulatory and political climate, state capital stock and EARM—were available for only one or two of the periods and one, growth in MVA, appeared only in the third period because of a change in methodology. The most significant overlap is in the first two periods with three times as many shared determinants as in the last two periods, suggesting that some important transformation occurred in the economic environment towards the end of the 20th century. Some of the determinants took a different form from period to period—that is, the value might be for one year or for growth over several years, and a determinant might be either a ranking among all US MSAs or a percentile of the highest value for that determinant among all US MSAs. The consequences of this transformation are revealed by an examination of the changes that have taken place in the nature of the determinants of urban competitiveness.

Table 6.

Determinants over the three periods

Determinants in all three periods

Educational attainment of the population

Cultural institutions

Firm size

Research centres/labour force

Determinants in the first two periods

Growth in per capita income

EARM in the labour force

Growth in population

Location in the Sun Belt

Research centres/MVA

Growth in the capital stock of the state

Determinants in the last two periods

Transport services

Health care facilities

One time only determinants

Period 1

Managers in the labour force

Period 2

Fiscal, regulatory, political climate

Growth in the labour force

Period 3

Growth in manufacturing value added

FIRE/labour force

Security

Leisure opportunities

Civilian employment

First, location in the Sun Belt, the band from Virginia through to southern California, ceased to be of importance in the third period. This could be reflective of a fundamental change that occurred as globalisation dramatically altered the competitive situation of urban economies in the US and elsewhere. The Centre became the principal region of strength in the US and transport emerged as a determinant of importance. Secondly, it is also noteworthy that ‘softer’ determinants, such as health care, security and leisure, replaced growth in population and per capita income, two determinants that were important in the earlier years and, as noted earlier, the sign for education of the labour force changed from negative to positive for the most recent period. This is most likely to be a reflection of the higher skills and educational attainment of today’s workers and of the transition from basic manufacturing to higher-technology niche manufacturing and to advanced services, including health care and education. Thirdly, we can note the observations made with regard to some of the other determinants in the discussions of Tables 3 and 4. Specifically, while avoiding excessive repetition: the aspects of a city that are attractive to an educated and skilled workforce (already enumerated); the continuing importance of urban amenities (culture goods and services); and, the importance of smaller ‘start-up’ or ‘spin-off’ firms.

Final Comments

We opened this paper by noting that there were different methodological approaches to the study of urban competitiveness and that each had its own advantages. Without commenting on the advantages or disadvantages of the other two approaches, we would like to finish by highlighting what can be accomplished using our methodology. First, our ranking of cities is done by utilising our Kresl–Singh methodology in which three variables that we assert are reliable indicators of urban competitiveness are used. This assertion has been supported by discriminant analysis, thus removing subjective analysis from our approach, in comparison with the two other approaches, as much as is possible. Secondly, when we have the ranking, we can then derive a set of determinants of urban competitiveness that are statistically verifiable. This gives us a smaller set of determinant variables than the other approaches but we have more confidence in the validity of these variables. Thirdly, when we repeat this study for different time-periods, we can reveal the increase and decrease in competitiveness of individual cities and of geographical regions over time. Fourthly, we can show how the importance of individual determinants has varied over these time-periods. Fifthly, using the Kresl–Singh methodology, we are able to present our determinants of urban competitiveness in a table that makes them useable by city leaders in designing an economic strategic plan for the development of their city’s economy in the near future. Unfortunately, it is not possible for us to ascertain the extent to which city leaders have utilised these findings in their policy formulation. We conclude that these five advantages do not obtain with the two other methodologies.

Footnotes

Appendix

Table A2.

The results of earlier studies

Results of the 1994 OECD study

Determinants of urban competitiveness

The increase in per capital income

The percentage of the population 25 years and older with a university undergraduate degree

The number of research centres/labour force

The share of the labour force categorised as ‘managers and professional’

A dummy variable for location in the Sun Belt and West

The share of the labour force in EARM (engineering, accounting, research and management)

The ranking of the city according to its cultural institutions

Determinants of growth in per capita income

The increase in the population

The increase in the percentage of the population with a university degree

The increase in the percentage of firms with 100 or more employees

The growth in investment in plant and equipment

Determinants in the EARM component of the labour force

The increase in the population

The increase in the percentage of the population with a university degree

Research centres/manufacturing valued added

The number of cultural institutions

Results of the 1999 Urban Studies Study

Determinants of urban competitiveness

The growth in per capital money income

Research centres/manufacturing valued added

The growth in the percentage of firms with more than 100 employees

The number in the labour force with more than a BA/BS degree

The share of EARM (engineering, accounting, research and management) component of the total labour force

The growth in the number of cultural institutions

The growth in the capital stock for the state—exports as a share of total output

Determinants of the share of the EARM component of the total labour force

The growth in the population

Transport services

Research centres/labour force

Location in the Sun Belt

Determinants in the growth of per capital money income

The fiscal, regulatory and political climate

The percentage of firms with more than 20 employees

The growth in the labour force

The number of cultural institutions

Results of the Present Study

Determinants of urban competitiveness

Growth in manufacturing value added, 1997–2002

Hospital beds per 100 000 in 1998

Percentage of the 25 and older population with a BA or BS degree

Finance, insurance and real estate employment as a share of the labour force

The number of cultural institutions

The percentage of firms with fewer than 100 employees

University and government research centres/labour force

Transport infrastructure and services

Determinants of the percentage of the 25 and older population with a BA or BS degree

Ranking in public security and crime

Ranking in leisure

Ranking in transport infrastructure and services

Civilian employment, 2000

Notes

References

Americans for the Arts (n.d.) Arts economic prosperity III: the economic impact of nonprofit arts and culture organizations and their audiences. Americans for the Arts Washington, DC.

Bloomberg (2009) Pittsburgh’s economic revival ( http://www.bloomberg.com/apps/news?pid=newsarchive&sid=aUY90NWSaEXQ ).

Cohon

(2009) Commentary: how Pittsburgh bounced back, CNN, 24 September (www.cnn.com).

Competitiveness Institute (n.d.) Website (www.competitiveness.com).

Corona

Doutriaux

Mian

S. A.

(2006) Building Knowledge Regions in North America: Emerging Technology Innovation Poles. Cheltenham: Edward Elgar.

Cumbers

MacKinnon

(2004) Introduction: clusters in urban and regional development, Urban Studies, 41(5/6), pp. 959–969.

Eurocities (2002) Eurocities mayors’ declaration. Brussels, 29 November (http://www.eurocities.eu/).

Florida

(2002) The Rise of the Creative Class. New York: Basic Books.

Florida

Tinagli

(2004) Europe in the Creative Age. London: Demos.

10.

Gaquin

D. A.

DeBrandt

K. A.

(Eds) (2007) County and City Extra:Annual Metro, City and County Data Book, 15th edn. Lanham, MD: Bernan Press.

11.

Gertler

M. S.

Florida

Gates

Vinodrai

(2002) Competing on creativity: placing Ontario’s cities in North American context. Ontario Ministry of Enterprise, Opportunity and Innovation, and the Institute for Competitiveness and Prosperity, Toronto, November.

12.

Global Urban Competitiveness Project (n.d.) Website (www.gucp.org/en/).

13.

Johansson

Paulsson

(2009) Location of new industries: the ICT sector 1990–2000, in: C. Karlsson, B. Johansson and R. R. Stough (Eds) Innovation, Agglomeration and Regional Competition, pp. 261–293. Cheltenham: Edward Elgar.

14.

Johnson

D. K. N.

Brown

(2004) How the West has won: regional and industrial inversion in US patent activity, Economic Geography, 80(3), pp. 241–260.

15.

Kresl

P. K.

(2007) Planning Cities for the Future. Cheltenham: Edward Elgar.

16.

Kresl

P. K.

Singh

(1994) The competitiveness of cities: the United States, in: OECD (Ed.) Cities and the New Global Economy, pp. 424–446. Melbourne: The Government of Australia and the OECD.

17.

Kresl

P. K.

Singh

(1999) Competitiveness and the urban economy: twenty-four large US metropolitan areas, Urban Studies, 36(5/6), pp. 1017–1027.

18.

Krugman

(1993) Geography and Trade. Cambridge, MA: MIT Press.

19.

Lever

W. F.

(2002) The knowledge base and the competitive city, in: Begg

(Ed.) Urban Competitiveness: Policies for Dynamic Cities, pp. 11–31. Bristol: The Policy Press.

20.

Markusen

(1996) Sticky places in slippery space: a typology of industrial districts, Economic Geography, 72(3), pp. 294–310.

21.

Markusen

King

(2003) The artistic dividend: the arts’ hidden contributions to regional development. Working Paper No. 315, Project on Regional and Industrial Economics, Humphrey Institute, University of Minnesota, MN.

22.

Markusen

Schrock

(2006) The distinctive city: divergent patterns in growth, hierarchy and specialisation, Urban Studies, 43(8), pp. 1301–1323.

23.

Maskell

Törnquist

(2001) Building a Cross-border Learning Region. Copenhagen: Copenhagen Business School Press.

24.

Matthiessen

C. W.

(1990) Vesteuropa 1990, Geografisk Tidsskrift, 90, pp. 45–46.

25.

Negrey

Zickel

M. B.

(1994) Industrial shifts and uneven development, Urban Affairs Quarterly, 30(1), pp. 27–47.

26.

Kresl

P. K.

(2010) Global Urban Competitiveness Report 2010. Cheltenham: Edward Elgar.

27.

OECD (1994) Cities and the New Global Economy. International conference presented by the OECD and the Australian Government, Melbourne, November.

28.

Pollard

Storper

(1996) A tale of twelve cities: metropolitan employment change in dynamic industries in the 1980s, Economic Geography, 72(1), pp. 1–22.

29.

Porter

(1990) The Competitive Advantage of Nations. New York: The Free Press.

30.

Rantisi

N. M.

(2004) The ascendance of New York fashion, International Journal of Urban and Regional Research, 28(1), pp. 86–106.

31.

Savageau

(2007) Places Rated Almanac, 25th edn. Washington, DC: Places Rated Books.

32.

Urban Studies (1999) Review Issue: Competitive Cities, 36(5/6).

33.

Urban Studies (2002) Review Issue: The Knowledge-based City, 39(5/6).

34.

Winden

W. van

Berg

L. van den

Pol

(2007) European cities in the knowledge economy: towards a typology, Urban Studies, 44(3), pp. 525–549.