The Urban Audit as an Instrument of Comparative City Analysis

Contents:

1.  Introduction

The last two seminars of our series launched an interesting exchange of experience between scholars and practitioners from the three participating capital cities, Berlin, Helsinki, and Stockholm on a variety of subjects in urban development. What has hitherto been lacking, however, has been a directly comparative approach.

The last two seminars of our series launched an interesting exchange of experience between scholars and practitioners from the three participating capital cities, Berlin, Helsinki, and Stockholm on a variety of subjects in urban development. What has hitherto been lacking, however, has been a directly comparative approach.

This paper seeks to examine whether and to what extent the EU Urban Audit project, the pilot phase of which has just been completed, and in which all three cities have participated, can be helpful in this regard.

2.  The Urban Audit Project

In 1997 the Directorate General XVI of the European Commission invited tenders for the project "Urban Audit", the contract being awarded to the British consulting consortium ERECO/ECOTEC ("European Economic Research and Advisory Consortium"). The project organisation provided for a British core team supported by a network of "correspondents" from member states of the EU. The German Institute of Urban Affairs was the national correspondent institute for Germany, organising cooperation with the nine German cities chosen for the project by the EU, namely Berlin, Dresden, Essen, Frankfurt am Main, Hamburg, Cologne, Leipzig, Munich, and Stuttgart.

With the Urban Audit project, the Directorate General XVI wanted to make a contribution to the comparative statistical assessment of the quality of life in European cities.The project was motivated by a growing interest within the EU in urban statistical data satisfying the requirements of international comparability. This interest was fired by the desire of the Commission to devote more attention to cities in traditional regional and structural policy, but also to make it easier than hitherto to assess the impact of various policy areas on cities ("urban compatibility audit").

Moreover, the Urban Audit, conceived as a pilot project, was intended by the EU to help cities establish a better basis for transnational comparative city studies. The results of a twelve-month pilot phase involving an initial 58 cities paves the way for an invitation to all European cities to join in this undertaking. In a few days, on 18th September 2000, a public presentation of the project is to take place with political participation in Paris-Créteil, at which the future course is to be discussed.

Overview 1: Cities Covered by the Urban Audit Project

The statistical offices of the German urban audit cities had formed a working group - cooperating with the German Institute of Urban Affairs - of the KOSIS Association in the Association of German Urban Statisticians. The participating cities also welcomed the Urban Audit as an opportunity of testing modes of cooperation with the EU, and in this spirit the project has also found the support of the statistics committee of the German Association of Towns and Cities. Furthermore, the governments of the German Urban Audit cities were also affirmative about cooperation in the projects when approached on the subject by the the Brussels office of the "Eurocities" network.

The terms of reference laid down by the EU addressed the 58 pilot-phase cities (Overview 1) and the 21 domains for which living conditions were to be presented in comparable form with the aid of (ultimately 108) indicators (Overview 2).

Overview 2: Domains of the Urban Audit Project

Two additional requirements were intended to serve the spatial differentiation of the results. All cities were to attempt to supply part of the statistical information additionally at the level of two strongly contrasting sub-city areas. This was not only to show disparities between European cities, but also to permit insight into internal disparities within cities. However, the participating cities hastened to point out that any comparison between European cities should not lead to additional discrimination of deprived sub-city areas. To obviate this, data on sub-city areas were encoded, since no geographical attribution containing the names of sub-city areas was necessary to show the extent of disparities within a city. We feel this is a good example of the political dimensions of the project and of statistical comparisons in general. Cities are concerned not to have their image harmed in international competition and do not want to run any risk in their direct relations with the EU of suffering any (financial) disadvantage. The EU has accordingly guaranteed that no data will be published without the individual approval of the participating cities.

Furthermore, some Audit cities were requested by the EU to supply data on the wider urban unit or conurbation as a whole to show relations with the surrounding areas. Among the cities requested to do so were Helsinki and Stockholm, but not Berlin. The German cities given this task were Frankfurt and Stuttgart.

In the Urban Audit project it was not easy to find a methodological approach for defining the wider territorial unit that was acceptable locally while being comparable between European cities.

Despite this and other difficulties, however, both practitioners and investigators are increasingly interested in comparative studies of European cities, since they provide answers to three types of question:

The very tight time-frame of the Urban Audit project laid down three phases limited to four months each:

Phase one: Investigation of statistical sources available at the local and national levels. No-one with experience in the matter will be surprised to learn that an abundance of definitional problems and difficulties with the varying availability of data are likely in such an international comparative study. The projected set of comparative tables needs to be complemented by comprehensive information on the metadata in which due attention is called to these problems.

Phase two: Collection of the statistical data themselves, where available in accordance with EU requirements. The German cities have agreed on a procedure essentially comparable to the production of the Statistical Yearbook of German Municipalities. This is to be accompanied by a "national report" to be prepared by the German Institute of Urban Affairs in coordination with the individual city reports.

Phase three: Presentation of the comparative results and assessments. This included the preparation of a manual to permit interested cities that had not been among the selected Audit cities, to join the project ("Urban Audit Manual"). It was also decided to publish the results in the Internet. What we have to say in the following section is based on this Internet presentation (http://www.inforegio.cec.eu.int/urban/audit). Finally, the EU has announced the publication of a book presenting the results of the project. On this subject two publications on the Urban Audit are meanwhile available for downloading from the above site.(1)

3.  Selected Comparisons Berlin - Helsinki - Stockholm

The amount of data originally planned for the project appears impressive at first sight. At the city administration level alone, 108 indicators are envisaged for the 21 "domains", with greater depth being provided by further information on internal differentiation by sub-city areas of about 100,000 inhabitants each. Moreover, Helsinki and Stockholm (but not Berlin) are among the Audit cities required to provide statistical information on the Wider Territorial Unit (WTU) or conurbation. Differentiation of information in time series covering 1981, 1991, and 1996 serve to describe longitudinal developments.

Overall, the somewhat tight time-frame of the project - about a year - set limits to the collection of the required data to the extent and in the quality planned. For the German cities it was also a problem that the country has failed to maintain the European census rhythm (there was no 1981 census), and for Berlin there is an additional data bottleneck caused by the unification of Germany in 1990, so that numerous indicator values are lacking especially for 1981. In the following tables, supplementary figures not recorded by the Urban Audit are given in brackets.

3.1  Population

Figure 1

Source: Landesamt für Datenverarbeitung und Statistik Brandenburg und Senatsverwaltung für Stadtentwicklung, Umwelt und Technologie, Bevölkerungsprognose für Berlin bis zum Jahre 2010, Berlin 1997, page 18.

Berlin is the largest Urban Audit city, with Helsinki and Stockholm counting as "medium size" cities (between 500,000 and 1 million inhabitants). The average Urban Audit city had a population of 741,000. Incidentally, London and Paris were not included in the project because they were considered "too large". In terms of geographical location, all three cities belong to the "North" group, the other geographical groups being "South", "Centre", and "Periphery".

Changes in population figures between 1991 and 1996 reveal clear differences between Helsinki and Stockholm on the one hand and Berlin on the other. With an increase of only 0.4%, Berlin is close to the Urban Audit average of 0.7%. Helsinki (8.1%) and Stockholm (5.8%) are closer to the maximum growth figure, which is 8.4% (Nantes).

A further difference is apparent in the share of the capital city population in the total population of the country in question. Helsinki has the highest share with 10%, while 8% of the Swedish population live in Stockholm and only 4% of Germany’s population in Berlin. The share of the respective WTU population in the national population exhibits the same sequence. This points to the nature of the urban system in Germany, where there are many large cities apart from the capital. As a result Berlin does not loom as large as the two other capitals in the national context.

3.2  Population structure

Figure 2: Structure of Population by Age and Nationality

57 cases.

56 cases.

The young age groups "0-15" and "16-24" are underrepresented in all three cities measured against the average for Urban Audit cities, respectively 17.1% and 13.6%. In contrast, all three are above the 53.2% average for the "25-64" age group. Berlin and Helsinki have the same proportion of people in the "65 and above" age group, below the overall average of 16,4%, while Stockholm is a little above this average with 18,3% of the population 65 and above.

These components can be summed to the group “Less than 16 years old and above 65”. With shares of about 30% the figures are identical for Berlin and Helsinki, with a higher proportion in Stockholm, 35%, which is also slightly above the overall average for Urban Audit cities of 33.3%. This means that Stockholm in the 1st quintile of Urban Audit cities has a heavier “demographic burden” of young and old for the economically active population to shoulder than Berlin and Helsinki.(2)

The three cities differ strongly in the proportion of non-national residents. The average proportion of nationals in all Urban Audit cities is 91.2%. Stockholm and Berlin have fewer, but Helsinki, with 99.2%, has markedly more, exceeded only by Seville with 99.5%. On the other hand, Stockholm and Berlin differ as to the origin of non-national residents, since the 11% “non-EU nationals” living in Berlin is comparatively high.

3.3  Households

Figure 3

Stockholm 1990.

58 cases.

57 cases.

52 cases.

The average size of households differs little between the three cities. In the average distribution of all Urban Audit cities, they are also in the same extreme quintile of smallest households, whereas the average figure for all 58 cities is still 2.4.

The share of one-person households in all three cities is correspondingly high, Stockholm showing a particularly high share of 55%. Only the Copenhagen figure of 60% is higher. The 35% overall average for Urban Audit cities is much lower. High proportions of one-person households are typical for large cities, and this is shown by the comparatively low figures for the Helsinki and Stockholm wider territorial units.

Again, the proportion of lone-parent households differs between the three cities. Berlin, with 7.6% is close to the 7.9% average for all cities, while Helsinki with 8.5% is a little above this figure and Stockholm with 4.7% markedly below.

There is a similar distribution pattern for lone-pensioner households. Berlin (13.5%) and Helsinki (12.8%) are about average in comparison with all Urban Audit cities (14.0%), while Stockholm has a far higher proportion of such households (19,6%).

3.4  Labour Market

Figure 4

Activity rate (%): Residents, who are economically active x 100/Residents of working age.

58 cases.

54 cases.

Source: Statistisches Jahrbuch Deutscher Gemeinden, 84. Jg. (1997), S. 470

The proportion of economically active people is high in all three cities. The Urban Audit average is 65.0%. The figures for Berlin (75%) and Helsinki (69%) are higher, and the figure for Stockholm (81%) is much higher, second only to Leeds (84.2%) among Urban Audit cities.

The situation is similar for female employment. Here, too, the three cities are above the Urban Audit average of 56%, rising from Helsinki (66%) to Berlin (70%) and finally Stockholm (80%). In this field, too, Leeds, with a top score of 82,7% is the only Urban Audit city with a higher figure.

In unemployment, Helsinki is close to the 8.4% Urban Audit average with a figure of 9%, while Stockholm, with only 5.5%, has far fewer jobless. Berlin has an exceptionally high rate of 15%. And it is not the case that unemployment is concentrated in former East Berlin. The figure for West Berlin meanwhile exceeds that for East Berlin.

3.5  Income and Poverty

Figure 5

Helsinki 1995.

32 cases.

19 cases.

Recipients of social security, but only "Laufende Hilfe zum Lebensunterhalt".

24 cases.

Median weekly household income assigns the three cities to three different quintiles in statistical distribution. Berlin has the lowest figure (350), a little above the statistical mean of all Urban Audit cities (347), which is, however, formed by only 32 cities. Stockholm with 357 shows a somewhat higher figure, and Helsinki has by far the highest level for the three cities with 461. This would mean that, despite a higher proportion of economically active population in Berlin and Stockholm, these two cities reach only about 75% of the Helsinki household income. However, the Urban Audit editors do point to the low international comparability of the data.

The poverty indicator "proportion of households receiving less than half of the national average household income" is informed only by Berlin and Helsinki. Both cities are slightly below the 23.2% average for all Urban Audit cities, the figure for Berlin being 23% and for Helsinki 18%. The editors remark at this point that the international comparability of the data is adequate, although it must be said that the median figure is derived from only 19 Urban Audit cities.

The indicator "proportion of households reliant on social security" can be interpreted as measuring policy reaction to the extent of poverty rather than poverty itself. The data from the three cities is surprising at first glance considering the preceding indicator: Berlin has a higher proportion of low income households than Helsinki, but Helsinki has a much higher score than Berlin when it comes to the proportion of households dependent on social security. The reason is that the definitions underlying the figures differ. The Urban Audit editors remark: "Germany: Data refer to recipients of social security, but only ‘laufende Hilfe zum Lebensunterhalt’ (the continuous transfers not related to contributions to the social security system)". For Finland, in contrast, they note: "The data presented concern the number of households that are currently in receipt of a ‘living allowance’, to overcome periods of financial difficulty, i.e. when income is insufficient for a ‘modest’ standard of living. The data do not include refugees, re-migrants or asylum seekers. These groups receive social security from the Ministry of Labour and not the municipality".

This is a good example of the “traps” in transnational comparisons of cities, even when limited to western Europe.

Otherwise, the Urban Audit provides data for the three cities on "health", "crime", "elections", "pre- and post school provision", "consumption of water", "climate", land use", and, in the culture and recreation domain, "concerts", "cinemas", and "museums".

4.  Conclusion

It is no easy task to come to an equitable conclusion about the Urban Audit Project, especially since we are dealing with a pilot phase. Nevertheless some remarks should be made.

1. The project was very ambitious considering the rather brief period allowed for implementation. There was clearly little time available for the central task of laying down comparable definitions and accessing comparable data sources. This was indicated not only by the German cities, which faced the additional problems of reconciling the peculiarities of local authority statistics in the structure of official statistics in Germany with the requirements of the project.

2. Surprisingly frequently, data is stated to be "not available". The organisers of the project comment: "Where information is not available, the reasons for this are classified as follows:

3. • N.Av1: data does not exist (sometimes information exists for part but not all of the indicators).
   • N.Av2: data does exist, but there is no approval to publish.
   • N.Av3: data exists but requires processing and / or purchasing prior to possible inclusion in the Urban Audit.
   • N.Av4: data was not provided in time for inclusion.
   • N.Av5: data missing for another reason.
   • N.Av6: the cell is not appropriate."

4. Compared with a "simple" statistical yearbook, it is an interesting advance to present the data regularly with qualifying remarks and calculations. They include the computation of quintiles and averages, as well as information on maxima and minima as aids to interpretation, as well as information on the international comparability of the data.

5. Although the data helps meet an initial need for information, the broad substantive spectrum of the "domains/indicators" means almost necessarily that these data are too superficial for specific detailed inquiries. As the sole source of data they are insufficient for such purposes, and comparative studies will regularly require complementary data sources.

6. On the other hand, it is a commendable basic intention to foster city comparisons at the European level and to make the information available to the (professional) public in modern form in the Internet. The current relevance of information usually associated with the Internet is limited, since the latest data are from 1996. At the launching of the project in mid-1998, they were the latest generally available.

7. So far, the Urban Audit project has given no great impetus to German local authority statistics. One reason is that many cities have been occupied for some time now in reorganising their local authority statistical information systems, and give higher priority to this task than to meeting the demands of the Urban Audit project for information. The situation may be different for cities that have hitherto not had such information systems.

8. This does not diminish the political dimension of the project. It also shows that the integration process in Europe has now reached the local authority level, and that comparative statistical information on European cities is also needed for rationally grounded local government policy. It will be interesting to see what decisions are made at the EU level on the question of continuing the project after the Paris conference. The interests of German local authority statistics will coincide with EU goals if local government statisticians serve not only as data providers but, keeping in mind the goals of municipal administration, can make a contribution in a spirit of partnership to the upgrading of European urban statistics.

Overall, assessment of what has been achieved to date in the project can be expressed with the well-known image: a glass of water can justifiably be described as either half full or half empty.

(1) Volume I: "Overview and comparative section" and Volume II: "Summary results for each city".

(2) In summary overviews, selected indicator values are compared with all other Urban Audit cities to the extent that the quintile to which the city in question belongs is stated. In this way explicit ranking is avoided, which is also not appropriate owing to the unre liability of the data, but which neverthe less gives a rough idea of the “position” of a city in the field of Urban Audit cities.