Navigation – Plan du site

AccueilTransversalitésDossiers1998La science régionale : Journées d...Optimal business location in Fran...

La science régionale : Journées d'études en l'honneur de Jean Paelinck

Optimal business location in France and in Europe

La question de la localisation optimale des entreprises en France et en Europe
Alain Sallez

Entrées d’index

Keywords :

optimal location
Haut de page


I am glad to dedicate to Jean the following paper which endeavours to illustrate this lesson.

Texte intégral

The research presented in this paper would have not been possible without the assistance of my students at ESSEC who conducted necessary surveys and simulations. I am grateful to all of them, and particularly to Eric Weisman who has been the student of J.H.P. Paelinck at Erasmus University.

Originally published in R. DOMANSKI and M. CHATTERJI (1996), eds, Wydawnictwo Naukowe PWN - Varsovia
Jean H.P. Paelinck's writing on urban dynamics had greatly interested me long before I met him in 1975 as President of the Association de Sciences Régionales de Langue Française (A.S.R.D.L.F), after the dramatic and premature death of our common friend Professor Jacques R. Boudeville. I had already appreciated the scholar, I discovered a man of great generosity, humanism, at the disposal of others ; children are not mistaken, mines immediately loved him.
The dramatic events of 1975 lead us to write in common a book dedicated to J.R. Boudeville and to prepare the 1975 congress of the ASRDLF on the theme "Energy and Regions", held at ESSEC in Cergy. Out of this encounter, grew a great affection, shared in both scientific, and in friendly, artistic or family circumstances. During a particularly painful period, Jean helped me to overcome the sadness of life. I wish to let him know how grateful I feel for his affectionate help.
How can the immense contribution of the Regional Scientist be acknowledge. Others will express this recognition better than I, but I feel very honoured to have benefited from his advice and to have been his co-author a number of times.
In particular, in a book written with him and two other former Presidents of the HSRDLF, Professors Tony Bailly and Bernard Guesnier, Jean H.P. Paelinck stated :
"Le problème spatial central de la science régionale peut être vu comme celui d'une dynamique d'un grand nombre de variables humaines dans l'espace géographique ... La science régionale aura comme mission de mieux préparer des actions et de montrer - si nécessaire - les limites des interventions publiques en matière régionale".
J.H.P. Paelinck

Is the concept of optimal location still relevant ?

1Companies' location is a major consideration both from the practical point of view of the regional planner as from that of the regional scientist. Since the beginning of the eighties, the reduction in firms' mobility in most European countries, notably in France, has given international investment: locations and delocations, with a strategic importance for governments. Countries and cities eagerly compete to attract multinational firms. This is the reason why the research underlying this paper has been backed by the French Agency: Délégation à l'Aménagement du Territoire et à l'Action Régionale (DATAR) in charge of regional planning in France. While this Agency has developed specific policies aiming at locating firms in depressed areas in recent decades, thus contributing to a better equilibrium of employment on the national territory, today it has to take into account the competitiveness of the French space amid European nations. In other words, this last remark means that firms should be located at an optimal point if this concept is meaningful.

2This question of optimality is also of particular interest for the regional scientist. In fact, since the writing of Alfred Weber, economists and geographers have better theories and models with which to analyse company location in two directions : the micro-economic and the macro-geographic . We have adopted the first track in line with A. Weber, T. Palender, E. Hoover, W. Isard , M. Webber and many others. But, while these authors have put forward differential running costs according to such location factors as transportation, salaries, energy, raw materials etc, we have endeavoured to introduce differential investment costs into the analysis. The mere observation of differences in land prices and public incentives to new investments for various locations (cities, regions, countries) shows the need for a broadening of the analytical framework.

3On this principle, we have developed some empirical research based on the analysis of forty-five industrial firms currently located in Europe during the eighties and the early nineties. Thirty companies for which we obtained the necessary accounting costs and location constraints through thorough internal investigation, were simulated in thirty-six urban sites in France, Germany, Luxembourg, Italy, Portugal, Ireland and Malta in order to calculate gains and losses associated with each location.

4The method, a discrete one, aimed at ordering the financial results that would obtain for each firm if it had been located on other sites according to four main French and European axies starting from the Paris Region (the focal point in France with a concentration of up to 23 % of the national employment). The comparisons, as is explained in the methodology section, were carried out on the basis of discounted differential costs: investment and running costs. For each company, a real negotiation was performed in at least six simulated towns, and the related costs evaluated.

5Our objective was to check the relevance of the optimal location concept in a non-homogeneous space composed of a limited discrete number of urban location sites. We wanted, in this way, firstly to refute the idea that most premises can be located without damage since they are becoming more and more footlose, and secondly to check the efficiency of public policies aiming at orienting firms towards specific zones in France and abroad.

6The findings of our research seem particularly novel. First, they refute the idea that the decrease in transportation costs and the development of electronic communications would make all locations equivalent for firms. The analysis confirms, on the contrary, the relevance of the optimum location concept. Second, the results demonstrate that national and international firms can economically locate manufacturing and research establishments outside the Paris Metropolis contrary to the prevailing thinking of many French prospectivists. It does not appear, in our view , that firms' and households' propensity to concentrate themselves in the Paris Region due to external economies, notably the size of the labour market and gains in information and transaction costs, is unavoidable. This view is also backed by the actual location choices of medium-sized and small firms or premises in Europe.

7Moreover, we can observe and anticipate the growing importance of an optimal location for firms confronted with a more and more competitive European and world market. With the exception of headquarters and specific strategic functions, it is likely that firms will locate their branches and subsidiaries in provincial cities if they can derive fringe benefits from such locations. In view of this, we can hope for a better balance of French spatial distribution of firms in the future and anticipate keener competition between peripheral European places such as Dublin, Malta, Portugal and eastern Germany, as revealed by our research.

8After a brief outline of the methodology adopted, we shall present

  1. the global findings in terms of optimality

  2. a number of results related to the efficiency of public incentives,

  3. an appraisal of the perception of location advantages by managers.


9The principle of the research consists in comparing differential cash-flows depending on location factors for a sample of thirty industrial firms located in Europe between 1982 and 1992. The comparison is made between the actual location chosen and at least of six other selected locations. The comparative cost calculation are based on full investment costs and differential running costs with the Paris Region sites. All costs are expressed in 1990 francs. The running costs cash-flows are considered for a 15-year period, using interest rates from 1 % to 20 %.

10A number of conventions had to be adopted, the main ones being that, for the 15 operating years, the firm:

  • keeps its annual output constant

  • uses the same raw material sources,

  • delivers to the same market places,

  • keeps production coefficients constant

11In such conditions, the optimal sites for each firm's unit can be expressed as

12Minimum Ci = Ii +(Di1 - Si1)x(1+a)-1 +... +(Dij - Sij)x(1+a)-j +...+(Di15 - Si15)x(1+a)-15


14i : index of considered sites of location,

15j : index of year, j varying from 1 to 15

16a : interested rate : calculation ranging from 1 % to 20 % margin

17Ii : investment costs in location i

18Dij : additional running costs in i for year j, compared with the Paris Region

19Sij : annual profits in i for year j, compared with the Paris Region

20Ci : total discounted costs for location i.

Considered Costs

Investment costs

21For each site the following costs have been evaluated by direct estimation of the needs of the firm with the managers, by negotiations with local developers and by gathering local statistics. Considered investment costs are:

  • location study

  • purchase of land (or actual value of lease)

  • construction (or actual value of lease)

  • relocation costs

  • special local construction tax ("taxe locale d'équipement")

  • equipment (offices, plants, warehouse, research facilities)

  • costs and allowances for employee displacement

  • redundancy allowances

  • training costs of new employee

  • governmental public subsidies ("prime d'aménagement du territoire" in France)

  • local public subsidies (prime régionale à l'emploi in France)

Differential running costs in comparison with the Paris Region

22These costs have been similarly evaluated with managers, local developers, and by the use of statistics and data banks.

23Main comparative costs:

  • Fiscal incentives (level and temporary exemption such as "taxe professionnelle in France)

  • tax on profits (level and local exemptions)

  • salaries

  • discounted loans granted by local authorities

  • transportation of products and raw materials

  • costs of executives travel

  • telecommunications costs.


24Not aiming at a statistical representativity, the sample of firms has been chosen among the so-called "foot-lose" industries. Four main characteristics have been considered: the size of the unit (number of employees and volume of sales), the type of industry ,the function of the business (production, R. and D, logistics ...), the location (according to public subsidies zoning).











10-50 employees





headquarters and administrations





Paris Region

(Ile de France)




51-100 employees









101-200 employees





R and D




Paris Basin

(200 Km around Paris)







Production and assembly lines





Rest of France








European countries

(excluding France)














Characteristics of the sample

- Electronics, Computers, Telecommunications


- Car industry


- Pharmaceutical industry, medical appliances


- Cosmetics, biotechnologies


- Film industry, Press


- Furniture, decoration


- Chemistry, rubber


- Industrial hardware and porcelain


- Airframe industry


- Food industry


- Optical instruments







Actual locations and sites retained for simulation

25Both actual and simulated locations are displayed along to four axes starting from Paris. This choice is justified from a national and an international point of view. A number of researchers (geographers, economists, Institut d'Aménagement et d'Urbanisme de l'Ile de France), have demonstrated that, since the middle of the 19th century, firms have moved from the core of Paris towards the suburbs, then to remote locations in France, according to extending far out directions depending on their initial Paris location (fidélité directionnelle)

26Similarly, but originated from other countries, foreign physical investments in France are mostly concentrated in regions close to the borders of origin as shown by DATAR maps.

27In view of this logic, sites of simulation for each firm have been selected along one of four specific axies originating in Paris :

  • North-East of France, Luxembourg, Eastern Germany

  • North-West of France, Ireland

  • North-West of France, Italy, Malta,

  • South-West of France, Portugal



European towns

(excluding France)



Provincial towns


Paris Basin towns


Paris Region towns


North East









Toul (54)

Verdun (54)

Strasbourg (64)


Auxerre (89)

Troyes (10)

Villenauxe (10)


Montreuil (93)

Marne la Vallée (77)







Angers (49)

Thouarcé (49)

Candé (49)

Baugé (49)


Le Mans (72)

Rouen (76)

Le Havre (76)


Cergy (95)

St Quentin en Yvelines (78)









Montpellier (34)

Clermont Ferrand (63)

Valence (26)




Nevers (58)



Evry (91)







Bayonne (64)

Toulouse (31)



Tours (37)

Vendôme (41)


Fresnes (94)

St Quentin en Yvelines (78)


List of simulated locations (cities)

(numbers indicate departmental codes)

(numbers indicate departmental codes)

The four axies of simulation

The four axies of simulation

Map of sites

Actual locations and optimality : a major economic stake ?

28Three main questions have oriented the analysis:

  • is the actual location of firms the best one from an economic point of view among the simulated sites along the firm's
    axis ?

  • Are the sites of the Paris Region competitive in respect of other sites ?

  • are the gains associated with an optimal location significant in comparison with the benefits generated by the firm's actual location ?

Under-optimality of choices in France, but right choices by multinational firms in foreign countries

29The main results summed up in the following have been calculated for a 10 % rate of interest. They remain very stable for any rate below 20 %.

30First finding, 17 units out of 30 are optimally located in the limited framework of the study. Optimal locations are revealed as such by axis of analysis :

  • Paris Region : 4 cases out of 5

  • Paris Basin : 1 case out of 4

  • Rest of France : 3 cases out of 11

  • Foreign : 9 cases out of s0

31We might suggest that French firms have studied their location less carefully than foreign firms considering that many French firms of the sample did locate in France, while the best choices had been made 8 times out of 20 in France and 9 times out of 10 in peripheral European cities. It would be hazardous to pursue further the analysis on this point. Nevertheless, let us propose three comments.

  • Might the under-optimal locations in France be associated with a prevailing importance of qualitative and even affective criteria in the minds of French business leaders ?

  • The decision by most international firms of the sample, to locate their branches , had been preceded by systematic location studies, comparing different sites in several European countries, which seems sensible considering the size of the investment. This remark applies notably to establishments settled in Dublin.

  • There is no single region whose towns or cities present an absolute advantage for all kinds of firms.

Sites and prevailing advantage revealed by simulations

The Paris Region too expensive for manufacturing plants ; towns of the Paris Basin and the peripheral of Europe competitive.

32A first step, simulating the location of the 30 firms only on French sites, reveals the lack of competitiveness of the Paris Region for plants, due to higher investment and running costs ; this is true even though we add a 7 % overproductivity to Paris area towns.

33Best locations in France for the 30 firms or establishments simulated in 28 sites :

  • Paris Region : 5 out of 30

  • Paris Basin : 14

  • Provinces : 11

34If foreign sites are introduced, and the main functions of the firms or establishments considered, the optimal locations appear as shown in the following table.




Paris Region

(Ile de France)


Paris Basin







Headquarters and administrative









R and D






 Manufacturing and assembly line

















Optimal locations for the 30 industrial business according to their main function

35It is interesting to explain what the advantages are, in terms of comparative costs by region corresponding to the above spatial optimal distribution.

The Paris Region towns

36Sites of the Paris Region are too expensive for production and assembly line since investment costs as well as salary expenses generally diminish as the distance from Paris grows. In particular, real-estate costs are so high that they cannot be compensated by urban economies. On the otherhand, headquarters and administrative (decision making) establishments benefit from the major "Hub" represented by Paris, which allows the minimization of communications and travel expenses.

Paris Basin and the Provinces

37Various costs such as salaries or incentives (subsidies, tax allowances) over compensate for the higher costs generated by transportation, travel and communications costs for firms located in towns of the Paris Basin and the Provinces. But the dominating major cost in favour of these towns is the low price of land in industrial parks.

38To be noted, the advantageous location in provincial towns for R. and D., which would be meaningless if we did not know from DATAR sources that 89 % of R. and D employment was located outside the Paris Region by multinational firms in 1994.

Sites on the periphery of Europe

39Only such foreign towns in the sample are attractive to firms. Luxembourg and Milano do not present comparative advantages as opposed to remote places which benefit from decisive differential costs in three fields : tax exemptions, incentives for foreign investments, low costs of skilled labour.

40In Ireland and Eastern Germany, subsidies and loans at lower rates are particularly efficient. Tax on profits at 10 % in Ireland is of foremost interest for highly profitable firms, while the sale or the lease of real estate, at an extremely low cost, explains major differential costs for investments. Malta holds simultaneously comparative advantages : very low wages, attractive taxation, lease of real estate by the Government at a very low cost. Finally, Portugal can be optimal for labour intensive firms.

Are gains resulting from an optimal location significant ?

41In the above sections, it has been shown that a hierarchy of locations actually exists for each firm or establishment. It has not been demonstrated that the economic advantage of an optimal location over a random choice merited a location study based on precise calculations.

42So, we have compared for each firm or establishment the gain resulting from an optimal location, along its geographical axis, over a location in the Paris Region, in order to evaluate the economic impact of an optimal or quasi-optimal location. The comparison thus relies on the annual difference between the two stated locations, for a 10 % interest rate. This difference is then compared with the annual mean (3 years) of the business profit before corporate tax , if it is a mono-site one, or for the branch (estimated profit on the basis of a split of the firm's profit).

43Let us recall, before considering the general results of these calculation for the sample of firms, that every firm simulated in a town of the Paris area was granted a 7 % productivity surplus corresponding to urban economies. In such conditions, if all the firms or establishments of the sample were optimally located (in the framework of our simulation), the median gain as above defined, would amount to 45 % of the business profit before tax.. notwithstanding of the limitations of the research, we thus feel confirmed to strongly hypothetise that an optimal location constitutes a significant economic variable for a firm, even today.

Are public policies efficient?

44Simulations have been concentrated on three kinds of public
policies : tax on profits, local tax on business ("taxe professionnelle" in France), national and local subsidies ("prime d'Aménagement du Territoire et prime régionale à l'emploi" in France) .

45Only Ireland (Dublin) and Malta offer a substantial differential rate of tax on profits compared with France. Even when one withdraws this advantage from the simulation, Dublin and Malta remain attractive for the same firms as before in 5 cases out of six.

46Similarly, we have simulated an abolition of both advantages concerning local taxes on business, i.e. temporary exemption in specific zones and differences in rate. The order of competitiveness of sites for each firm is not modified, even if the differential gains are reduced. This result obviously questions the efficiency of this kind of public policy, at least on the national scale.

47Thirdy, we have simulated the suppression of advantages resulting from national and local subsidies according to national zonings in 1990 and in 1995. Let us recall that in France, these national subsidies represented 35 - 50 000 francs per created job within a limit of 17 % to 25 % of the total investment (real estate plus equipment) according to a spatial zoning (see map below) in 1990. These figures have been increased by 40 % in 1995, but remain far below those permitted by Brussels in peripheral European regions. Similarly, the withdrawal of subsidies does not substantially modify the ranking of sites, since only one optimal site is modified in the case of the 1995 rate of subsidies.

48Finally, after having simulated the withdrawal of the three kinds of public aids, one at a time, we actually simulated the simultaneous suppression of all public aids. The order of optimal sites was only marginally modified.

49These results might suggest that public spatial policies are not very efficient on a national scale, even if they represent an appreciable aid for some firms. To be more efficient, the zoning should be more selective and the amount of subsidies higher. It is doubtful whether the present subsidies could be suppressed, even if they are inefficient, due to the keen European competition of countries and cities to attract multinational firms' new branches.

Perception of location factors and cost advantages by managers

50Knowing the impact of qualitative factors : quality of life, infrastructure, image of the city, political and business environment ... for location decision making, business leaders in the sample were questioned on

  1. the relative importance of qualitative and quantitative factors,

  2. their evaluation of the quantifiable factors considered for the location of their firm or establishment.

Quantifiable factors

Qualitative factors

Price of land

Proximity of clients

Proximity of suppliers

Proximity of subcontractors

Cost of energy

Cost of raw materials

Wage level

Financial incentives

Labour attitudes

Business services

Household services

Quality of life

Image of the city

Local political support

Quality of business park

Efficient transportation infrastructure availability

Location factors submitted to business heads' appraisal

51Five qualitative factors : transportation infrastructure availability, quality of life, image of the city, local political support, quality of business park, were, on average, chosen first. This point confirms the considerable weight of qualitative factors in the decision- making process.

52However, because of our hypothesis concerning the importance of economic factors, for each firm or establishment, we have compared the perception of managers with actual prevailing differential costs resulting from our calculations for each firm. It is not our purpose to push the analysis into further detail, but we shall retain the main outcome which shows a ratio of 50 % of the accuracy of managers' appraisal with the prevailing differential costs due to location factors according to our own simulations ; a finding which can be judged pessimistic or optimistic according to one's personal judgement.


53Within the scope of the simulations permitted by our cost oriented research, we can clearly conclude that

  1. an optimal location is of economic importance, even nowadays, for a firm

  2. public policies aimed at orientating firms towards specific zones look relatively inefficient from a strictly economic point of view ;

  3. considering the growing competition between firms in the European market, it seems likely that business will attach growing importance to an optimal (or quasi-optimal) location for their new plants or offices both from the quantitative and from the qualitative point of view ;

  4. some peripheral cities of Europe will probably be perceived as increasingly attractive locations for international physical investments.

54Of course, the sample of our research is too narrow to provide a statistical evidence to the above assumptions. The convergence of many findings resulting from our simulations confers at least a fair presumption of the reliability of our conclusions, which tend to show some limits of public intervention in regional planning, one of the wise teachings of Professor J.H.P. PAELINCK.

Graphs of total discounted differential cash-flows (y axis) for different locations according to interest rates (x axis) for two firms of the sample

Graphs of total discounted differential cash-flows (y axis) for different locations according to interest rates (x axis) for two firms of the sample

Notes :

1J.R. Boudeville, Functional Regional Analysis : an elementary exposition of some selected topics edited by J.H.P. Paelinck and A. Sallez. Editions de l'ASRDLF 1978

2This kind of logic is highly correlated with the French road and railroad networks in the shape of a spider's web with Paris in the center

3This decision was justified by results advanced by R. Prud'homme and M. P. Rousseau in France (1992) and by Segal, Sveikanskas, Moorar, Henderson in U.S.A.

4For the sample, the distribution of the ratio : annual gains resulting from an optimal location over a Paris area one/annuel business profit before tax is :

0 to 10 % = 20 % ; 10 % to 30 % = 28 % ; > 30 % = 52 %

5Ireland : 10 % for foreign firms exporting 90 % of their production.

Malta : 0 % for 10 years for firms exporting 95 % of their production ; 5 % in other cases

France : 33,3 % except in a few specific conversion zones

Luxembourg : same rate as in France with a possible rebate of one fourth for a new location.

Haut de page


BENKO G. - LIPIETZ - Les régions qui gagnent - PUF - 1992

BENKO G. - Les nouveaux espaces industriels dans le monde in Géographie des Technopôles - Masson - 1991

BIZET B. - 10992 - L'entreprise dans les Etats de la CEE - Delmas

DAMETTE F. , 1994 - La France en Villes - La Documentation Française


FERNIOT B. - in "La Décentralisation Industrielle" - IAURIF - 1976

GERI- Banque de France, 1994 - Localisations et performances : la situation des entreprises industrielles régionales en 1992 - Centrale des Bilans

HENDERSON, J.Y., Urban Development, Theory fact and Illusion, Oxford : Oxford University Press, 1988

Invest in France - DATAR, 1994 - Décisions d'investissements étrangers en France annoncées en 1994. Paris.

ISARD W. - "Methods of Regional Analysis". MIT Press 1960

PEREUR J. - "La localisation des unités de production" in Analyse économique spatiale sous la direction de Cl. Ponsard - PUF - 1988

PRUD'HOMME R., ROUSSEAU M.P.- 1992 - Les avantages de la concentration parisienne, l'OEIL

RALLET A. - 1994 - La polarisation de l'Espace in Communiquer demain - dirigé par Pierre MUSSO - DATAR - Ed. de l'Aube

RALLET A. - Réseaux de télécommunication - Localisation des activités informationnelles et hiérarchie urbaine - Communication au colloque IFRESI- Pir Villes - Lille 17-18 mars 1994

SAVY M. - VELTZ P. - 1993 - Les nouveaux espaces de l'entreprise - DATAR - Editions de l'Aube

SVEIKAUSKAS, L. ; GOWDY J., and FUNK, M. ; Urban Productivité : City Size or Industry Size, Journal of Regional Science, Vol. 28, n° 2, 1988

TIEBOUT C.H. - 1956, A pure theory of local expeditures in Journal of Political Economy

WEBBER M. - 1972 - Impact of Uncertainty on Location - MIT Press

WEBER Alfred, Uber den Standort der Industrien -Tübingen : JCB Mohr, 1909

Haut de page

Table des illustrations

Titre (numbers indicate departmental codes)
Fichier image/png, 38k
Titre The four axies of simulation
Fichier image/png, 105k
Fichier image/png, 8,2k
Titre Graphs of total discounted differential cash-flows (y axis) for different locations according to interest rates (x axis) for two firms of the sample
Fichier image/png, 7,4k
Haut de page

Pour citer cet article

Référence électronique

Alain Sallez, « Optimal business location in France and in Europe », Cybergeo: European Journal of Geography [En ligne], Dossiers, document 51, mis en ligne le 13 mars 1998, consulté le 08 août 2022. URL : ; DOI :

Haut de page


Alain Sallez

Professor at ESSEC, France

Haut de page

Droits d’auteur

Licence Creative Commons
La revue Cybergeo est mise à disposition selon les termes de la Licence Creative Commons Attribution 4.0 International.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search