The three-year period from 1998 to

Transcription

1 ISSN Traduction de la Note d information 02.53, parue en novembre 2002 The increase in gross domestic expenditure on R&D has been one of the highest over the past decade, especially after a sharp upturn in Growth of industrial research expenditure, slightly below that of the GDP, is still faster than that of the government sector. Funding by the government sector for industrial research has increased at the same rate as industrial-sector research expenditure, thus contributing to healthy industrial-sector research. The number of individuals employed in research rose by 3.7% in 2000, as did the overall level of qualifications, with almost 7% more researchers. Forecasts for 2001 suggest continued increases in research expenditure, with the government sector playing akeypart. R&D expenditure in France in 2000 First estimates for 2001 The three-year period from 1998 to 2000 was one of steady growth of the French economy, owing both to domestic demand, especially household consumption, and to exports. Firms improved their profit margins and started to invest again. In 2001, under the effect of a slackening world economy, the GDP rose by only 1.8% in volume, compared to 3.8% in On the other hand, price increases were twice their 2000 levels. After substantial research expenditure increases in 1999, the years 2000 and 2001 were a period of consolidation. In a less favourable economic climate in 2001, the gross domestic expenditure on R&D (GERD) rose by 4.1% in value and 2.6% in volume,totalling 32.2bneurosasopposed to 30.9bn euros the previous year. Since the GDP rose less steeply than the GERD in 2001,theshareofGERDintheGDPmoved up to 2.20%, from 2.19% in 2000 and 2.18% in This increase in R&D expenditure is imputed to both the government sector (+4.3% in value) and the industrial sector (+4% in value). SINCE 1998 THE EXECUTION AND FINANCING OF R&D HAVE BEEN STABLE All funding by French private- and publicsector organizations, of research performed in France and abroad, makes up the Gross National Expenditure on R&D (GNERD). This GNERD totalled 31.44bn euros in 2000 and is estimated at euros for From the early 1990s, the structures responsible for the of execution (GERD) and financing (GNERD) of research changed as the weight of the industrial sector increased. The share of government-sector funding TABLE I GNERDandGERDtrends(million of euros) (3) 2001 * Funding GNERD Annual growth rate in volume [%] (2) - 1,7 0,4 1,6 3,5 4,4 2,4 Government-sector funding (1) Industrial-sector funding Government-sectorfunding/GNERD[%] 50,10 49,50 44,70 44,40 45,40 45,50 Execution GERD Share of the GERD in the GDP [%] 2,34 2,31 2,18 2,18 2,19 2,2 Annual growth rate in volume [%] (2) - 0,6 0,3 1,2 3,7 1,2 2,6 Execution by government sector (1) Execution by industrial sector Execution by industrial sector / GERD [%] 61,80 62,10 62,30 63,20 62,50 62,40 Source : MEN-DPD C3 (1) Public- and private-sector administrations (state, higher education and not-for-profit institutions) (2) Assessed on the basis of GDP price indexes (3) Break: see methodological note * Estimates

2 TABLE II Funding of public-sector research, excluding defense, in 2000 (in million of euros) State subsidies constantly declined and has been below 50% since From 1998 to 2000 this trend seemed to level off, with firms accounting for approximately 62% of the GERD and government-sector funding between 44% and 45% of the GNERD. Several factors explain this trend. The steep decline of the Defence Ministry s research budget was offset by a slight increase in contracts with firms, and the civil budget for research and development (CBRD) remained stable. Although the share of the CBRD in the GDP continued to decline, variations in volume from one year to the next were stable (around 1%) in In 2001, government-sector funding is expected to rise more steeply than industrial-sector funding and to exceed 45%, given the Defence Ministry s projected research budget increases. PUBLIC-SECTOR RESEARCH: AREVIVALIN2001 Government-sector GERD totalled 11,605m euros in 2000 and is expected to increase by 2.9% in volume, to total 12,105m euros in A revival is thus forecast, after two years of very little or even no growth. In 2000, expenditure by scientific and technological public-sector organizations (STPO) remained stable in terms of volume. This was due to cuts in CNRS (Centre national de recherche scientifique) domestic expenditure, while all the other STPO enjoyed a more favourable situation. Expenditure by industrial and commercial public-sector organizations(icpo) dropped by 5.8% in terms of volume, with the constant method. This decrease is concentrated in the CNES (Centre national d études spatiales) and the CEA (Centre d études atomiques). The constant increase in universities research spending (+4% in volume) was not enough to offset these two trends /1999 in volume (%) R&D State contractsand Total subsidies other income R&D contractsand other income Total budget excl defence ,60-5,80-0,20 Of which ICPO ,30-9,90-4,20 STPO ,60-5,40-0,20 Universities ,40-5,90 1,90 Source : MEN-DPD C3 The increase in the government-sector GERD (GOVERD + HERD) forecast for 2001 ( m euros) is expected to result primarily from the steady increase in expenditure by the Defence Ministry (+ 180m euros, +23%), taking into account the recorded expenditure on simulated nuclear tests, that of the CNRS(over 100m euros,up6% invalue) and that ofotherstpo(+ 75meuros,up6% in value). IN 2000, FOR THE SECOND YEAR RUNNING, EXTERNAL INCOME OF PUBLIC-SECTOR RESEARCH DECLINED Total The total R&D budget of public-sector organizations and services is based on two types of complementary funding: subsidies and external income, usually from contracts. Over three-quarters of the income of public-sector civilian research is from subsidies. 55% of all subsidies are from the civil budget for R&D (CBRD) and 23% are from other ministerial contributions: primarily the Youth Ministry, National Education and Research Ministry (funding for higher education, 2,763m euros) or the Ministries of Infrastructure, Transport, Housing, Tourism and Maritime Affairs. Contract-generated income and own income complete subsidies. Research contracts signed by public-sector research provides 16% of its funds, i.e. 2,172m euros. It is above all within the public sector its that research contracts are signed (for a total of 1,104m euros); research organizations in the sector are linked by a complex web of sub-contracting of research work. They also have contracts with firms (for a total of 592m euros) and receive 475m euros from abroad, of which 200m euros from other European countries. Finally, 6% of public-sector research income ( 832m euros) is from own resources. Income from intellectual property rights, donations and legacies each amount to 135m euros. The year 2000 confirmed the trend towards a decline in non-subsidy income forecast in 1999 and an increase in income from subsidies. Contract-generated income decreased, with the constant method, by5.8% in volumecomparedto1999,while subsidies increased by 1.6%. Over the year, the CBRD increased by 0.6% in volume while university research funding increased by 3.0%. All public-sector research, ICPO, STPO and universities, recorded a drop in non-subsidy income. The decrease in contracts with firms is the main explanation. PUBLIC-SECTOR FUNDING FOR FIRMS TOOK A NEW UPWARD TURN IN 2000 From 1992 government funding of firms regularly declined. The year 1999 marked an inversion of this trend with increases in volume of both civil public-sector funding for research and grants from the Defence Ministry. The same trend was recorded in 2000 with comparable funding for civil research and a steady increase in funds from the Defence Ministry. With 2,259m euros public funding of industrial-sector research was higher than the mean recorded over the previous six years. Contracts signed by the Defence Ministry, which account for two-thirds of publicsector funding, total 1,498m euros, up 6.3% in value and 5.6% in volume. Forecasts for 2001 show an even bigger increase. Funding for civilian research is linked primarily to large technological programmes (aeronautics, space, electro-nuclear, and information and communication industries) but also to a multitude of grants from ministries or public research institutions such as the CEA (atomic energy) and the CNES (space). In 2000 all this funding together attained a level only slightly inferior to that of In 2000, funds provided by the department of civil aeronautics (Direction des programmes aéronautiques -DPAC)for industrial research reverted to their 1996 level. However, funds from the DPAC are characterized by sharp fluctuations from one year to the next due to time lapses between the execution of R&D work and the notification of payments. Irrespective of Page 2

3 TABLE III Origin of public-sector funding for firms (in millions of euros) these fluctuations, funding in 2000 was high once again, after decreasing in 1997 and 1998, and by comparison with the mean over six years. Overall funding from the Ministry of Industry also stalled between 1996 and 1997 and has remained at a level of around 150m euros since ANVAR grants included, there was an overall increase in volume of 1.6%. The volume of funding from the Research Ministry increased by 23.1% in This rise reflects the 18% increase in grants from the Technological Research Fund (FRT) and the Scientific Support Fund (FNS), of which only a little more than a third was earmarked for the industrial sector. In 2000 government thus financed 11.4% of the industrial-sector GERD, a percentage identical to that of Publicsector funding increased at the same pace as the BERD, whereas it had increased more rapidly in1999.itsgrowthisexpected to be higher than that of the BERD in FIRMS CONTINUED THEIR RESEARCH EFFORTS IN 2000 Industrial-sector research expenditure totalled 19.3bn euros in Although the increase in research expenditure was lower than in 1999 (+2.9% in terms of volume, compared to +5.2% in 1999), it was still well above increases recorded since A favourable economic situation since 1998 and investments in new technologies encouraged firms to invest in research. Increasing research expenditure is attended by growing numbers of people employed in research, especially researchers. This automatically induces a faster increase in research staff s salaries (+4.4% comparedto3.5%involumein1999).thisis the first time since 1992 that the number of individuals employed in research has risen Mean over 6 years in volume (in %) Defence funding ,6 Public-sector civilian funding of which ,4 Civilian aeronautics programmes ,2 Min. of industry and ANVAR* ,6 Min. responsibl for research ,1 Total funding ,1 Source : MEN-DPD-C3 * Agence nationale pour la valorisation de la recherche (national agency for the promotion of research) substantially for two consecutive years. While expenditure on equipment hardly increased in 1999, it rose steeply in PREDOMINANCE OF A FEW LARGE COMPANIES AND GROUPS Research expenditure is concentrated in large companies: 113 firms each employ over 100 researchers, a total of over 60% of all researchers employed by firms. They account for two-thirds of domestic research expenditure and receive 86% of all government grants. This concentration is even more marked if we look at the subsidiaries of large groups. Alone, the 13 majorfrench groups in research account for half of all industrial research expenditure. The car and telecommunications industries, the two main branches of research, are on a par Four branches each account for over 10% of the BERD. Together they account for 50% of research expenditure in France, compared to 47% in In 2000, as in 1999, the main research branch in France was the car industry, an area with average technological intensity whereas the other main research branches are high-tech. For the second consecutive year research spending in the car industry rose steeply (+6.4%, after +18.1% in 1999). This leading position is explained by the weight of this industrial sector in which France has a number of leading players (car and parts manufacturers). The second largest research branch radio, TV & communication equipment and components is driven by new information and communication technologies(nict), e- lectronic components and telephony. Since 1992 theweightofthissectorinthegerdhas increased by close to four points, from 1.6bn euros to 2.7bn euros in Research efforts in the pharmaceutical sector stalled after increasing substantially in 1999(+8.7%). This was the first time since 1992 that the pharmaceutical industry s share in R&D expenditure dropped. From 1992 to 2000, its overall share climbed by three points (from 9.5% to 12.4%). Aeronautical construction, France s main branch of research until 1996, is now in fourth place with hardly more than 10% of the BERD. Since 1992 this sector s share in industrial research expenditure has fallen by seven points. The sector was badly hit by the fact that a number of programmes reached their term and that government funding declined in the 1990s. Defence funding has remained at a relatively low level for several years although it has started to increase again in the instrumentation industry (precision, measurement and optical instruments). Consequently, the growth curve of this sector, France s fifth largest research branch, accounting for 6.8% of industrial R&D, has been positive since Among the other branches of research, those related to NICT show varying trends. While transport and telecommunication (especially the latter) and information technology services have grown over the past few years, computer equipment manufacturers are still on a downward slope. The popularization of computer equi- TABLE IV Breakdown of the BERD in the main research branches Domestic expenditure on R&D Public funding by firms in million d euros of BERD en M d euros en % en % Car industry ,8 0,5 Radio, TV & communication equipment and components ,7 18,7 Pharmaceutical industry ,4 0,5 Aeronautical and spatial construction ,2 37,5 Medical, precision and optical instruments ,8 18 Chemical industry ,1 2,4 Transport and communication services 995 5,1 0,5 Machine and equipment manufacturers 921 4,8 10,9 Engineering services 564 2,9 3,5 Information technology services 496 2,6 1,6 Office machine and equipment manufacturers 286 1,5 1,2 Other research branches ,1 4,7 TotalBERD ,0 100,0 Source MEN-DPD C3 Page 3

4 TABLE V Employee numbers (excl. Defense) (1) (2) Researchers in industrial sector Researchers in government sector of which PhD students Total R&D staff in industrial sector Total R&D staff in government sector Total number of researchers Total number of R&D staff Source MEN-DPD C3(1) Revised (1) Revised (2) Methodological change pment and manufacturers reorientation towards other areas (e.g. information technology services, telephony) explain the growing weakness of this sector in France. GROWTHINNUMBERS AND QUALIFICATION OF RESEARCH STAFF The total numbers of personnel employed in 2000 for R&D activities, excluding defence, was 319,373 full-time equivalents (FTE), of whom 44% were employed by the government and 56% by firms. Staff numbers rose by 3.7% from 1999, with a more marked increase in numbers of researchers (6.9%). In firms the overall increase was 3.6% while the increase in numbers of researchers was 7.5%. Since 1992 increasing numbers of R&D staff has been accompanied by higher overall qualifications and thus more researchers. Over an eight-year period, the number of researchers increased by 16,300 while support staff declined by 3,000 FTE. Researchers accounted for 46% of all R&D staffin 2000although there are wide disparities from one sector to the next: 57% for the communication sector (with over 61% for the precision instrument branch) and only 34% for the car industry. The government sector (excluding defence) has approximately 141,700 FTE on its payroll. Almost 50% of these employees are researchers or lecturing researchers, 13% are PhD students and a little under 40% have technical and administrative support positions. The two main pillars of French publicsector research, the CNRS and the universities, employ 60% of all personnel. The main employer is the universities, with 58,000 FTE of which 31,900 are lecturinger- researchers. Since 1992, close to 15,000 new lecturing researcher posts (not including secondary school teachers lecturing in higher education) have been taken into account in the higher education budget. In terms of research, this is an increase of 7,500 FTE. The CNRS has 27,600 FTE on its payroll. The remaining 40% of employees in research work for other institutions. STPO (excluding the CNRS) employ about 18,000 FTE, of which 8,500 work for INRA and 5,300 for INSERM. In the ICPO (21,200 FTE), the main employers are the CEA (12,000), the CNES (2,600), ONERA (1,900) and IFREMER (1,400). In 2000, beneficiaries of research grants provided for in the CBRD accounted for 11,524 FTE out of a total of 18,117 PhD bursars and grant recipients. The training through research effort remained steady throughout the entire period in respect of both the CBCRD (from 1993, 3,800 research grants were allocated every year and 3,900 in 2000) and charitable organizations and associations, especially in the health sector. REGIONAL SPECIALIZATION The intensity of R&D varies, depending on the region, and differences stem not only from the economic strength of each one. The ratio of R&D staff to all salaried jobs enables us to highlight regional characteristics. In all regions taken as a whole this ratio is 1.42%, with 0.82% for the industrial sector and 0.61% for the government sector, but stark regional differences exist. Six regions exceed the mean ratio: Îlede-France (2.50%), Midi-Pyrénées (1.97%), Rhône-Alpes (1.60%), Languedoc-Roussillon (1.48%), Auvergne (1.43%) and Franche-Comté (1.42%). But the reasonsforthisintensity of R&Dare not identical; they stem from a high intensity Permanent staff and the others Of the 202,700 natural persons remunerated by public-sector research (excluding defence), 150,600 are civil servants or permanent contract workers, 20,600 are PhD bursars and research grant holders, and 20,000 are natural persons with temporary positions. Hence, French public-sector research is based essentially on permanent staff although it also functions with a substantial number of temporary employees, especially in the universities and university hospitals (CHU). TABLE VI Status of public-sector research personnel (excl. defence) in 2000 Permanent Temporary Not specified* Total Researchers Research engineers and technicians Other S/total PhD students Total Source: MEN-DPD C3 * Questions on status were not generalized throughout the survey. PhD students excluded, two-thirds of all employees are researchers, a little under 10% are technicians, and a quarter are administrative support staff. This group is characterized by a wide diversity. Different types of job contract attract qualified manpower to public-sector research, enhancing research teams both qualitatively and quantitatively. Temporary research positions, particularly at universities, include associate or guest researchers, both French and foreign (especially in the framework of high-level practical training periods organized by the Foreign Ministry), whose presence contributes towards the openness of the university. These statuses also allow employment of young researchers as a transition between training through research and stable employment. These are mostly «attachés temporaires d enseignement et de recherche» (temporary teaching and research posts) and young researchers paid from contract-generated income. But this type of work contract also corresponds to a search for quantitative flexibility, especially for university hospitals where over a third of the employees have temporary jobs. In STPO, use of these forms of work contract is more frequent for non-research staff, especially technicians. Their remuneration may be from laboratories contract-generated income. Page 4

5 GRAPH 2 Share of R&D employment in total salaried employment in mainland France in 2000 Total métropole régionalisé Total hors Île de France Corse Limousin Champagne-Ardenne Poitou-Charentes Basse-Normandie Picardie Bourgogne Franche-Comté Haute-Normandie Auvergne Lorraine Nord-Pas-de-Calais Alsace PaysdelaLoire Centre Aquitaine Languedoc-Roussillon Bretagne Midi-Pyrénées PACA Rhône-Alpes Île-de-France either in both the government and industrial sectors (Île-de-France, Midi-Pyrénées and Rhône-Alpes), or in the industrial sector only (Franche-Comté and Auvergne), or else in the government sector only (Languedoc-Roussillon). In view of the strong concentration of research in both the industrial and government sectors in Ile-de-France, this region is unique. It employs 41% of research personnelalthough itaccountsforonly23%of all employment in mainland France. The Les régions sont classées dans l'ordre croissant de l'effectif de R&D Businesses Administrations 0,0 0,5 1,0 1,5 2,0 2,5 «R&D/Total employment» ratio is almost double that of the national mean. The Midi-Pyrénéesregion,specializedin aeronauticaland spatialresearch,employs 5.6%ofallresearchstaff,thusranking fourth, although it ranks eighth for total employment, with 4.1%. Rhône-Alpes, which ranks second in the economy, with 9.9% of all employment, has not only a powerful industrial fabric but also a strong R&D potential. The region employs 12.4% of all R&D staff. In this classification, Franche-Comté stands out. It ranks twentieth on the list of employers (all categories together) but fifteenth when it comes to R&D jobs. This region, where car manufacturing is the dominant industry, is characterized by relative specialization in R&D, with an «R&D/Total employment» ratio higher than the national mean. In Languedoc-Roussillon the high number of public-sector organizations (three-quarters of R&D jobs are government funded) contribute towards this position. In view of the weight of Ile-de-France, it is preferable to compare the other regions to a national mean excluding Ile-de- France. The «R&D/Total employment» ratio then drops to 1.10%, with 0.1% for industry and 0.49% for the government sector.four other regions are then aboveor close to this mean: Provence-Alpes-Côte d Azur (PACA), Brittany, the Centre and Alsace. In Brittany and the Centre it is industrial research that lifts the ratio over the mean. In Alsace and PACA the opposite is true. Philomène Abi-Saab, Monique Bonneau, Catherine David, Florent Favre et Emmanuel Weisenburger, DEP B3 International comparison of the role of the state in R&D funding for the industrial sector Since the beginning of the 1990s the US, UK, Germany and France have reduced the value of government funding (excluding tax credits for research) for research undertaken by firms. This reduction is particularly noteworthy in France and the US and corresponds to the decrease in military funding for firms. Until 1998 the US funded industrial-sector research more than did France and Germany. Trends in this funding show a continuous decline in three countries (US, UK and Germany) since then, contrasted with an upturn in France. Nevertheless, a recent increase in military R&D is observed in the US and France. At the same time the UK has strengthened its tax incentives to support R&D. International comparison of the role of the state in R&D funding for the industrial sector (in %) BERD / GERD Share of the BERD financed by public grants Ratio USA 75,3 20,8 12,3 11,1 UK 65,6 12,5 10,2 8,8 France 64,0 16,4 (1) 9,9 9,9 Germany 70,8 9,9 7,6 7,2 Japan 71,2 1,1 1,8 1,7 (1). Refunds for advances subtracted. Source : OECD, MSTI An essentially French population and openness towards Europe and Africa In 2000 the nationality of over 80% of researchers (excluding beneficiaries of research grants and PhD bursaries) at universities, STPO and ICPO was identified. Over 90% of the researchers in this group are French. The origin of foreign researchers reflects the international scope of French public-sector research and the area of influence it has built up. Over half of the foreign researchers mobilized by French public research are European. Citizens of European Union countries account for up to 40% of all foreign researchers, while those of other European countries account for a little over 15%. Geographic and cultural proximity, historical bonds and the structuring effect of European research programmes probably partly explain this pre-eminence. Africa is second on the list, with over 20%, followed by Asia. North America is less present, with only 9% of the foreigners working in French public-sector research. TABLE VII - Foreign researchers in French universities, STPO and ICPO in 2000 Number of researchers Breakdown of foreigners per country (%) French Foreign Of which European Union countries Other European countries North America South America Africa Asia Other 65 1 Not specified Total Source: MEN-DPD C3 Page 5

6 Methodological note The results presented for the year 2000 are not directly comparable to those of Several methodological changes were made. These changes enhance the quality of data on public-sector research but introduce breaks in series of figures. They include: The evaluation of universities research expenditure has been improved; social charges and civil allowances are now taken into account better. Certain organizations changed their answers in our survey to align them with the pluri-annual contract-plan binding them to the state. Use of a revised questionnaire made it possible to clarify certain concepts (contracts, nature of own resources, number of employees) and induced a change of response by certain organizations. Given the multiplicity of sources of such breaks, it was difficult to produce two series of data for 2000: one by means of a constant method and the other with the new evaluations. The exercise was nevertheless performed on the GERD and GNERD. Thus, in terms of the constant method, the GERD increased in volume by only 1.6% between 1999 and 2000 (compared to 4% with the new method). This increase, inferior to that of the GDP (+3.8% in volume) is the result of a 2.9% increase in the BERD and a 0.7% decrease in the government sector GERD (i.e. GOVERD + HERD (government + higher education expenditure on R&D)). Likewise, the GNERD increased by 1.9% in volume in terms of the constant method (as opposed to 4.4% after methodological changes), the result of an increase of 2.1% in the industrial-sector GNERD and 1.2% in the government-sector GNERD. On the other hand, estimates for 2001 are comparable to data from 2000 and differences between the two years reflect the increase in government-sector R&D, especially in the military sector. GRAPH 1 Rates of changes of volume from 1999 to / / /2000 5,2 5,9 3,2 3,8 1,8 3,7 4 2,6 3,7 1,6 2,6 2,9 2,5 1,2 2,8 1,2 2,8-0,7 GDP GERD GERD (constant method) Source : MEN-DPD C3 BERD GOVERD + HERD GOVERD + HERD (constant method) SOURCES AND DEFINITIONS The above results are drawn from surveys carried out in the government and industrial sectors by the research statistics office. The survey on firms was carried out in 2000 on about 10,050 firms and 50 technical research centres. The survey is exhaustive for large firms and sampled for a part of small and medium-sized enterprises (at a ratio of 1 to 2). In the government sector the survey was performed on ministerial bodies and services which finance and/or execute research work, the engineering «grandes ecoles» not under supervision of the Ministry in charge of Higher Education, and not-forprofit associations that contribute to the national R&D effort. Results for 2000 are drawn from a reviewed survey. GROSS DOMESTIC EXPENDITURE ON R&D (GERD): This expenditure corresponds to R&D work performed on the national territory, irrespective of the origin of the funds. It includes current expenditure (R&D staff salaries and running expenses) as well as capital expenditure (purchases of equipment needed for in-house R&D as well as real estate operations during the year). NATIONAL EXPENDITURE ON R&D (GNERD): This aggregate measures the financial effort of the national economic actors, without double counts, irrespective of the destination of the funds. The difference between the GERD and the GNERD represents the balance of R&D exchanges between France and other countries, including international organizations. RESEARCH BRANCH: This refers to the economic sector benefiting from R&D work, grouped here into 25 categories based on the French list of economic activities (nomenclature d activités française NAF). GOVERNMENT FUNDING OF INDUSTRIAL R&D: this includes government contracts and grants for R&D in industry. It does not include the tax credit for research, a tax incentive set up in 1983 and based on the evolution of R&D expenditure. CIVIL BUDGET FOR R&D (CBRD): the survey on funds spent on R&D in public-sector organizations and services enables us to identify, for 2000, 7,248m euros of the 8,331m euros listed in the BCRD as ordinary expenditure and subsidies. This discrepancy is explained by a definition of the field of the survey consistent with international statistical definitions relative to R&D. The «Cité des sciences et de l industrie» and funding from ANVAR and the Industry Ministry for innovation are therefore excluded. Moreover, all this income is considered tax exclusive. Direction de l'évaluation et de la prospective Directeur de la publication Claudine PERETTI Rédactrice en chef Francine LE NEVEU Maquette et impression DEP bureau de l'édition SERVICE VENTE DEP, bureau de l'édition 58 bd du Lycée, VANVES ABONNEMENT ANNUEL France : 42,69 euros Étranger : 45,73 euros Page 6