1jrc june 2008 statistical research at the joint research centre andrea saltelli,...
TRANSCRIPT
1JRC June 2008
Statistical research at the Joint Research
Centre
Andrea Saltelli,[email protected]
NTTS Conference, Brussels,
February 2009
2JRC June 2008
ERAWATCH Unit at JRC-IPTS (Seville)
Contact: Pietro Moncada Paterno’ Castello (JRC IPTS)[email protected]
3JRC June 2008
4JRC June 2008
- The EU Industrial R&D Investment Scoreboard: analysis of 1000 EU and 1000 non-EU top investing companies in R&D
- The EU Survey of Business R&D- Economic and policy analysis of corporate R&D.
Industrial Research and Innovation at JRC - Seville http://iri.jrc.ec.europa.eu/
5JRC June 2008
Some Results (1): Nature of the R&D investment gap
88 % 12 %EU - US
Sectoral composition effect Underinvestment effect
Total: -1.8%
Breakdown of EU-US gap in R&D intensity into sectoral composition and underinvestment effects.
Note: In this figure, only companies of similar R&D size are considered, i.e. the top EU (391) and the US (563) with R&D investment above a common threshold (€23 million in the latest year).
Source: JRC-IPTS calculations (2008) - Analysis of 2007 EU Industrial R&D Investment Scoreboard
- EU's R&D intensity deficit is largely explained by the different industrial structure (sectoral composition effect).
Industrial Research and Innovation
6JRC June 2008
R&D rises productivity much more in high-tech sectors than in low-tech ones
Some Results (2): Econometrics of R&D & firm productivity
Industrial Research and Innovation
-6-4
-20
-6-4
-20
-8 -6 -4 -2 0 -8 -6 -4 -2 0
High Low
Medium Total
Pro
duct
ivity
/Em
ploy
ee
R&D stock/employeeGraphs by R&D Intensity sector groups
R&D Stock/Employee vs. Productivity/Employee
Source: European Commission, JRC –IPTS (2008) Analysis of 2007 EU Industrial R&D Investment Scoreboard
7JRC June 2008
Agrilife Unit at JRC-IPTS (Seville) Contact: Marc Müller (JRC IPTS)
8JRC June 2008
Building an Agro-Economic Modelling Platform:Disaggregated Agricultural Social Accounting Matrices for EU27 (AgroSAM)
9JRC June 2008
• Minimisation of Cross Entropy Measure (CE), subject to accounting constraints:
– AgroSAMs have to match EuroStat control totals
– CE allows specifying confidence in data (higher confidence in cereal, oilseed, and dairy data, lower in fodder crops)
• Contribution of EU27 AgroSAMs to GTAP database (2008)
• Future Developments
– Spatial coverage: Regional SAMs (NUTS2)
– Annual coverage: Compilation of SAMs until
• 2005 based on observations• 2010 based on projections
E x p e n d i t u r e s
Activities Commodities Factors
Transactions
Institutions Total
Agriculture
Industry Activities
Services
Domestic production
Agriculture
Industry Commodities
Services
Intermediate demand
Domestic consumption
Exports
Labour Factors
Capital
Payments for fixed factors
Income from abroad
Trade Trade margins Transactions
Taxes Taxes on activities
Taxes on commodities
Direct taxes
Enterprises
Households
Government
Savings-Investment
Savings
R e v e n u e s
Institutions
Rest of the world
Imports
Distribution of factor
income
Transfers
Total
CAPRI Data
EuroStat Data
Unbalanced AgroSAM
1 min lns s ss
CE W W W Cross Entropy Minimand
s.t. 2 Y Y Final AgroSAM entry Y equals prior information
Y times correction factor kappa κ 3
exp s ss
W b
Kappa is defined as exponential function of bounds b and associated weights W
4 3, 1.5,0,1.5,3sb SIG Bounds b are defined as symmetric interval centred at 0; range of the interval depends on exogenously set standard deviation SIG
5 1; 0 1s ss
W W Weights W have to add up to 1 and range between 0 and 1
6 accounting identities for Y Totals of rows and columns in the AgroSAM have to be equal; associated quantities also have to be balanced
Balancing procedure
10JRC June 2008
Axel Tonini (JRC IPTS, Seville), Roel Jongeneel (LEI, The Hague),2008, Modelling dairy supply for Hungary and Poland by generalised maximum entropy using prior information, European Review of Agricultural Economics 35 (2) (2008) pp. 219-246.
Müller, M. and I. Pérez Domínguez (2008): Compilation of Social Accounting Matrices with a Detailed Representation of the Agricultural Sector (AgroSAM). Presented at the 11th Annual Conference on Global Economic Analysis, Helsinki, Finland. Müller, M., I. Pérez Domínguez, and S.H. Gay (2009, forthcoming): Construction of Social Accounting Matrices for EU27 with a Disaggregated Agricultural Sector, IPTS Technical Documentation.
11JRC June 2008
Spatial Data Infrastructures Unit of JRC-IES (Ispra)
Contact: Jean [email protected]
12JRC June 2008
Spatial Data Infrastructures Unit (IES)• The Spatial Data Infrastructures Unit was established in 2006 as the JRC's
response to new policy priorities. One such new priority regards the creation of a European Spatial Data Infrastructure, with a particular focus on the development and implementation of distributed information systems for environmental monitoring through in-situ and Earth observation techniques according to the
INSPIRE Directive adopted in February 2007 • Our mission is to coordinate the scientific and technical development of the
Infrastructure for Spatial Information in Europe (INSPIRE), support its implementation within the Commission and the Member States, evaluate its social and economic impacts, and lead the research effort to develop the next generation
of spatial data infrastructures
13JRC June 2008
Harmonised multi-resolution geographical grid (IES)
Context: One method of storing spatial information is by using geographical gridsEqual area grid suitable for generalising data, statistical mapping, analytical work
INSPIRE Directive complemented by Implementing Rules foresees the definition of a Pan-European Grid based on a commonly agreed reference system (ETRS89-LAEA)Grid defined as hierarchical (power of 10) with associated coding system
JRC’s role as technical coordinator of INSPIRE : - identify user requirements and develop recommendations for grid specifications - testing use case implementations
0,0
Grid 100 km
Grid 10 km
Applications:European Population Grid (JRC-EEA)Multi-scale Soil Information System (Soil Action)Eco-pedological Map for the Alpine Territory (Soil Action)SRTM DEM for Europe, Corine Land Cover, LUCAS, …
15JRC June 2008
LUCAS (Land Use/Cover Area frame Statistical Survey)
Role of JRC on LUCAS 2006: • Optimising the sample • Efficiency assessment
Main task for 2009: • Helping Eurostat to adapt 2006 sample to new
priorities
• 2001-2003
• 2006
Relative efficiency
16JRC June 2008
Population density grid of the EU
• Fine resolution (1ha)• Downloadable from EEA
• LUCAS • Reference data
• Downscaling
Initial data: population per commune
17JRC June 2008
In Europe• 35 countries covered• 11 crops monitored• 33 years of meteo and agrometeo data (daily
data from ~3000 stations)• 20 crop’s indicators are daily simulated by crop
models• 21 years of low resolution satellite information
MARS Crop Yield Forecasting System
7th FWP
Mars FOOD
Mars STAT
18JRC June 2008
Econometrics and Applied Statistics Unit, JRC-IPSC (Ispra)
19JRC June 2008
ECOTRIM (Riccardo)
20JRC June 2008
QUEST III (Riccardo)
21JRC June 2008
22JRC June 2008 Rickety NumbersInternational rankings of higher education
lack statistical robustness Michaela Saisana, Beatrice D'Hombres, Andrea Saltelli
UNE ÉTUDE QUI
MET EN CAUSE LE CLASSEMENT DE
SHANGHAÏ
See www.lemonde.fr Saturday 15 November
2008 http://www.lemonde.fr/archives/article/2008/11/14/vers-un-classement-europeen-des-universites_1118448_0.html
23JRC June 2008
1-5
6-1
0
11-1
5
16-2
0
21-2
5
26-3
0
31-3
5
36-4
0
41-4
5
46-5
0
51-5
5
56-6
0
61-6
5
66-7
0
71-7
5
76-8
0
81-8
5
86-9
0
91-9
5
96-1
00
101-1
05
106-1
10
111-1
15
116-1
20
121-1
25
126-1
30
SJTUrank
Harvard Univ 100 1 USAStanford Univ 89 11 2 USAUniv California - Berkeley 97 3 3 USAUniv Cambridge 90 10 4 UKMassachusetts Inst Tech (MIT) 74 26 5 USACalifornia Inst Tech 27 53 19 1 6 USAColumbia Univ 23 77 7 USAPrinceton Univ 71 9 11 7 1 8 USAUniv Chicago 51 34 13 1 9 USAUniv Oxford 99 1 10 UKYale Univ 47 53 11 USACornell Univ 27 73 12 USA… … …Univ California - San Francisco 14 9 14 3 11 3 7 10 4 3 3 3 6 1 6 1 18 USA… … …Duke Univ 10 6 13 11 6 3 7 6 3 1 3 1 9 9 7 1 3 1 32 USARockefeller Univ 4 10 23 26 1 3 3 3 3 3 4 4 6 3 1 1 1 32 USAUniv Colorado - Boulder 19 39 30 11 1 34 USAUniv British Columbia 20 60 20 35 CanadaUniv California - Santa Barbara 9 9 10 3 10 6 7 6 11 4 6 3 4 7 1 1 36 USAUniv Maryland - Coll Park 6 37 44 9 4 37 USA… … …Ecole Normale Super Paris 7 9 4 6 7 6 4 9 6 7 4 3 3 4 3 3 1 6 4 73 FranceUniv Melbourne 1 20 17 31 23 1 6 73 AustraliaUniv Rochester 1 10 7 16 24 14 10 10 6 1 73 USAUniv Leiden 3 6 9 23 24 13 14 9 76 Netherlands… … …Univ Sheffield 1 21 26 21 9 13 7 1 77 UKTohoku Univ 4 1 7 1 4 17 19 3 3 3 19 7 3 4 4 79 JapanUniv Utah 4 4 6 1 4 9 6 16 7 13 4 9 6 6 1 79 USAKing's Coll London 4 6 9 29 17 14 10 1 6 3 1 81 UKUniv Nottingham 1 6 10 21 21 10 17 7 4 1 82 UKBoston Univ 3 1 6 3 6 11 1 4 3 13 14 10 10 10 83 USA… … …Legend:Frequency lower 15%Frequency between 15 and 30%Frequency between 30 and 50%Frequency greater than 50%
Simulated rank range
24JRC June 2008
Global sensitivity analysis
2000 2004 2008
Courses every year
Venice
Brussels
Ispra
…