Impact of New Technologies

1
Economic impact of agricultural biotechnology in
the European Union Transgenic sugar beet and
maize
Dissertationes de Agricultura, No. 713, Jozef
Heuts-auditorium, Landbouwinstituut, Faculteit
Bio-ingenieurswetenschappen, Katholieke
Universiteit Leuven, 1 September 2006, 1600pm
Matty Demont Promoter Prof. E. Tollens Jury
President Prof. G. Volckaert Jury Members Prof.
E. Mathijs Prof. J. Swinnen Prof. J.
Vanderleyden Prof. J. Wesseler
2
IntroductionAgBiotech adoption in the world
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
3
IntroductionAgBiotech adoption in the world

• Most of the recent agbiotech innovations have
  been developed by private sector
• Therefore, the central focus of societal interest
  is not on the ROR of RD, but on distribution of
  gains among stakeholders in the technology
  diffusion chain

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
4
IntroductionAgBiotech adoption in the world
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
5
IntroductionAgBiotech adoption in the world
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
Upstream
Downstream
6
IntroductionAgBiotech adoption in the world

• Upstream private sector is highly consolidated
• Existence of market power and extraction of
  monopoly rents

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
7
IntroductionAgBiotech adoption in the world

• Alston, Norton Pardey (1995) (ANP)
• Moschini Lapan (1997)
• Widely used in agbiotech impact literature

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
8
(No Transcript)
9
(No Transcript)
10
IntroductionAgBiotech adoption in the world

• Farmers capture sizeable gains
• Size and distribution of welfare effects of the
  first generation of GE crops are function of
• 1. Adoption rate
• 2. Crop
• 3. Biotech trait
• 4. Geographical region
• 5. Year
• 6. National policies (Ch.1) and IPR protection
• 7. Assumptions and underlying dataset (Ch.4)

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
11
Upstream Average 37
12
IntroductionAgBiotech adoption in the world

• However, benefit sharing seems to follow a
  general rule of thumb
• 1/3 upstream vs. 2/3 downstream
• This rule of thumb seems to be valid for both
  industrial and developing countries

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
13
IntroductionAgBiotech adoption in the EU

• De facto moratorium on GM crops October 1998
  May 2004 (Syngenta Bt 11 maize)
• 1998-2002 Adoption stagnated at 25,000 ha Bt
  maize in Spain, doubled afterwards
• 2006 5 Bt maize growing EU Member States Spain,
  Portugal, France, Czech Republic, Germany
• De facto moratorium implies a cost to society
  deadweight cost or benefits foregone of GM crops

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
14
IntroductionAgBiotech adoption in the EU

• We need to know this cost in ex post, but also
  for future decisions in ex ante

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
15
IntroductionHypotheses

• The first generation of agbiotech innovations
  could and can significantly contribute to
  productivity and welfare in EU agriculture
• The largest share of total welfare creation is
  captured downstream (farmers, processors,
  manufacturers, distributors and consumers)
• Conventional benefit-cost analysis cannot capture
  uncertainty and potential irreversibility
  regarding environmental effects. It can be
  extended by a real option approach to assess
  maximum tolerable levels of irreversible
  environmental costs that justify a release of
  these innovations in the EU

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
16
IntroductionHypotheses

• 4. Some of the variability of welfare estimates
  reported in literature can be explained by the
  modeling of supply shift in conventional
  equilibrium displacement models

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
17
IntroductionCase studies
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
18
IntroductionCase studies
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
19
IntroductionCase studies
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
20
IntroductionCase studies
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
21
IntroductionHerbicide tolerant (HT) sugar beet
in EU-15
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
22
IntroductionBt Bacillus thuringiensis maize
in Spain
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
23
MethodologyHerbicide tolerant (HT) sugar beet
in EU-15

• Farm level analysis
• - assume standard HT replacement programs
• - compare costs with observed programs
• - assume technology pricing (see data)

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
24
MethodologyHerbicide tolerant (HT) sugar beet
in EU-15

• Aggregation to the global level through standard
  methodologies
• - Alston, Norton Pardey (1995) (ANP)
• - 3 regions EU, ROW beet, ROW cane
• - Dynamic world price behaviour
• - Moschini, Lapan Sobolevsky (2000) (MLS)
• - Former EUs CMO sugar
• - Technology spillovers included
• - Non-spatial no intra-EU trade flows
• - Disaggregated supply 16 prod. blocks
• - Aggregate EU demand

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
25
MethodologyHerbicide tolerant (HT) sugar beet
in EU-15

• Real option approach (Wesseler Weichert, 1999)
  decision to release GM crops in EU is one under
  flexibility, irreversibility, and uncertainty
• Neo-classical decision criterion benefits
  costs
• Include an additional safety factor to take
  into account uncertainty irreversibility
• Decision criterion benefits gt costs by a factor,
  the so-called hurdle rate (estimated through
  historical gross margin series)
• Calculate break-even points maximum tolerable
  costs

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
26
MethodologyBt Bacillus thuringiensis maize in
Spain

• Farm level analysis
• - standard damage abatement function
• - damage stochastic (lognormal)
• - calibrated on real corn borer damage data
• Aggregation to national level
• - Alston, Norton Pardey (1995) (ANP)
• - small, open economy
• - Oehmke Crawford (2002) Qaim (2003) (OCQ)

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
27
Data

• Ex ante (HT sugar beet in the EU-15)
• - No adoption of the new technology
• - No farm level impact data, only field trials
• - Assumptions 1. Yield impact
• 2. Technology pricing
• - Sources expert opinions, literature, economic
  theory, national surveys, Eurostat
• - Stochastic simulation
• Ex post (Bt maize in Spain)
• - Scarce data sources
• - Data mining (e.g. corn borer damage)
• - Sources literature, national surveys
• - Stochastic simulation

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
28
Results
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
29
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
30
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
31
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
32
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
33
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
34
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
35
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
36
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
37
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
38
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
39
ResultsHerbicide tolerant (HT) sugar beet in
the EU-15
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
40
ResultsHerbicide tolerant (HT) sugar beet in
the EU-15
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
41
ResultsBt Bacillus thuringiensis maize in
Spain
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
42
ResultsBt Bacillus thuringiensis maize in
Spain
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
43
ResultsBt Bacillus thuringiensis maize in
Spain
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
44
Model evaluation

• 5 methods of supply shift calculation
• CIR Change in Revenue
• ANP Alston, Norton Pardey (1995)
• ANP1 ANP with supply elasticity 1
• OCQ Oehmke Crawford (2002) Qaim (2003)
• MLS Moschini, Lapan Sobolevsky (2000)

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
45
Model evaluation
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
46
Model evaluation
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
But if ? 0 ? ANP ANP1 OCQ
47
Model evaluation
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
48
Model evaluation

• ANP method seems not robust at first sight

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
49
Model evaluation

• Lets have a look at the differences between the
• 5 methods

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
50
Model evaluation

• No systematic differences between the models when
  fed with stochastic market data, except between
  ANP1 and OCQ

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
51
Model evaluation

• Thus, if we are only interested in the mean
• value, given stochastic market data, model choice
  
• does not matter

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
52
Model evaluation

• In other words
• Data uncertainty gt model uncertainty

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
53
Model evaluation

• Not surprisingly, supply elasticity plays a major
  role in ANP sensitivity

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
54
Model evaluation

• Lets have a look at the variance comparisons

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
55
Model evaluation

• Remember

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
56
Model evaluation

• ANP is significantly less robust than CIR, OCQ
  and MLS, but not ANP1
• ? the ANP vs. ANP1 discussion on supply
  elasticity is irrelevant, given stochastic data

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
57
Model evaluation

• ANP is significantly less robust than CIR, OCQ
  and MLS, but not ANP1
• OCQ and MLS significantly more robust and hence
  preferred methods

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
58
Model evaluation
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
59
Model evaluation
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
60
Model evaluation
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
61
Conclusions

• The first generation of agbiotech innovations
  could and can significantly contribute to
  productivity and welfare in EU agriculture
• The largest share of total welfare creation is
  captured downstream (farmers, processors,
  manufacturers, distributors and consumers)
• Conventional benefit-cost analysis cannot capture
  uncertainty and potential irreversibility
  regarding environmental effects. It can be
  extended by a real option approach to assess
  maximum tolerable levels of irreversible
  environmental costs that justify a release of
  these innovations in the EU

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
62
Conclusions

• Some of the variability of welfare estimates
  reported in literature can be explained by the
  modeling of supply shift in conventional
  equilibrium displacement models
• Recommend simplified and transparent model in
  combination with stochastic data mining
• The real question is whether we want to produce
  information or whether we want to produce a model

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
63
Conclusions

• The crippling flaw in much environmental and
  natural resource economics is that most
  practitioners believe that the models we build
  offer a clear and plausibly reliable mapping
  into propositions about the world of facts they
  presume to depict. All models are wrong, but some
  are more wrong than others
• (Bromley, 2005, p. 29).

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
64
Acknowledgements

• Parents
• Siska De Borger
• Prof. E. Tollens, promotor
• Prof. G. Volckaert, jury president
• Jury Members Prof. E. Mathijs, Prof. J.
  Vanderleyden, Prof. J. Swinnen Prof. J.
  Wesseler
• Josée Verlaenen, Godelieve Vanzavelberg, Odette
  Moria
• Collegues Centre Agr. Food Economics
• VIB, K.U.Leuven, European Commission, Monsanto
• Experts (zie p. ii, iii) co-authors
• Audience

Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements
65
MethodologyHerbicide tolerant (HT) sugar beet
in EU-15
?PSEU b a d c ? 0 ?CSEU 0
66
MethodologyBt Bacillus thuringiensis maize in
Spain
Introduction Methodology Data Results Model
evaluation Conclusions Acknowledgements