Transfers out of poverty: research insights, policy progress, and positive outcomes

1
Transfers out of poverty research insights,
policy progress, and positive outcomes

• Jere R. Behrman
• 20 February 2007
• PORIA Case Studies
• Methods Planning Workshop

2
Background

• Central objective Assuring Household Access to
  Food From Subsidies toward Safety Nets and
  Other Modern Instruments
• IFPRIs early work on social protection paid
  particular attention to considering ways of
  reforming food subsidies

3

• Over time, in part because of influence of
  IFPRIs research, attention shifted to innovative
  approaches for social protection. Two strands
• understanding famines improving mechanisms to
  mitigate droughts floods.
• Evaluating effectiveness of programs such as
  conditional cash transfers that sought to ensure
  minimum levels of food consumption for poor while
  also investing in human capital of subsequent
  generations.
• (Also, e.g., gender, household allocations,
  insurance, employment programs)

4
Output from the POR

• Books and Papers
• Workshops and Conferences
• Meetings, Presentations, Briefings
• Citations in ISI, Google Scholar

5
Proposed impact assessment approach and
procedures

• Primarily focus on the demand for this type of
  information of governments and NGOs.
• Survey of policy makers in nations that have
  modernized their safety-net to learn how they
  acquired their guidance, sources of information
  (including IFPRI), methods of analysis used in
  interpreting the experience, etc., and then
  documenting the proportion of budgets allocated
  to such activities, and the oversight mechanisms
  for ME.

6

• 2. Identify and, where possible, quantify
  influence by meta-analysis of evaluation
  studies,
• 3. identify and quantify impacts and, where
  possible, a reasonable estimate of the dollar
  value of impacts and costs of the POR

7
Challenges 1.Survey Policy Makers

• How identify respondents, gain access? (Key
  people may be very few, e.g., Mexico Levy,
  Gomez-Leon, Zedillo, Fox)
• How identify IFPRI impacts?
• More than some case studies? Resource costs?

8
2. Identify Influences

• Distinguish influence from impact
• Question of influence of what presumably IFPRI
  input (and not, say, PROGRESA).

9
3. Impacts and Costs

• How measure impact (Not, e.g., some IR 8 variety
  that can be traced) and how value?
• How identify IFPRI contribution from that of
  national (and other) partners?
• How get resource costs for relevant venture?

10
How Might Proceed?

• Case Study of IFPRI/PROGRESA (cherry not
  limon).
• Clearly international public good aspects
• Data, studies, inferences for other countries
• Direct references in interviews with national
  policy makers and international policy advisors
• Document as possible channels of possible
  influence (citations in policy-oriented
  organizations such as World Bank and IDB and in
  academic venues, interviews)

11

• Great challenge identifying IFPRI influence or
  impact from that of PROGRESA itself. Possibly
  desirable to undertake jointly with PROGRESA
  (Oportunidades).
• Very limited resources.
• Risk that seen as cherry already picked (e.g.,
  Shoufias, Levy).
• Other options for this case?
• Other cases? Limon?

12
Impact of the In-Trust Agreements betweenFAO
the CGIAR centres on germplasm availabilityand
value

• Elisabetta Gotor
• Institutional Learning and Change Unit
• Bioversity International
• 20-21 February 2007

13
Presentation Outline

• Historical Background
• Approach to the Study
• Procedure of the Study
• Counterfactual
• Accountability
• Methodology
• Alternatives

14
Convention on Biological Diversity1992-Resolutio
n 3 of the Nairobi final Act

• Issue of Sovereign Right of Nations over Their
  Genetic Resources

Is a Germplasm a private properties or a public
good ? WTO-TRIPS Art. 27.3(b) Members shall
provide for the protection of plant varieties
either by patents or by an effective sui generis
system or by any combination thereof. The
provisions of this subparagraph shall be reviewed
four years after the date of entry into force of
the WTO Agreement.
15
Keystone International Dialogue

• Crucible Group

In-Trust Agreements - 1994
CGIAR
FAO
16
SINGER System-wise Information Network for
Genetic resources
Genetic Resources Policy Committee of the CGIAR
SGRP System wide Genetic Resources Programme
17
Approach to the Study

• Demand led approach

Starting Point
INPUTS
In Trust Agreements
OUTCOMES of the policy
OUTCOMES
18
Procedure of the study
Identify Bioversity International role played
for the achievement of the In-Trust Agreement?
Main Objective
First Part

• How Ideas and positions evolve throughout the
  policy
• making process?
• What perspectives do key parties in the policy
  making
• process have on how the policy change was
  achieved?

19
Procedure of the study
Assess the impacts of the In Trust-Agreements
on conservation and access to germplasm
Main Objective
Second Part
Counterfactual What Would have occurred without
the In trust Agreement?
Without a regulation would donor countries
have continued to donate germplasm?
Without a regulation how host countries would
have considered the acquired germplasm?
Would the 2001 Treaty have came through?
20
Procedure of the study
Estimate the value of the additional germplasm
conserved because of the policy
Main Objective
Third Part
Accountability What is the empirical evidence of
the In-Trust Agreement??
Assess the Actual Flows of germplasm
SINGER System-wise Information Network for
Genetic Resources
Estimate the additional amount of germplasm made
available or maintained by CGIAR genebanks
21
Methodology to assess phase 2 and 3

• Two-step procedure
• Assess the impacts of the In Trust-Agreements on
  conservation and access to germplasm
• Hierarchical correlated count data model
• 2. Valuing its incremental impact on the overall
  value of plant genetic resources because of
  the policy in place
• Conventional methodology Choices

22
Step 1 Assess the impacts of the In
Trust-Agreements on conservation and access to
germplasm
Correlated Count Data Chib, S. and R.
WinkelmanMarkov Chain Monte Carlo Analysis
of Correlated Count Data. Journal of Business
and Economic Statistics19 (2001) 428-435
Panel count data Chib, S., E. Greenberg and
R. WinkelmanPosterior Simulation and Bayes
Factors in Panel Count Data Models Journal
of Econometrics86 (1998) 33-54
Hierarchical Model Lindley, D. V. M. and A. F.
M. SmithBayes Estimates for the Normal Linear
Model Journal of The Royal Statistical Society
Series B Methodological34 (1972) 1-41 Chib, S.
and E. GreenbergHierarchical Bayes Modeling
inPalgrave Dictionary in EconomicsSecond
Edition
23
Fundamamental Observational Equation

• yijk ?(xijk, zijk) uijk

Where i denotes a site j denotes a species k
denotes a time period y denotes accessions (or
distributions) thus Yijk denotes the number of
accessions (distributions) occurring at site i
for species j at time k
24
Explainingyijk ?(xijk, zijk) uijk

• xijk denotes treatments at site i for species
  j at time k
• zijk denotes controls at site i for species j
  at time k
• uijk denotes errors at site i for species j
  at time k

25
Explainingyijk ?(xijk, zijk) uijk

• What do I mean by Treatments?
• Anything that caused accessions or distributions
  to change in the time series at site i for
  species j in a fundamental way
• The main treatment of interest, the over-arching
  concern and the hypothesized driver of difference
  pre-and post enactment is the In-Trust
  agreement
• Ad hoc at present, but a binary treatment ( 0
  before enactment, 1 post enactment) may suffice

26
Explainingyijk ?(xijk, zijk) uijk

• What do I mean by Controls?
• Anything that neither caused accessions (or
  distributions) to change in the time series at
  site i for species j in a fundamental way but
  may give rise to differences across sites and
  across species.
• Problematic If accessions are supply-driven we
  need an in-depth analysis of research resources
  devotes to species at sites and at times
  complicated
• Problematic If distributions are demand-driven
  we need in-depth analysis if the
  demand-structures at sites across species and
  across time complicated

27
Explainingyijk ?(xijk, zijk) uijk

• What do I mean by Errors?
• Anything that is unexplained by treatments or
  controls across sites, species and time periods.
• Correlations?
• Problematic Complicated by the site-species-time
  hierarchy
• Problematic Complicated by the fact that yijk
  denotes a count
• Correlations across time (same species and site)
• Correlations across sites (same species and
  time)
• Correlations across species (same site and time)
• Correlations in counts?
• Complicated

28
Hierarchical Modelling A General Structure
?
?
?
29
A General Structure
Overarching
?
?
?
Units
Sub-Units
Micro-Units
30
Applied to Accessions and Distributions Impact
of the In-Trust Agreement
?
?
?
Across the sample
Across the sites
Across species
Across time periods
31
Fundamental

• ?i is a random draw from ?(.?)
• ?ij is a random draw from ?(.?i)
• ?ijk is a random draw from ?(.?ij)
• where ?(.?) denotes a probability density
  function conditioned by parameters ?

32
Bayesians Like to Hierarchically Model

• How data are generated
• Added flexibility
• Versatility
• Choices available

33
Fundamental

• The hierarchical ordering
• Hypothesized
• sites ?(.population)
• species ?(.sites)
• periods ?(.species)

Alternative Two periods ?(.population) sites
?(.periods) species ?(.sites)
Alternative One species ?(.population) sites
?(.species) periods ?(.species)
34
Many alternative formulations

• Which model is best and how to measure best ?
• Gelfand, A. and A. F. M. Smith.
• Sampling-Based Approaches to Calculating
  Marginal Densities. Journal of the American
  Statistical Association 85(1990)3972-985
• Chib, S.
• Marginal likelihood from the Gibbs
  Output.Journal of the American Statistical
  Association 90(1995)41313-21

35
Posterior predictive distribution from the
correlated panel count model p(yy) The p for
an unobserved Sample y given the observed
sample y Embodies the sample that would have
existed had the agreement not been enacted
Posterior predictive distribution from the
correlated panel count model p(yy)
Posterior distribution from the correlated
panel count model p(?y)
36
Second Step

• Use information in p(yy) to estimate value of
  plant genetic resources because of the policy in
  place
• Conventional methodology Choices

37
Alternatives

• Recent additions to the literature
• Standard total-factor productivity study
• Search studies
• Evenson, R. E. and Y. KislevA Stochastic Model
  of Applied Research Journal of Political
  Economy84 (1976) 265-281
• Koo, B., P. G. Pardey and B. D. Wright et
  al.Saving Seeds The Economics of Conserving
  Crop Genetic Resources Ex Situ in The Future
  Harvest Centers of the CGIARCABI Publishing
• Gollin, D., M. Smale and B. SkovmandSearching
  an Ex Situ Collection of Genetic
  Resources.American Journal of Agricultural
  Economics82 (2000) 812-827(2004)

38
Case 3 WorldFish
39
Impact Assessment of Community Based Fisheries
Management(CBFM) in Bangladesh
Diemuth E. Pemsl
Presentation for SPIA POR IA Workshop20 21
February 2007, IFPRI Washington
40
Project Background - Bangladesh

• Rich inland fishery resources (42 of total fish
  production)
• Fish most important animal protein
• 80 of rural households catch fish for
  subsistence
• Siltation and encroachment (resource degradation)
• Broadly two types of waterbodies
• - haor, baor, lakes and beels (revenue
  oriented) ? Short-term lease, no incentive to
  conserve, over extraction
• - floodplains and rivers (open access, no
  management) ? Commons dilemma persists

41

• Two important elements of Community Based
  Fisheries Management (CBFM)
• Institutions
• a) Participatory, involves all
  stakeholders
• b) Linkages with local level government
• Fisheries Management
• a) Habitat restoration
• b) Conservation measures

42
Description of the POR research
Research Phase
Dissemination Phase
43
Outcomes of CBFM POR

• Developed and tested CBFM approaches (including 2
  elements)
• Increased awareness of wider public and
  government bodies of need for improved
  management of inland fisheries
• General understanding of the CBFM concept by
  decision makers e.g. DoF officers in charge
• Informing decision makers for upcoming policy
  changes
• Foundation of 130 community based organizations
  (CBO) and provided training/assistance in CBFM
  and legal support

44
Outcomes of CBFM POR (continued)

• Policy changes for 117 project water bodies
  (transfer of use rights, reduction of lease fees
  and abolition of taxes)
• Conservation practices are being followed in
  project water bodies, including establishment of
  186 fish sanctuaries
• Fish production in project water bodies increased
  on average by 20 with no change in the control
  sites over same period
• Income from fish is up 21 for project
  participants compared to 15 for fishers in
  control sites
• Fish abundance in project water bodies increased
  at 17 per year with increasing trends in
  biodiversity whereas control sites experienced
  significant declines in fish abundance

45
Methodological approaches for POR IA
Supply-led
Demand-led

• Inputs
• Outputs
• Dissemination
• and Uptake
• Influence
• Impact

Trace back to project outputs
Identify outputs first
Attribute impact
Start from specific recent policy change
46
Impact pathway for CBFM research
Graph forthcoming
47
CBFM impact categories
Direct impacts

• Productivity and sustainability of inland
  fisheries
• International public good (IPG) community based
  management systems and habitat restoration
  technology
• Awareness of general public and policy makers
• Contribution to policy changes (and impact of
  those)

Indirect impacts

• Subsequent changes in fish price?
• Others?

48
Proposed IA methodology
How to assess/quantify?
Costs ? project docs, partner
estimates Outputs ? inventory of publications
by topic, others e.g. of people
trained Dissemination ? stakeholder surveys
awareness Uptake/Influence ? stakeholder
surveys adoption Impact ? sum up
assigning value counterfactual,
sustainability elicit share of WorldFish
49
Indicators for impact of CBFM

• Farm-level
• Net income, consumption/nutrition changes?
• Inter and intra-household allocation of benefits
• Water body-level
• sustainability of fisheries
• total yield/capacity
• biodiversity
• CBOs
• social capital?
• Policy changes
• Role of CBFM work to influence policy
  changes(based on feedback from decision makers,
  partner NGOs, donors, other players working in
  the field

50
Methodological challenges for IA

• Counterfactual
• Control sites (with NGO activities beyond CBFM)
• Speeding up the process? (Who are) other players?

• Attribution
• (Who are) other players? WorldFish share on
  influence?

• Sustainability
• Trial phase versus sustainable change/adoption?
• How sustainable are changes once project stops

51
Methodological challenges for IA(continued)

• Evaluation and assessment of benefits
• Assessment/measurement of non-market
  benefitsAdding up different types of non-market
  benefits
• Identify and measure direct and indirect effects
  (e.g. increase in the fish catch vs changes in
  fish prices and thus income changes?)

• Equity implications
• Value of benefits arising to different
  stakeholders (different producers, consumers,
  government)? Question of benefits for poor
  versus non-(or less) poor

52
Next steps for POR IA of CBFM

• What data are available/have been collected for
  the CBFM project that can be used for POR IA?
• What information is missing?
• Identify data that have to be collected
• Decide on methodologies
• Funds available from CBFM project?
• Allocate tasks for IA study
• Literature review of previous work (especially
  with regard to policy changeand policy processes
  in Bangladesh)

53
Work plan and time line

• Final CBFM project workshop in Dhaka 6 7 March
  2007
• Field visit to project sites
• Planning session and briefing with project
  leader/staff
• Develop detailed work plan and time line for POR
  IA
• Submit detailed work plan to SPIA by 15 March
  2007

54
Thank you
55
Policy Change in Dairy Marketing in Kenya and
East Africa Economic Impact and Pathways to
Influence from Research
PORIA Methods and Planning Workshop 20-21
February 2007, IFPRI HQ, Washington, DC
56
Kenya Dairy Policy (written policy)

• Main Issues
• Outdated document, 1958 Kenya Dairy Act
• Protects interests of large scale settler dairy
  producers
• Criminalizes activities of small-scale production
  and marketing
• Ignores contribution of smallholder producers and
  traders to economic development

57
Activities of SDP (1997-2005)

• Low cost and appropriate technologies
• Training programs on safe milk handling
• Improve standards of milk processing in informal
  standards
• Provision of incentives for improved milk
  handling
• Supportive milk dealer certification system

58
Milk Marketing Chain (Omore et al, 1999)
59
Policy Change

• Behaviour changes among SSMVs, KDB agents, and
  other milk handlers
• Reduced transactions costs
• Bribes paid
• Milk lost/poured
• Milk cans confiscated
• Etc

60
Impact Pathways
61
Study Activities

• Qualitative study of impact pathways - activities
  and processes leading to policy change in Kenya
• Ex-post economic impact assessment using a
  multi-market model

62
Methods Qualitative Assessment(Research and
Policy in Development (RAPID) Outcome Assessment
Framework)

• Episode studies of specific policy changes
• - Tracking back from policy changes to identify
  key actors, events and influences, and assessing
  their relative importance
• Case study analysis of specific research projects
  
• Tracking forward from specific research and
  related activities to assess their impact
• Outcome mapping approaches
• Identifying changes in behaviour of key actors
  and analysing what influenced these changes

63
Methods Economic Impact Assessment

• Model Attributes
• 2 output markets (milk, maize)
• 4 input markets (fertilizer, labour, napier,
  concentrates)
• 1 proxy for non-food expenditures (housing
  sector)
• 3 Production systems (subsistence, major dairy
  enterprises, limited dairy enterprises)
• 4 household consumption groups (urban rich, urban
  poor, rural rich, rural poor)

64
Methods Economic Impact Assessment

• Data Requirement
• Food production by consumption group
• Food imports and exports
• Input use by production system
• Expenditure on consumption goods
• Value of per capita consumption
• Commodity prices
• Consumption data
• Household income
• Elasticities
• (output ss elasticities wrt input and output
  prices, input dd elaticities wrt input and output
  prices, income elasticities, etc)

65
Methods Economic Impact Assessment

• Specific policy evaluated ex-post
• Reduction in transactions costs resulting from
  changes in the dairy policy
• Policy variables monitored
• Consumer and producer prices
• Supply of commodities by production system
• Consumption levels by households
• Household income

66
Potential Regional Impacts (Regional economic
impact assessment not promised to SPIA)

• ECAPAPA facilitation of learning of cross-border
  lessons in the dairy sectors
• Rationalization and harmonization of policies,
  regulations and standards in the dairy industry
• (Kenya, Uganda, Tanzania, Rwanda, Burundi)

67
Assessing the Impact of CIFORs Influence on
Policy and Practice in the Pulp and Paper Sector

• David A. Raitzer

68
Overview of the Research

• Study of the political economy of the
  Indonesian pulp and paper sector
• Basic problem massive investments in pulp
  mills, but plantations not being developed as
  fiber source
• Reliance on conversion of natural forest for pulp
• Encouraged by subsidies, poor institutions and
  information asymmetries.

69
Identified causes of pulp overcapacity

• Subsidies
• Cheap wood (gt2 cubic meter in royalties)
• Reforestation Fund allocations
• Soft loans from State banks
• Poor regulations
• Related-party loans from conglomerate owned banks
• Use of conglomerates and shell companies
• Financial mark-up schemes
• Poor investor due diligence
• Little involvement of forestry experts
• Rarely use independent audits of forestry
  operations
• Export credit guarantees reduce risk to financial
  institutions

70
Context

• 15 billion of debt accrued by Asia Pulp and
  Paper (APP) and Asia Pacific Resources
  International Ltd (APRIL)
• Study released on Bloomberg newswire in December
  2000 as APP tried to renegotiate its debt
• APP defaulted on 13.5 billion in March 2001
• 1.3 billion of Indonesian non performing loans
  assumed by the Indonesian Bank Restructuring
  Agency
• Creditors made conditional agreements with APP on
  other debts
• Profits on Paper spurred high profile campaigns
  by NGOs.

71
Impact pathways

• NGOs
• Profits on Paper spurred high profile
  campaigns by WWF, Friends of the Earth, the
  Indonesian Environmental Forum and others.
• Increased attention to effects of new mills
  worldwide
• Increased pressure for responsible investing
  worldwide
• Pressure for environmental provisions in debt
  restructuring agreements for APP and APRIL
• Continued reporting on operations

72
Impact pathways

• Creditors
• Requested profits on paper directly
• Use in lawsuits
• Inserted environmental requirements (audits, etc)
  in Master Restructuring Agreement for APP
• Increased attention to environmental safeguards
• Increased attention to risk
• Other investors
• Increased attention to environmentally and
  socially responsible investing
• Increased attention to risk
• Potential motivation for formation of Equator
  Principles

73
Impact pathways

• APRIL and APRIL Improvements since study
• Environmental audits/reporting
• Tesso Nilo National Park (APRIL, 200k ha)
• Conservation set asides (400k ha)
• Committed to legal wood sourcing
• Certification (APRIL concession)
• Community programmes (APRIL)
• Proposed UFS pulp mill
• NGO lobbying led to non issuance of MIGA
  political risk guarantees

74
Improvements in Indonesian forestry regulations

• NGO campaigns led to boycotts of Indonesian pulp
• GoI aims to regain consumer confidence
• Ministerial Decree 101/Menhut-II/2004 on the
  'acceleration of plantation development and pulp
  and paper industry raw material supply
• Now repeatedly presented by Ministry as a
  requirement that all pulp must be harvested from
  plantations by 2009 even though this is not what
  the Decree states

75
Global improvements in due diligence and
responsible investment practices

• NGO campaigns related to the APP/APRIL experience
  may have influenced
• the Equator Principles, which most of the largest
  international banks now follow
• policies and commitments concerning pulp-paper
  and/or forestry investments by large financial
  institutions
• FLEGT action plan commitment to improve due
  diligence for forestry investments by EC
  financial institutions

76
CIFOR pulp sector research
NGOs
Creditors
Other investors
World Bank
Gov Agencies
APRIL
APP
Equator Principles
FLEG initiative
Forestry Ministry
UFS
No mill
Fiber Sourcing Decree
Less corruption
More CSR
Changes in SFM
77
Assessment methods

• Supply and demand led
• Supply led - document direct influence on
  intermediate processes that are believed to drive
  outcomes, so as to identify their counterfactual
  course
• Demand led - identify impact of intermediary
  processes (that were influenced).
• Put the two together to identify impact of
  influence.

78
Assessment methods

• Interviews with representatives of NGOs and
  bilateral agencies that were actively involved
• Changes to attention to the Indonesian pulp
  sector
• Changes to pulp fibre supply shortfalls
• Involvement in lobbying that led to the Equator
  Principles
• Origins of the FLEGT action plan and bilateral
  UK-Indonesia MOU
• CIFORs influence on campaigns and plausible
  counterfactuals
• Triangulation
• Changes in sustainability of APRIL and APP
  practices
• Perceptions of Ministerial Decree
  101/Menhut-II/2004

79

• Interviews with APP and APRIL creditor
  representatives
• Changes in awareness of fiber sourcing
• Influences for sustainability requirements of
  debt restructuring agreements
• CIFORs influence on subsequent changes and
  plausible counterfactuals
• Interviews with APRIL and APP managers
• Changes made to operations
• Role of creditor requirements and external
  assessments mandated by creditors
• Role of NGOs
• Perceptions of counterfactual without creditor
  requirements scenarios

80

• Interviews with MIGA and World Bank Indonesia
  staff
• Considerations regarding proposed United Fiber
  Systems mill in Southern Kalimantan
• Fiber supply appraisal
• Factors leading to political risk guarantee
  decision
• Role of NGOs
• Interviews with officials from Indonesian
  Ministry of Forestry and Forest Council
• Realized and anticipated effects of Ministerial
  Decree 101/Menhut-II/2004
• Implementation and enforcement mechanisms
• Influences that led to Decree
• CIFORs influence on subsequent policy changes
  and plausible counterfactuals

81

• Interviews with headquarters International
  Finance Corporation Staff
• Current pulp and paper involvement
• Procedures for appraisal of pulp and paper
  financing
• Changes in appraisal procedures and investment
  patterns
• Influences on changes in investment procedures
  and patterns
• Role of CIFORs findings and plausible
  counterfactuals
• Telephone interviews with officials of banks that
  adopt deforestation policies
• Changes in investment patterns due to policy
• Specific examples of averted investments
• Influences on deforestation policy
• CIFOR attributable influence and plausible
  counterfactuals

82
Impact supplement

• Indonesian forester - interviews with other pulp
  producers and provincial government officials
• Expectations about Ministerial Decree
  101/Menhut-II/2004
• Plantation development rates to date
• Anticipated plantation development
• Forest clearing rates over time
• Changes attributable to Ministerial Decree
• Depending on findings, use counterfactual for
  Ministerial Decree to assess impact attributable
  to CIFOR

83
Methodological challenges

• Policy reform in the context of corruption and
  deception
• Differing presentations of Ministerial Decree
• Companies with histories of falsification
• Unstable context, reforms could be retracted
• Controversial research on a high stakes subject
  neutrality difficult
• Non-market benefits, valuation difficulties

84
Methodological issues for discussion

• How to handle agencies for which influence
  imposed costs?
• APRIL
• APP
• MIGA
• Degree of confidentiality in interviews
• How to query about confidential decision
  processes (e.g. WB Board deliberations)
• Presentation of key informant responses

85

• Economic Assessment of a Change in Chemical
  Registration Policy due to IRRI Research on
  Pesticide Use and Philippine Farmer Health

20-21 February 2007 IFPRI HQ Washington DC
86
Aim

• The aim of this study is to measure the economic
  benefits attributable to a change in pesticide
  registration policy and assess what proportion of
  the benefits policy change can be attributed to
  IRRI research on pesticide use and Philippine
  farmer health.

87
Objectives

• Understand
• pest ecology and the relationship between pest
  infestation and yield loss
• farmer KAP in pest management
• pesticide market
• researchers, chemical companies and policy
  makers perceptions of pest-related yield losses
  and control methods.
• Determine the degree of awareness of the research
  among FPA and DoA.
• Determine the degree to which this research
  influenced the banning of category 1 chemicals.
• Develop a plausible counterfactual scenario of
  policy evolution in absence of the IRRI
  pesticide-health research.
• Measure the benefits under the counterfactual and
  actual scenarios to quantify the magnitude and
  distribution of these benefits.

88
Research-gtpolicy change-gtimpact
Impact of pesticides on human health 1987
Other research C of M (1985) Dr Rola (1987) IRDC
Task Force to write new pesticide policies
Fertilizer and Pesticide Authority
IRDC Institutional Change
IPM research and results
Civil society / media
Department of Agriculture
Meetings Italy, Germany and Switzerland
Pesticide suppliers
Bureau of Agricultural Extension
Farmers
89
Theoretical framework

• Expected utility function, EU(p, r),
  incorporating health costs due to pesticide
  exposure is specified as
• EU(p, r) EU(py wx HC)
• where
• p profit
• r risk
• p output price
• y output, a random variable whose distribution
  is conditionally defined on the input vector x
• w vector of input prices
• x input vector
• HC health cost as a result of illness due to
  pesticide exposure

90
Health Cost Model

• Health costs are associated with total pesticide
  use pesticide exposure pesticide hazard
  category and other farmer characterisics.
• ln HC f(LAGE, WTHT, DS, DD, LTD, LDI, LDH)
• where
• Ln HC log of health costs
• LAGE log of famers age
• WTHT farmer weight / height
• DS dummy for smoking
• DD dummy for alcohol
• LTD log of total dose of pesticides
• LDI log of dose of insecticides
• LDH log of dose of herbicides

91
Counterfactual

• EU(p, r) wo90 - EU(p, r) w90
• and/or
• EU(p, r) wo07 - EU(p, r) w07
• need to determine
• Toxicity relationship between category 1-2 and
  3-4 pesticides
• Price relationship between category 1-2 and 3-4
  pesticides
• Efficacy relationship between category 1-2 and
  3-4 pesticides

92
Economic Surplus

• Expected change in production costs due to fall
  in health costs k shift
• Number of farmers affected adoption level
• Change in producer and consumer surplus total
  benefits
• Cost of projects plus other costs(?)
• Returns to POR benefits/costs

93
Research-gtpolicy change-gtimpact attribution
IRRI Research
Other research Other information
Policy task force Partnerships
FPA Institutional arrangements
IRDC Institutional research
IPM Other information
Civil society / media Coercion
DoA Political change
Meetings Italy, Germany and Switzerland
Pesticide suppliers Competing interest
BAE Competing interest
Farmers Competing interest
94
Data sources

• Project documents and outputs
• Literature review / databases
• other PORIA studies
• other relevant pesticide use/health studies
• pest ecology and the relationship between pest
  infestation and yield loss (actual)
• pesticide market
• number of farmers affected by policy change.
• Interviews with Dr Pingali, Dr Palis, Dr Rola, Mr
  Max Obuson, Frank Conejo, FPA, and chemical
  companies.
• researchers, chemical companies and policy
  makers perceptions of pest-related yield losses
  and control methods
• determine the degree to which this research
  influenced the banning of category 1 chemicals.
• Household surveys in the three locations
• input/output data
• farmer KAP in pest management
• farmer perceptions of changes in pest ecology and
  optimal pest management practices
• farmer information sources on best pest
  management practices

95
Timeline

• February initial meetings and questionnaire
  design
• March Literature review / databases.
• April June encoding, data analysis, meetings
  with key players/policy makers
• July August initial finings, draft report
• September November follow-up meetings, model
  refinement
• December Final report