Synthesis: March 1, 2008

SYNTHESIS March 1, 2008

What we have discussed • Overview of methods on TFP measurement and explanation • Presentations and discussions: – Commodity studies (livestock and poultry; traditional exports; high value crops; aquaculture; grains) – Cross-cutting themes (R&D, Extension; environment; Human capital; Social capital)

Commodity Livestock and poultry

Measuring TFP Stochastic frontier

Explaining TFP Multivariate regression, enterprise level

Data source Panel data (2000); Recent panel to be surveyed

Traditional export crops

Translog cost

Embedded in translog?

Yield

Yield function equation Embedded in translog?

Panel data (1997); recent panel to be surveyed National (regional?) secondary data Panel data (1994?); recent panel to be surveyed

High value crops

Aquaculture

Grains

Translog (with age of tree stock) Yield

Yield function equation

Decomposition

Qualitative assesment

Yield

Yield function equation Multivariate regression, enterprise level

Malmquist (via DEA)

Available survey data (WorldFish, BFAR) Shrimp recent data to be surveyed (pseudo-panel) National (regional?) secondary data Available survey data (To be examined)

Cross-cutting themes • R&D – Construct measures; institutional evaluation – Case study of R&D impact of TFP in rice

• Environment and Natural Resources – Upland: two stage, switching regression (1st determine adoption; 2nd use in yield regression) – Lowland: collect data, possibly run an impact regression?

Cross-cutting themes • Human capital – Explain farm output (?), by years of schooling (distinguished by level) and other factors – Using Bicol farms panel data

• Social capital – Measure institutions according to intermediate outcomes (property rights, transaction costs) – Use available data from various sources

Common issues • Need to start from behavioral model (e.g. Lanzona) – Define decision-maker’s motivation, constraints, derive actions – Adoption decision (ENR study) – Industry structure (commodities) – Endogenous technology choice (commodities)

Common issues • Definition of TFP – include technical efficiency? How about institutional constraints? • Short panels – Very little technical change – Resort to enterprise level differences (few policy handles)

• Resort to partial productivity measures for policy variables (trad, HVC, aqua)

Future steps • Agriculture sector as a whole • Pilot studies (link to commodity studies?) • Synthesis – how to ensure that the various studies converge – Are TFP results comparable across commodities? – Achieving consistency between agri as a whole, and individual commodity results – How would results from the cross-cutting issues results be incorporated, if at all, in other studies?

• Future scenarios (based on multi-market models)