1 / 36

Leonardo Becchetti Pierluigi Conzo Giuseppina Gianfreda

Leonardo Becchetti Pierluigi Conzo Giuseppina Gianfreda. Market access, organic farming and productivity: the determinants of creation of economic value on a sample of Fair Trade affiliated Thai farmers. 2009. What is Fair Trade?.

eman
Télécharger la présentation

Leonardo Becchetti Pierluigi Conzo Giuseppina Gianfreda

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Leonardo Becchetti Pierluigi Conzo Giuseppina Gianfreda Market access, organic farming and productivity: the determinants of creation of economic value on a sample of Fair Trade affiliated Thai farmers 2009

  2. What is Fair Trade? Economic phenomenon aimed to promote inclusion of marginalised famers through: • Improved market access • Capacity building • Environmental sustainability • Export services • Price stabilisation • Provision of a premium used for investment or development of local public goods

  3. Controversial issues • FT premium  market distortions ? (LeClair, 2002) • Price premium corrects market distortion because of monopsonistic market power of local intermediaries. • Successful innovation for moral hazard on producer’s investment (Reinstein and Song, 2008). • Access to education, credit and markets. • Does FT promote capacity building and inclusion of farmers in internatioanl markets?

  4. Aim of the analysis • Hypothesis we tested: • Affiliation years enhance economic value • Methodological approaches to control for endogeneity and potential selection bias: • Instrumental variable • Propensity score evaluation • Only-treatment sample analysis • Separate Organic farming and FT effects

  5. Fair Trade in Thailand

  6. The dataset CounterfactualAnalysis Questionnaire 75 questions on various measures of qualitative and quantitative well being.

  7. Descriptive Findings1) socio demographic variables, cooperative membership and affiliation years Treatment and Control samples do not present significant differences in terms of socio-demographic characteristics

  8. Descriptive Findings2) Price and sale conditions Affiliated farmers obtain better conditions than control farmers also when selling to local cooperatives  Bargaining power / organic premium recognised by local market

  9. Descriptive Findings3) Productivity, income, wages and investment • Treatment and control samples are 90% significantly different in terms of productivity (income from agriculture/h worked). • Creation of economic value (per capita income from agriculture) is significantly different.

  10. Descriptive Findings4) Consumption expenditure and self-consumption • Difference in income from agriculture is higher when self consumption (market-valued) is considered

  11. Descriptive Findings4) Consumption expenditure and self-consumption • Visible + Invisible (self produced) FOOD CONSUMPTION (MKT value) • SELF CONSUMPTION adds 27% to total family income in BakReua • SELF CONSUMPTION adds 31% to total family income in Kud Chun • Standard of Living rises from 6.17 to 7.87 $/day in PPP in BakReua • Standard of Living rises from 4.69 to 6.14 $/day in PPP in Kud Chun

  12. Descriptive Findings5) Savings, Debt and wealth Wealth measured by indirect indicators related to housing accomodation (durables owned) Affiliated farmers appear relatively better off in terms of financial conditions

  13. Econometrics FindingsSummary • Objectives: • check whether the difference in the creation of economic value is confirmed when controlling for relevant factors (i.e. education, location, age, sex, n. of children, etc.) • FT vs organic certification effects: which of those prevails? • Selection bias and endogeneity issues  Instrumental variables; propensity score; only-treatment sample analysis • Robustness check: adding “invisible” income from self consumption

  14. 1) ORGANIC CERTIFICATION EFFECT –OLS estimation • Eq.1  marginal effect of 1 year of organic certification = 818 TBT (2% more than average income from agriculture in control group) • Eq. 2  Organic certification result persists when controlling for size of FT premium. • FT premium  more saving/less debt, but Certification Years has effects on income only if invested in capacity building  cannot explain marginal effect of the treatment. • Productivity and commercialization gains widened the income gap across year. • Eq. 3 and 4  Certification Year have a stronger impact in Bak Reuea (higher income and productivity area).

  15. 1) ORGANIC CERTIFICATION EFFECTEndogeneity and 2SLS estimation • INSTRUMENTS: • Affiliated farmer’s distance from the cooperative  cost of bringing product to cooperative  affiliation. Relevant instrument: farmers are “locked” in their area and did not change it after the “treatment”. [eq. 5] • Exogenous memorable events  positive and negative economic consequences • (i.e., increase in market price, shock on production, lottery winning, etc. / relative’s death, disease, car accident, etc. •  idiosyncratic shocks [eq. 6] • DIAGNOSTICS for testing exogeneity: • Wooldrige’s heteroskedasticity-robust score and regression tests : instruments are exogenous if the added variables (residual from a modified specification in which instruments replace selected endogenous regressors) in the standard non instrumented equation are not significant. null of exogeneity is not rejected at 99% when using only DISTANCE. • Sargan test on overidentifying restrictions  can’t reject null hypothesis in the specification with more than one instrument (eq. 6) but null of exogeneity is rejected

  16. 1) ORGANIC CERTIFICATION EFFECTEndogeneity and 2SLS estimation • Certification age is positive but significant only at 10% confidence level •  Better estimates when replacing organic with FT affiliation years

  17. 1) ORGANIC CERTIFICATION EFFECTEndogeneity and Propensity Score Matching Compare treatment and control producers with P.S.M. approach • I specification: • avoid variables with positive impact on income • Include AGE, N. CHILDREN, GENDER AND GEOGR. LOCATION. • II specification: • Add SCHOOL YEARS and JOB EXPERIENCE (not significant in explaining income)

  18. 1) ORGANIC CERTIFICATION EFFECTEndogeneity and Propensity Score Matching Compare treatment and control producers with P.S.M. approach In both cases the difference between TREATMENT and CONTROL sample is significant and strong

  19. 1) ORGANIC CERTIFICATION EFFECTEndogeneity and Only-Treatment sample analysis Estimation of AFFILIATION YEARS effects in the subsample of affiliated producers only. • AFFILIATION YEARS Much weaker effects • Non-significant for Kud Chun area

  20. 2) FAIR TRADE AFFILIATION EFFECT • Previous variables have been rescaled by introducing an upper bound of 6 years for all farmers with organic certification longer than 6 years. • FT AFFILIATION YEARS  significant and stronger in magnitude Re-estimate previous specifications replacingORGANIC CERTIFICATION years with FT AFFILIATION years

  21. 2) FAIR TRADE AFFILIATION EFFECT • FT YEAR effect (OLS): • Eq. 1  …is stronger (1,350 TBT) • Eq. 2  …moves to 1,458 when introducing FT premium. • Eq. 3 and Eq. 4  …remains significant when calculated separately in the two areas.

  22. 2) FAIR TRADE AFFILIATION EFFECT Endogeneity and 2SLS estimation • FT YEAR effect (2SLS): • Eq. 5 and Eq. 6  …remains significant(also when calculated separately in the two areas). • Better Exogeneity tests  Single instrument Equation does not reject null of exogeneity at more than 5% confidence level.

  23. 2) FAIR TRADE AFFILIATION EFFECT Endogeneity and Only-Treatment sample analysis • FT YEAR effect (on Treatment): • Eq. 1  …gets stronger • Eq. 2  …remains significant after correcting for the 2008 FT PREMIUM

  24. 3) FAIR TRADE AFFILIATION VS ORGANIC CERTIFICATION Which effects prevail? • FT and organic certification years are highly correlated (.92) • Test whether one prevails on the other: • Estimate the the BASE and RESTRICTED model with both variables • Use DAVIDSON-MCKINNON (1993) TEST FT affiliation effect is stronger

  25. 3) FAIR TRADE VS ORGANIC CERTIFICATION Which effects prevail? – Davidson-McKinnon test • Eq. 2 Predicted Var. = FT AFFILIATION YEARS • Eq. 3 Predicted Var. = ORGANIC CERTIFICATION YEARS The predicted dependent variable from the FT affiliation estimate is 5% significant in the organic certification estimate while it is not for the opposite case

  26. 4) Robustness check • We repeated all the previous estimates by adding the market value of agricultural products produced and consumed in the household (INCOME FROM AGRICULTURE + SELF-CONSUMPTION). • Results are similar • FT affiliation effect is confirmed under the different specifications and methodological approaches

  27. Dep. Variables: control group, area 1, age, n. of children, school years, male, married, divorced, years in agriculture, temp. employees, land size Dep. Variables: control group, area 2, age, n. of children, school years, male, married, divorced, years in agriculture, temp. employees, land size 4) Robustness check Greater impact of FT and OC Better specification is with FT aff. years Better results from exogeneity test: no reject the null of exogeneity at 10%

  28. Interpretation of our findings • FT AFFILIATION affects creation of economic value more than ORGANIC CERTIFICATION YEARS. Why? • Double bounus FT: price premium to farmers and premium to organization (to be invested for innovation and local public goods). • Marketing gains generated by FT. Consider that affiliated producers sell more.

  29. Interpretation of our findings • FT AFFILIATION affects creation of economic value but notPRODUCTIVITY PER WORKED HOURS. Why? • Higher worked h: affiliated farmers on average work 20 days/year more in agriculture. • h. worked increase with affiliation years: farmers below the median affiliation years work on average 1,461h/year against 1,723h/year for those above it.

  30. Interpretation of our findings …in other words: • ORGANIC FARMING increasing labour intensity activity  taken alone, not helpful in improving productivity • FT AFFILIATION YEARS decisive to improve productivity. Why? • Improved market access(alternative tradechannelprovision) • “social premium” tobeinvested in capacity building and farmer’s welfare

  31. Conclusions • Additional FT affiliation years  positive and signficant effect on income from agriculture • This effect does not translate into higher productivity since affiliated workers work progressively more hours. • Only when considering FT affiliation years, results remains robust after controlling for endogeneity and selection bias.

  32. Conclusions • Affiliated farmers sell more and enjoy higher self-consumption share • Improved marked access  Affiliation effect • Observed income from agriculture and productivity effect = downward biased ConcurringFT affiliationisprobablycrucial in determining a nonnegativeproductivityand per capita incomedifferencebetweenorganic and conventionalfarmers

  33. The link between foreign market access and access to education …work in progress • Restrospective panel data  evaluate the effect of FT on schooling decisions across the past 20 years. • The probability of school enrolment in families with > 2 children is significantly affected by affiliation years

  34. The link between foreign market access and access to education …work in progress • PROBABILITY OF GOING TO SCHOOL  positively correlated with BIRTH ORDER. • From 84% for the 1st falls to 71% for the 5th and 53% for the 6th. • Fair trade affiliation seems to matter for children who come after the second.

  35. The link between foreign market access and access to education …work in progress • PROBABILITY OF GOING TO SCHOOL (irrespective of the age order) in smaller and larger families • = 88% in single child families; = 51% in families with 6 children. • The difference between affiliated and non affiliated families becomes more significant as far as the number of children grows.

  36. Thankyou “Amici di Alessandro Bolondi” association is fully awknowledged

More Related