Table 1 shows that out of the 761 listed by the
GIAR as having been published from 1995-2008, only 83 listed impact focusing on welfare indicators such as income ornutrition/health status.
able 1: CGIAR Impact Assessment Studies
Impact evaluations focusingon income as an outcomevariableImpact evaluations focusingon (income ornutrition/health) as anoutcome variableAll ImpactEvaluations2008 0 02007 1 22006 4 42005 0 02004 4 52003 5 6Total 1995-2008 67 83 761
As of August 2009http://impact.cgiar.org/ Neither the Poverty Action Lab (J-PAL) nor the International Initiative on Impact Evaluation(3ie) have undertaken or commissioned many agricultural project impact studies. As of mid2009 the project database search at the Poverty Action Lab website shows 25 healthevaluations, 38 in education, only 5 in agriculture (and these are all in Kenya)
. And only 2of 18 funded applications in round 1 of 3ie (the International Institute for Impact Evaluation)funding were awarded to agriculture projects (irrigation, low cost farm equipment)compared to 6 in health. Presumably this reflected some combination of low submissions(perhaps due to the size of funding chunks available) and lack of quality of submissions
.Second, the aggregate data on the impacts between agricultural growth and income ornutrition are inconclusive.
ross-country econometric work (Ligon and Sadoulet, 2008)reported in the 2008 World Development Report shows that a 1% gain in GDP originating inagriculture generates a 6 % increase in overall income for the poorest 10% of thepopulation. This compares with a 4% increase in overall income for the next poorest, and 3%for the subsequent decile. In stark contrast, GDP growth originating in non-agriculturesectors generates zero growth for the poorest 10% of the population, a 1% increase inincome for the next 10% and a 2% increase thereafter. A more recent empirical study by
hristiansen et al. (2010) comes to similar conclusions. Using cross country econometricevidence they report Irrespective of the setting, a one percent increase in agricultural percapita GDP was found to reduce the total $1-day poverty gap squared by at least 5 timesmore than a one percent increase in GDP per capita outside agriculturep 30. For a large setof countries within a cross-country regression framework, Loayza and Raddatz (2009) foundthat growth in labour intensive sectors was the most poverty reducing.
ross-countryregressions simply represent average associations between variables. It is useful to contrasttheir results with careful large country time series studies. IFor Brazil Ferreira et. al. (2006)found that growth in the service industries was the most poverty reducing for the 1985-2004 period. For India, Datt and Ravallion (2010) found that pre-1991, rural growth wasmore poverty reducing than urban growth, but for the post 1991 period the reverse heldtrue.