CoChat
Sign Up
Cited by 2,180Agricultural Innovations and PracticesInnovation Diffusion and ForecastingPolicy Transfer and Learning

Learning about a New Technology: Pineapple in Ghana

Conley, Timothy G, Udry, Christopher R·2010·American Economic Review
DOI 10.1257/aer.100.1.35
Open this paper in CoChat — free account

Abstract

This paper investigates the role of social learning in the diffusion of a new agricultural technology in Ghana. We use unique data on farmers' communication patterns to define each individual's information neighborhood. Conditional on many potentially confounding variables, we find evidence that farmers adjust their inputs to align with those of their information neighbors who were surprisingly successful in previous periods. The relationship of these input adjustments to experience further indicates the presence of social learning. In addition, applying the same method to input choices for another crop, of known technology, correctly indicates an absence of social learning effects. (JEL D83, O13, O33, Q16)

AI snapshot

Written from the paper’s full text.

TL;DR

Conley and Udry use direct communication data from Ghanaian pineapple farmers to test whether farmers learn from information neighbors’ experiments, and find that farmers adjust fertilizer (and labor) inputs toward surprisingly successful neighbors’ practices, especially inexperienced farmers, with no analogous learning for established maize-cassava. They interpret these patterns as social learning, robust to controls for geographic growing conditions, credit/financial ties, clan, religion, and alternative network definitions.

Key findings

  • Farmers are more likely to change fertilizer use after bad news at their lagged input level: logit coefficient 12.32 [3.72]; a one-SD increase in this bad-news index (.12) raises the logit index by 1.5, while the 75th–25th percentile range of estimated logit indices is 2.8. Bad news at alternative levels lowers the probability of changing (coefficient -2.98 [1.91]; one-SD increase ~.13 lowers logit index ~.15).
  • For magnitude, the good-news input index M is positive and significant (0.99 [.16]); a one-SD increase in M (~4) is associated with ~4 cedis per plant increase in fertilizer use, greater than median fertilizer use per plant among users. Inexperienced farmers drive this: M*Inexperienced 1.09 [0.22] vs M*Experienced 0.10 [0.32], statistically different at 3% level.
  • Farmers respond more to news from experienced neighbors (1.02 [0.17] vs -0.13 [0.37] for inexperienced neighbors) and same-wealth neighbors (1.03 [0.18] vs -0.41 [0.32]), differences p<2%; big-farm vs small-farm and same-soil vs different-soil differences are not statistically significant.
  • Pineapple labor: M-tilde coefficient 0.95 [.37]; one-SD increase in M-tilde (396) associated with ~376 cedis per sucker increase, one-fifth median labor use per sucker (1,845). Maize-cassava: M-tilde coefficient 0.04 [0.14], virtually zero; geographic-neighborhood index 0.25 [0.14] (one-SD increase ~590,000 cedis/ha vs mean labor input 650,000 cedis/ha), attributed to growing conditions not learning.

Methods

  • Data: two-year survey (1996–98) of ~240 households in southern Ghana; four villages near Nsawam/Aburi, three where pineapple farmed; random sample of married individuals (60 couples/triples per village), 15 rounds ~6 weeks. Final estimation sample: 107 plantings by 47 farmers.
  • Information neighborhoods from direct communication: link if either farmer answered yes to “Have you ever gone to ___ for advice about your farm?”; median 2 info neighbors among pineapple farmers (max 33). Geographic neighbors within 1 km (median 12, max 25); alternative info metrics: roster of contacts (farm info only/full list) and predicted advice from logit. Financial neighborhoods used as controls.
  • Empirical strategy: model local learning about fertilizer productivity; construct indices of good/bad news at lagged and alternative input levels and index M of input levels associated with good news; estimate logit for whether fertilizer use changes and OLS for magnitude of innovation, controlling for geographic neighbors’ inputs to absorb correlated growing conditions, plus wealth, clan, religion, village, round, experience; spatial GMM standard errors (Conley 1999). Analogous labor regressions for pineapple and maize-cassava.

Limitations

  • Stated: small sample size leads to examining partitions one variable at a time rather than jointly; short panel prohibits including outcomes of lagged own experiments (only 14 farmers have more than 2 fertilizer changes).
  • Stated: data are early in learning process, so value of learning cannot be estimated well; inexperienced farmers often make mistakes, and the paper does not find increases in realized profits for those who change inputs. Labor results are cautioned because labor and fertilizer are not likely separable and are less robust to specification changes than fertilizer results.
  • Stated/evident: current pineapple farmers may be selected on ability to learn, potentially overstating responsiveness of farmers in general; information-neighborhood definitions have drawbacks (ask-advice based on random sample yields smaller-than-actual neighborhoods; roster measures include pairs who may not discuss farming and depend on subjective understanding); endogeneity of neighborhoods from unobservables is an important concern, though predicted-neighborhood results mitigate it. Evident: Monte Carlo check shows M coefficient becomes insignificant when added measurement-error SD is 4 (fertilizer mean 4, SD 7).

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1257/aer.100.1.35

Cites

107 of 107 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 10 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § 1 Introduction · p. 5

    Farmers are more likely to change their fertilizer use after information neighbors who use similar amounts of fertilizer achieve lower than expected profits.

    “(1) he is more likely to change his fertilizer use after his information neighbors who use similar amounts of fertilizer achieve lower than expected profits;”
    Quote in the paperSubstantial claim
  • Findingown data
    § 1 Introduction · p. 5

    Farmers' responsiveness to news about fertilizer productivity in their information neighborhood is much greater if they have only recently begun cultivating pineapple.

    “(3) his responsiveness to news about the productivity of fertilizer in his information neighborhood is much greater if he has only recently begun cultivating pineapple;”
    Quote in the paperSubstantial claim
  • Findingreasoning
    § 1 Introduction · p. 5

    The conclusions about fertilizer learning hold when conditioning on the fertilizer use of physically nearby farmers who experience unobserved growing conditions highly correlated with the farmer's.

    “These conclusions hold when conditioning on the fertilizer use of farmers who are physically nearby and who therefore experience unobserved growing conditions that are highly correlated with his.”
    Quote in the paperSubstantial claim
  • Findingown data
    § Abstract · p. 1

    Farmers in the study region adopt surprisingly successful information neighbors' practices, conditional on potentially confounding factors including common growing conditions, credit arrangements, clan membership, and religion.

    “We present evidence that farmers adopt surprisingly successful neighbors' practices, conditional on many potentially confounding factors including common growing conditions, credit arrangements, clan membership, and religion.”
    Quote in the paperSubstantial claim
  • Findingown data
    § 1 Introduction · p. 5

    Farmers respond more to news about fertilizer productivity on plots cultivated by experienced farmers and farmers with wealth similar to theirs.

    “(4) he responds more to news about the productivity of fertilizer on plots cultivated by experienced farmers and farmers with wealth similar to his.”
    Quote in the paperSubstantial claim
  • May combine two claims. 59
  • Classified as probably not a substantial claim. 30
  • The quote only partly states this claim. 1

  • May combine two claims.

    Findingown data
    § 1 Introduction · p. 5

    The conclusions about fertilizer learning are robust to a variety of different definitions of information flow between farmers, and conditional on the fertilizer use of farmers with whom the farmer has financial ties.

    “In addition, they are robust to a variety of different definitions of information flow between farmers, and conditional on the fertilizer use of farmers with whom he has financial ties.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 1 Introduction · p. 5

    Farmers increase (decrease) their use of fertilizer after information neighbors achieve unexpectedly high profits when using more (less) fertilizer than the farmer did.

    “(2) he increases (decreases) his use of fertilizer after his information neighbors achieve unexpectedly high profits when using more (less) fertilizer than he did;”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 3.1 Communication and Knowledge Data · p. 15

    There is evidence of spatial correlation in information link patterns: proximity increases link probability, but distance is not the sole determinant of links.

    “There is evidence of spatial correlation in link patterns as the marginal effect of proximity is to increase link probability, but distance is not the sole determinant of links.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results · p. 26

    The point estimate of the effect of good news at alternative levels of fertilizer use on the probability of changing is positive, and that of the effect of good news at the lagged level of use is negative, as expected, but these coefficients are very small and statistically indistinguishable from zero.

    “The point estimate of the effect of good news at alternative levels of fertilizer use on the probability of changing is positive, and that of the effect of good news at the lagged level of use is negative, as expected, but these coefficients are very small and statistically indistinguishable from zero.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • The authors state this as an assumption: Within farmer i's information neighborhood there is a set of H farmers using input level x_k,t sufficiently recently and sufficiently geographically close to k that they have common objective growing conditions.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • May combine two claims.

    Findingown data
    § Results · p. 27

    In Table 5 column A, the coefficient on the index of good news in the farmer's information neighborhood is positive, clearly statistically significant, and large.

    “Table 5, column A presents the results of estimating equation (9). The coefficient on the index of good news in the farmer's information neighborhood is positive, clearly statistically significant and large.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • May combine two claims.

    Findingown data
    § Results · p. 26

    Observations of bad news at alternative levels of fertilizer use decrease the predicted probability of changing; the coefficient is negative and significantly different from zero at the 12% level.

    “Observations of bad news at alternative levels of fertilizer use decrease the predicted probability of changing. The coefficient is negative and significantly different from zero at the 12% level.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • The authors state this as an assumption: Within farmer i's information neighborhood there is a set of H farmers using input level x_k,t sufficiently recently and sufficiently geographically close to k that they have common objective growing conditions.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • May combine two claims.

    Findingown data
    § Results · p. 27

    There is no aggregate trend in fertilizer use over the sample period; none of the round indicators is individually significant, nor are they jointly significantly different from zero.

    “There is no aggregate trend in fertilizer use over our sample period. None of the round indicators is individually significant, nor are they jointly significantly different from zero.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results · p. 27

    A one standard-deviation increase in M (about 4) is associated with an increase in fertilizer use of approximately 4 cedis per plant, which is greater than the median level of fertilizer use per plant of those farmers who use fertilizer.

    “A one standarddeviation increase in M (about 4 ) is associated with an increase in fertilizer use of approximately 4 cedis per plant, which is greater than the median level of fertilizer use per plant of those farmers who use fertilizer.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • May combine two claims.

    Findingown data
    § Results · p. 26

    Observations of bad news at the farmer's lagged input level strongly increase the predicted probability that he will change input levels; the estimated coefficient is positive, significantly different from zero (p-value 0.001), and large.

    “Observations of bad news at the farmer's lagged input level strongly increase the predicted probability that he will change input levels. The estimated coefficient is positive, significantly different from zero (p-value of 0 . 001 ) and large.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • The authors state this as an assumption: Within farmer i's information neighborhood there is a set of H farmers using input level x_k,t sufficiently recently and sufficiently geographically close to k that they have common objective growing conditions.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • May combine two claims.

    Findingown data
    § 3.2 Pineapple Production in Ghana · p. 16

    The official recommendation is 400 Kg. of fertilizer/hectare, more than 10 times the mean fertilizer application observed in the sample; only 4 of the 208 plantings observed exceeded the recommended level.

    “The recommendation is 400 Kg. of fertilizer/hectare, which is more than 10 times the mean fertilizer application observed in our sample. Only 4 of the 208 plantings we observed exceeded the recommended level of fertilizer application.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 3.2 Pineapple Production in Ghana · p. 19

    Pineapple farmers are much more likely to be male: 70% of pineapple farmers are male, while only 27% of non-cultivators are male.

    “Second, pineapple farmers are much more likely to be male: 70% of pineapple farmers are male, while only 27% of those who do not cultivate pineapple are male.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results · p. 26

    Inexperienced farmers seem more likely to change input levels, but while the size of the coefficient is large, it is not estimated with precision (different from zero at only approximately the 20% level).

    “As would be expected in virtually any model of learning, inexperienced farmers seem to be more likely to change input levels, but while the size of the coefficient is large, it is not estimated with precision (it is different from zero at only approximately the 20% level).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results · p. 27

    The estimated probability of changing fertilizer levels is strongly increasing in the average absolute deviation of farmer i's lagged inputs from his geographic neighbors, providing evidence of the importance of positively serially and spatially correlated unobserved shocks to the productivity of fertilizer.

    “The estimated probability of changing fertilizer levels is strongly increasing in the average absolute deviation of farmer i 0 s lagged inputs from his geographic neighbors, providing evidence of the importance of positively seriallyand spatially-correlated unobserved shocks to the productivity of fertilizer.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • The authors state this as an assumption: Within farmer i's information neighborhood there is a set of H farmers using input level x_k,t sufficiently recently and sufficiently geographically close to k that they have common objective growing conditions.

  • Probably not a substantial claim: may not pass S4 (based).

    Findingnone given
    § 1 Introduction · p. 5

    There is no evidence of similar learning in an established maize-cassava technology in these villages.

    “but that there is no evidence of similar learning in an established maize-cassava technology in these villages.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 3.2 Pineapple Production in Ghana · p. 19

    Over 20% of experienced pineapple farmers within each village have approached each other for advice about farming, while only 6% of non-pineapple farmers are in each others' information neighborhood.

    “Over 20% of experienced pineapple farmers (within each village) have approached each other for advice about farming, while only 6% of non-pineapple farmers are in each others' information neighborhood.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 10 Appendix 3: Constructing Indices of News · p. 51

    In our data many of the plantings are less than a full 5 periods before t1 so some information about their outcome is known to farmer i at t1, but not necessarily all of it.

    “However, in our data many of the plantings are less than a full 5 periods before t 1 so some information about their outcome is known to farmer i at t 1 , but not necessarily all of it.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 6.1 Alternative Information Neighborhoods · p. 29

    Across alternative definitions of the information neighborhood in Table 6, the coefficient on M is statistically significant and large for inexperienced farmers.

    “Regardless of the precise definition of the information neighborhood, the coefficient on M is statistically significant and large for inexperienced farmers (the standard deviation of M is approximately 3 . 5 for the first two information neighborhoods, and about 1 for the 'Predicted Advice' neighborhood).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 6.2 Robustness to Assumptions on Specification · p. 32

    Adding the lagged fertilizer level x_i,t-1 to the regression changes the coefficients on M_i,t by a magnitude comparable to other alternative specifications, and for inexperienced farmers M remains a significant predictor.

    “However, to be confident our results are not an artifact of mean reversion, in column F we add the lagged fertilizer level x i,t -1 to the regression. The coefficients on M i,t change by a magnitude comparable to some of our other alternative specifications and for inexperienced farmers it remains a significant predictor.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 6.3 Learning about Optimal Labor Use in Pineapple and in Established Crops · p. 35

    The coefficient for geographic neighborhood experiments for maize-cassava is large (one standard deviation increase associated with increased labor input of 590,000 cedis/ha, compared to mean labor input of 650,000 cedis/ha) and on the conventional boundary of statistical significance.

    “The coefficient is large (a one standard deviation increase in the index of experiments in the geographic neighborhood is associated with an increased labor input of 590,000 cedis/ha., compared to mean labor input of 650,000 cedis/ha.) and on the conventional boundary of statistical significance.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 6.3 Learning about Optimal Labor Use in Pineapple and in Established Crops · p. 34

    A one standard deviation increase in M (396) is associated with an increase in labor use of approximately 376 cedis per sucker, which is one-fifth of the median labor use per sucker on pineapple plots (1,845).

    “The coefficient is also large: a one standard deviation increase in ˜ M (which is 396 ) is associated with an increase in labor use of approximately 376 cedis per sucker, which is one-fifth of the median labor use per sucker on pineapple plots (which is 1 , 845 ).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data
    § 10 Appendix 3: Constructing Indices of News · p. 58

    The pseudo R-squareds for the change in input use regressions are .34 and .36 for columns A and B respectively.

    “Pseudo R-squareds .34 and .36, columns A and B respectively.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 7 Conclusion · p. 35

    A farmer increases (decreases) his fertilizer use after someone with whom he shares information achieves higher than expected profits when using more (less) fertilizer than he did.

    “Looking at the magnitude of innovations in fertilizer use, we find that a farmer increases (decreases) his use after someone with whom he shares information achieves higher than expected profits when using more (less) fertilizer than he did.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 6.2 Robustness to Assumptions on Specification · p. 31

    Including soil characteristics in the conditioning set leaves the core result unchanged: the coefficient on M is positive, large and statistically significant for inexperienced farmers but not for experienced farmers.

    “In column E we include information on soil characteristics in the conditioning set. We lose some observations by doing so, because soil testing was not completed on all plots, but once again the core result is unchanged: the coefficient on M is positive, large and statistically significant for inexperienced farmers but not for experienced farmers.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results · p. 28

    In column E, both coefficient estimates are large, positive, and statistically significant.

    “Column E presents analogous estimates with a partition of M depending upon the size of the farms in i 0 s information neighborhood. Both coefficient estimates are large, positive, and statistically significant.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingnone given
    § 10 Appendix 3: Constructing Indices of News · p. 53

    There is an occasional complication caused when the application window for a farmer's previous planting is still open when a new planting starts.

    “There is also an occassional complication caused when the application window for a farmer's previous planting is still open when a new planting starts.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 6.2 Robustness to Assumptions on Specification · p. 31

    When the geographic neighborhood is defined as within 500 meters, inexperienced farmers change fertilizer use in the direction of good news experiments by their information neighbors, but experienced farmers do not.

    “In column D, we alter the definition of the geographic neighborhood so that only plots within 500 meters fall within a plot's geographic neighborhood. Again, we find that inexperienced farmers change their fertilizer use in the direction of inputs associated with good news experiments by their information neighbors, but that experienced farmers do not.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data
    § 10 Appendix 3: Constructing Indices of News · p. 56

    The logit model predicting ask for advice has a sample size of 490 and a pseudo R-squared of .12.

    “Logit MLE Estimates, Sample Size = 490, Pseudo R-squared =.12.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 7 Conclusion · p. 35

    Farmers are more likely to change input levels upon receipt of bad news about the profitability of their previous level of fertilizer use, and less likely to change when they observe bad news about the profitability of alternative levels of fertilizer.

    “Our findings suggest that farmers are more likely to change input levels upon the receipt of bad news about the profitability of their previous level of fertilizer use, and less likely to change when they observe bad news about the profitability of alternative levels of fertilizer.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 6.3 Learning about Optimal Labor Use in Pineapple and in Established Crops · p. 34

    The coefficient of the learning index M for maize-cassava is virtually zero; a one standard deviation increase in M is associated with an increase in labor use of 11 thousand cedis per hectare, while the mean change in labor use is 153 thousand cedis and its standard deviation is 700 thousand.

    “The coefficient of the learning index ˜ M is virtually zero (at the point estimate, a one standard deviation increase in ˜ M is associated with an increase in labor use of 11 thousand cedis per hectare, while the mean change in labor use is 153 thousand cedis, and its standard deviation is 700 thousand).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 6.1 Alternative Information Neighborhoods · p. 30

    The result that bad news about alternate levels is associated with a statistically significant decrease in the probability of changing obtains for both roster of contacts measures, but not for the predicted neighborhoods measure.

    “Our result that bad news about alternate levels is associated with an statistically significant decrease in the probability of changing obtains for both roster of contacts measures, but not that based on predicted neighborhoods.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning
    § 6.2 Robustness to Assumptions on Specification · p. 32

    The coefficient on Γ_i,t changes most dramatically when lagged input levels are added, unsurprising because lagged input levels are spatially correlated and provide an alternate control for spatially correlated growing conditions.

    “The coefficient on Γ i,t changes the most dramatically; this is unsurprising since the lagged input levels are spatially correlated and so provide an alternate control for spatially correlated growing conditions.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data
    § 6.3 Learning about Optimal Labor Use in Pineapple and in Established Crops · p. 35

    Maize-cassava farmers adjust labor inputs in the direction of successful 'experiments' in their geographic neighborhood.

    “We see in column C that maize-cassava farmers adjust labor inputs in the direction of successful 'experiments' in their geographic neighborhood.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 6.2 Robustness to Assumptions on Specification · p. 30

    Under alternative timing assumptions in columns A and B of Table 7, the coefficient on M is positive and significantly different from zero for inexperienced farmers, and not for experienced farmers.

    “In both cases, the coefficient on M is positive and significantly different from zero for inexperienced farmers, and not for experienced farmers.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results · p. 28

    In column D, the index is an important and significant predictor for same-category neighbors but not for different-category neighbors.

    “Column D presents a partition depending on whether i 0 s information neighbor is in i 0 s wealth category (both rich or both poor). The index is an important and significant predictor for same category neighbors but not for different category neighbors.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • May combine two claims.

    Findingown data
    § Results · p. 28

    In column C, the coefficient of the index for experienced farmers' results is large and significant, while for inexperienced farmers it is not.

    “Column C defines M separately for inexperienced and experienced farmers in i 0 s information neighborhood. The coefficient of the index for experienced farmers' results is large and significant and for inexperienced farmers it is not.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • May combine two claims.

    Findingown data
    § 6.2 Robustness to Assumptions on Specification · p. 32

    In Monte Carlo experiments adding artificial measurement error, the estimated coefficient on M becomes insignificant at conventional levels when the standard deviation of the added noise is 4, while Γ remains a significant predictor.

    “The estimated coefficient on M becomes insignificant at conventional levels when the standard deviation of the added noise is 4 while Γ remains a significant predictor.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Findingnone given
    § 10 Appendix 3: Constructing Indices of News · p. 53

    Our farmers do not all plant the same number of pineapple plants per plot and they have observed differing amounts of plants' outcomes in the past.

    “One main issue remains: our farmers do not all plant the same number of pineapple plants per plot and they have observed differing amounts of plants' outcomes in the past.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § 6.3 Learning about Optimal Labor Use in Pineapple and in Established Crops · p. 33

    In focus group interviews, pineapple farmers reported uncertainty regarding the optimal amount of labor-intensive weeding; they were not sure how important it is to keep weeds clear of the plot.

    “In focus group interviews, pineapple farmers reported uncertainty regarding the optimal amount of labor-intensive weeding; they were not sure about how important it is to keep weeds clear of the plot.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data
    § 2 A Learning Model · p. 8

    The paper collected data on both the geographic location of plots and the flows of information between farmers, allowing it to distinguish the effects of spatially-correlated growing conditions from social learning.

    “We are able to distinguish the effects of spatially-correlated growing conditions from social learning because we collected data on both the geographic location of plots and the flows of information between farmers.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § 3.2 Pineapple Production in Ghana · p. 16

    In the empirical application, farmer wealth is included as a conditioning variable because it is likely strongly associated with variation in the opportunity cost of capital across farmers.

    “In the empirical application, we include farmer wealth as a conditioning variable, because this is likely to strongly associated with any variation in this opportunity cost across farmers.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given
    § Local Learning · p. 8

    The local learning rule is consistent with multiple kinds of learning, including Bayesian learning with independent priors over the elements of the support of x.

    “This local learning rule is consistent with multiple kinds of learning, including Bayesian learning with independent priors over the elements of the support of x .”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmers are trying to learn about the responsiveness of output to fertilizer.
    • The authors state this as an assumption: Farmers do not know the production function; it is the object of learning.
    • The authors state this as an assumption: Learning is local: a new observation with input level x only affects g(x) and not g(˜x) for ˜x ≠ x.
    • The authors state this as an assumption: Farmers are learning in the sense that g increases when they observe higher shock-adjusted output than expected and decreases when they observe lower than expected output.

  • Classified as probably not a substantial claim.

    Study design and conductown data
    § 3.1 Communication and Knowledge Data · p. 14

    The paper's measure of information availability is based on direct data about conversations between individuals.

    “We base our measure of information availability on direct data about conversations between individuals.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Interpretationnone given
    § Base Regression Specifications · p. 24

    Local learning implies that α1 and α4 are less than zero and α2 and α3 are greater than zero.

    “Local learning implies that α 1 , α 4 &lt; 0 and α 2 , α 3 &gt; 0 .”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given
    § Measuring Subjective Expectations, and Good/Bad News · p. 21

    Observation of the indicator d_i,k,t = 1 will result in farmer i adjusting g at input level x_k,t upwards.

    “Of course observation of the indicator d i,k,t = 1 will result in farmer i adjusting g at input level x k,t upwards.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.

  • May combine two claims.

    Study design and conductown data
    § 3.2 Pineapple Production in Ghana · p. 18

    All plots were mapped using global positioning system equipment, yielding more accurate measures of plot size and location than available in most surveys in LDCs and making it possible to distinguish explicitly between effects of information connections and geographic proximity.

    “In addition, all plots were mapped using global positioning system equipment. This procedure yields much more accurate measures of plot size and location than are available in most surveys in LDCs and makes it possible for us to distinguish explicitly between the effects of information connections and those of geographic proximity.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning
    § Base Regression Specifications · p. 24

    The sign and magnitude of changes in response to bad news are ambiguous because they are determined by the previously second-best input level, so the authors confine attention to the effect of good news on innovations.

    “The sign and magnitude of changes in response to bad news are ambiguous since they are determined by the previously second-best input level. Therefore, we confine our attention to the effect of good news on innovations.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning
    § 3.2 Pineapple Production in Ghana · p. 16

    Farmers argue the recommended fertilizer level is too high because fertilizer is 'too expensive', which the authors believe reflects an opportunity cost of capital much higher than the interest rate used to calculate the recommendation.

    “Farmers argue that the recommended level is too high because fertilizer is 'too expensive', which we believe reflects an opportunity cost of capital to these farmers that is much higher than the interest rate used to calculate the recommendation.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning
    § Implications for Actions · p. 11

    The paper's strategy uses geography and timing to identify spatially- and serially-correlated growing conditions, and direct data on information linkages to distinguish learning from similar growing conditions.

    “Our strategy is to use geography (and timing) to identify spatially- and serially-correlated growing conditions, and direct data on information linkages to distinguish the effects of learning from those of similar growing conditions.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation
    § Base Regression Specifications · p. 24

    The error term v_i,t is permitted to be conditionally heteroskedastic and spatially correlated across plots as a general function of their physical distance using the spatial GMM approach of Conley (1999).

    “Finally, the error term v i,t is permitted to be conditionally heteroskedastic and spatially correlated across plots as a general function of their physical distance using the spatial GMM approach of Conley (1999).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Conley (1999)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based). May combine two claims.

    Interpretationnone given
    § Base Regression Specifications · p. 24

    M_i,t, the index of input levels associated with good news, should predict the magnitude and direction of changes and should therefore have a positive coefficient β1.

    “M i,t , defined in (6), is our index of input levels associated with good news that should predict the magnitude and direction of changes, and should therefore have a positive coefficient β 1 .”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.
    • Rests on an assumption: The information conveyed by an information neighbor's planting depends on the number of plants involved.
    • Rests on an assumption: The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.
    • The authors state this as an assumption: One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data
    § 3.1 Communication and Knowledge Data · p. 14

    The question 'Have you ever gone to ___ for advice about your farm?' was reliably answered and is transparently related to the learning process under study.

    “We use responses to this question as our benchmark definition of information neighbors because during the field research it appeared reliably-answered and it is transparently related to the learning process under study.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisereasoning
    § Construction of Indices Reflecting Innovations in Information · p. 21

    A given farmer plants pineapple only occasionally rather than every period, so the authors do not always observe a farmer's reactions to individual observations of (x, w, y).

    “A given farmer plants pineapple only occasionally, rather than every period. Hence we do not always see a farmer's reactions to individual observations of ( x, w, y ) .”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § Base Regression Specifications · p. 24

    Γ_i,t is a crucial control for movements in x_i,t induced by correlated growing conditions.

    “Γ i,t defined in (7) is our crucial control for movements in x i,t induced by correlated growing conditions.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Premisereasoning
    § Measuring Subjective Expectations, and Good/Bad News · p. 20

    High, positive spatial and serial correlation in growing conditions shocks permits estimation of the rational expectations analog of the objective expectation of profits for plots within a farmer's information neighborhood.

    “High, positive spatial and serial correlation in growing conditions shocks permits estimation of the 'rational expectations' analog of E i,t π k,t +9 ( x k,t , w k,t ) : the objective expectation of π k,t +9 ( x k,t , w k,t ) for k within i 0 s information neighborhood for a particular but unknown (to the econometrician) value of w k,t .”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Within farmer i's information neighborhood there is a set of H farmers using input level x_k,t sufficiently recently and sufficiently geographically close to k that they have common objective growing conditions.

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given
    § 1 Introduction · p. 3

    Direct data on information interconnections is typically unavailable to economists.

    “Direct data on information interconnections is typically unavailable to economists.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning
    § Base Regression Specifications · p. 24

    The specification includes a regressor analogous to Γ_i,t with the neighborhood definition based on financial rather than geographic neighborhoods, motivated by alternate explanations that significant β1 estimates might be caused by omitted variable bias because information neighbors share common access to credit arrangements.

    “In addition, it includes a regressor analogous to Γ i,t with the neighborhood definition based on financial rather than geographic neighborhoods. This is motivated by alternate explanations that would suggest that significant β 1 estimates might be caused by omitted variable bias because information neighbors share common access to credit arrangements.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation
    § 3.2 Pineapple Production in Ghana · p. 16

    There is agronomic evidence that pineapple yields are very responsive to fertilizer and that the impact varies with local soil conditions and moisture patterns.

    “There is agronomic evidence that pineapple yields are very responsive to fertilizer and that the impact varies with local soil conditions and moisture patterns (Abutiate (1991); Purseglove (1972)).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Abutiate (1991)

      Not checked in the free run.

    • cites Purseglove (1972)

      Not checked in the free run.

  • May combine two claims.

    Premisereasoning
    § 1 Introduction · p. 3

    The set of assumptions relating observed relationships to unobserved information flows is critical for measuring social learning but can rarely be tested because of data limitations.

    “This set of assumptions is critical for the measurement of the extent of social learning, but can rarely be tested because of data limitations.”
    Quote in the paperSubstantial claim

  • The quote only partly states this claim.

    Interpretationreasoning
    § Implications for Actions · p. 9

    In the local learning model, the model offers no prediction for the direction or magnitude of an input change induced by bad news about an input level.

    “However, bad news about x k,t provides no information about which among multiple alternatives the farmer will choose if he changes inputs in response to it.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmers are trying to learn about the responsiveness of output to fertilizer.
    • The authors state this as an assumption: Farmers do not know the production function; it is the object of learning.
    • The authors state this as an assumption: Learning is local: a new observation with input level x only affects g(x) and not g(˜x) for ˜x ≠ x.
    • The authors state this as an assumption: Farmers are learning in the sense that g increases when they observe higher shock-adjusted output than expected and decreases when they observe lower than expected output.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Study design and conductnone given
    § Base Regression Specifications · p. 25

    Endogeneity of neighborhoods arising from individuals choosing neighbors who would be good to learn from is not a concern.

    “Endogeneity of neighborhoods that arises from individuals choosing neighbors who would be good to learn from is also not a concern.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationnone given
    § Base Regression Specifications · p. 25

    The parameters of equations (8) and (9) are not parameters of a structural model and are best interpreted as those for a prediction of innovations in behavior.

    “The parameters of equations (8) and (9) are not parameters of a structural model and are best interpreted as those for a prediction of innovations in behavior.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Premisereasoning
    § 1 Introduction · p. 2

    If there are multiple adopters of a new technology in similar circumstances, the process of learning about the new technology may be social.

    “If there are multiple adopters of the new technology in similar circumstances, as is often the case with an innovation in agriculture, then the process of learning about the new technology may be social .”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisereasoning
    § 3.2 Pineapple Production in Ghana · p. 17

    The empirical work focuses on profits rather than yields because quality variation in pineapple output implies there is no natural unit for yield, and because the cost of other inputs, especially labor, should be deducted from the value of output.

    “In our empirical work, we focus on profits rather than yields because quality variation in pineapple output implies there is no natural unit in which yield can be measured. In addition, the cost of other inputs (most importantly labor) should be deducted from the value of output.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Interpretationreasoning
    § Base Regression Specifications · p. 24

    The authors expect unobserved shocks to growing conditions to be positively spatially and serially correlated and thus α5 > 0.

    “We expect unobserved shocks to growing conditions to be positively spatially and serially correlated and thus α 5 &gt; 0 .”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.

  • May combine two claims.

    Study design and conductreasoning
    § 9 Appendix 2: Alternative Measures of Information Connections · p. 51

    The ask-for-advice measure is based on a random sample of other farmers, and so yields estimates of the information neighborhood of a farmer that are smaller than his actual information neighborhood.

    “The ask for advice measure is based on a random sample of other farmers, and so yields estimates of the information neighborhood of a farmer that are smaller than his actual information neighborhood.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § 8 Appendix 1: Social Learning in a Dynamic Context · p. 42

    This prediction replicates that of the myopic model, demonstrating that the myopic model's predictions can survive the introduction of forward-looking behavior.

    “This prediction replicates that of the myopic model, demonstrating that the myopic model's predictions can survive the introduction of forward-looking behavior.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Farmers are Bayesian learners.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given
    § 9 Appendix 2: Alternative Measures of Information Connections · p. 51

    The roster of contacts measures include some pairs who probably do not discuss farming activities, and depend upon the respondents' subjective understanding of 'significant conversations about farming.'

    “The roster of contacts measures include some pairs who probably do not discuss farming activities, and depends upon the respondents' subjective understanding of 'significant conversations about farming.'”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation
    § 10 Appendix 3: Constructing Indices of News · p. 58

    The analysis uses spatial GMM (Conley 1999) standard errors that allow for heteroskedasticity and correlation as a function of physical distance.

    “Logit MLE point estimates, spatial GMM (Conley 1999) standard errors in brackets allow for heteroskedasticity and correlation as a function of physical distance, see footnote 20 for details.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Conley 1999

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationnone given
    § 7 Conclusion · p. 36

    Inexperienced farmers often make mistakes by switching to what is truly a suboptimal input level after seeing it perform surprisingly well in a small number of experiments.

    “Inexperienced farmers often make mistakes by switching to what is truly a suboptimal input level after seeing it perform surprisingly well in a small number of experiments.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Interpretationreasoning
    § Details · p. 48

    A period one term is absent from ∂V1L/∂g1(H) because L rather than H has been chosen; the derivative is the marginal change in expected output times the probability that H will be chosen in period two, averaged over the values of the growing conditions shock that may realize in period 2.

    “A period one term is of course absent from ∂V 1 L ∂g 1 ( H ) as L rather than H has been chosen and the derivative is again the marginal change in expected output times the probability that H will be chosen in period two, averaged over the values of that the growing conditions shock may realize in period 2 .”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning
    § Details · p. 48

    The second term of ∂V1H/∂g1(H) corresponds to the marginal change in profits, w2, times the subjective probability that H will be chosen in period two, averaged over prospective shock values w2.

    “The second term corresponds to the marginal change in profits, w 2 , times the subjective probability that H will be chosen in period two, averaged over prospective shock values w 2 .”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § 8 Appendix 1: Social Learning in a Dynamic Context · p. 42

    This increases the relative value to the farmer of choosing that level of fertilizer use, and thus increases the probability that that level will be chosen in the following period.

    “This increases the relative value to the farmer of choosing that level of fertilizer use, and thus increases the probability that that level of fertilizer use will be chosen by the farmer in the following period.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmers are Bayesian learners.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given
    § 10 Appendix 3: Constructing Indices of News · p. 58

    A full set of round dummies were included in the regressions but not reported.

    “A full set of round dummies were included but not reported.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § 7 Conclusion · p. 36

    Information has value in these villages, as do the network connections through which that information flows.

    “Information, therefore, has value in these villages, as do the network connections through which that information flows.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § Details · p. 50

    There is a unique critical value of w1 ( ˆw1) defined implicitly when V1H( ˆw1) equals V1L( ˆw1), such that if w1 ≤ ˆw1 the farmer chooses x1 = L, and otherwise x1 = H.

    “It can be shown as in the myopic model, there is a unique critical value of w 1 ( ˆ w 1 ) defined implicitly when V 1 H ( ˆ w 1 ) equals V 1 L ( ˆ w 1 ) , such that if w 1 ≤ ˆ w 1 the farmer chooses x 1 = L , and otherwise x 1 = H .”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data
    § 7 Conclusion · p. 35

    These findings hold when controlling for common input usage due to similar unobservable growing conditions via conditioning on actions of geographic neighbors, for credit arrangements using a notion of financial neighborhoods, and across several information metrics.

    “These findings hold when controlling for common input usage due to similar unobservable growing conditions via conditioning on actions of geographic neighbors, for credit arrangements using a notion of financial neighborhoods, and across several information metrics.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationnone given
    § 8 Appendix 1: Social Learning in a Dynamic Context · p. 42

    If a neighbor achieves particularly high growing-conditions-adjusted output at a particular level of fertilizer use in period t, then a farmer's beliefs about the productivity of that level shift upward.

    “If his neighbor achieves particularly high growing-conditions-adjusted output at a particular level of fertilizer use in period t , then a farmer's beliefs about the productivity of that level of fertilizer shift upward.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Farmers are Bayesian learners.

  • May combine two claims.

    Interpretationreasoning
    § Details · p. 50

    If growing conditions over adjacent periods is sufficiently strongly positively correlated, the condition for ∂V1H/∂g1(H) > ∂V1L/∂g1(H) will hold; similarly, if the range of growing conditions outcomes is sufficiently restricted, this condition will hold even if growing conditions were not serially correlated.

    “If growing conditions over adjacent periods is sufficiently strongly positively correlated, this condition will hold; similarly, if the range of growing conditions outcomes is sufficiently restricted, this condition will hold even if growing conditions were not serially correlated.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning
    § Details · p. 48

    The first term of ∂V1H/∂g1(H) reflects expected profits in period one being linear in expected shock-adjusted output.

    “The first term of ∂V 1 H ∂g 1 ( H ) simply reflects expected profits in period one being linear in expected shock-adjusted output.”
    Quote in the paperNot a substantial claim
79 of 79

  • The authors state this as an assumption.

    Farmers are trying to learn about the responsiveness of output to fertilizer.

    “The basic form of this model is that farmers are trying to learn about the responsiveness of output y i,t +1 on plot i to a discrete-valued input x i,t which we will call fertilizer:”

  • The authors state this as an assumption.

    The productivity shock is expectation zero and IID across farmers and time.

    “ε i,t +1 is an expectation zero productivity shock that is IID across farmers and time.”

  • The authors state this as an assumption.

    Farmers do not know the production function; it is the object of learning.

    “Farmers do not know the function f ; it is the object of learning.”

  • The authors state this as an assumption.

    Growing conditions are positive, exogenous, influence the marginal product of fertilizer, and are correlated across farmers and time.

    “The variable w i,t is a positive, exogenous growing conditions variable influencing the marginal product of x i,t that is correlated across farmers and time.”

  • The authors state this as an assumption.

    Growing conditions are observable to farmers but not to the econometrician.

    “We assume the w i,t are observable to farmers but not the econometrician.”

  • The authors state this as an assumption.

    Farmers' subjective expectations can be defined.

    “We will not be specific about how the farmer's beliefs evolve, but simply posit that farmer i 's subjective expectations, E i,t ( · ) , can be defined.”

  • The authors state this as an assumption.

    Learning is local: a new observation with input level x only affects g(x) and not g(˜x) for ˜x ≠ x.

    “We consider learning rules that are local in the sense that a new observation with input level x will only affect g i,t +1 ( x ) and not g i,t +1 (˜ x ) for ˜ x = x.”

  • The authors state this as an assumption.

    Farmers are learning in the sense that g increases when they observe higher shock-adjusted output than expected and decreases when they observe lower than expected output.

    “Weassume in addition that farmers are learning in the sense that g i,t +1 ( x ) -g i,t ( x ) &gt; 0 when they observe higher shock-adjusted output than expected ( y w &gt; g i,t ( x )) , and likewise reduce their subjective expectation when they observe lower than expected y w .”

  • The authors state this as an assumption.

    The baseline model assumes farmers are myopic.

    “The time t problem faced by a myopic farmer i is to choose the input, x, to maximize his subjective expectation of time t +1 profits for this input choice”

  • The authors state this as an assumption.

    Inputs are discrete.

    “Since inputs are discrete, this is nothing more than choosing input level x ∗ i,t so that:”

  • The authors state this as an assumption.

    The opportunity cost of capital is the same across farmers.

    “In our theoretical discussion, we incorporate the opportunity cost of capital in the price of fertilizer and assume it is the same across farmers.”

  • The paper relies on this without support.

    A plot planted at time t receives its crucial fertilizer inputs during periods t+1 through t+4, chemically forced flowering occurs approximately at t+5, and the pineapple harvest is generally complete by t+9.

    “A plot planted at time t receives its crucial fertilizer inputs during periods t +1 through t +4 . Chemically forced flowering occurs approximately at t +5 , at which point the eventual size of the crop begins to become apparent. The pineapple harvest is generally complete by t +9 .”

  • The paper relies on this without support.

    Plots have to be harvested on a single day for efficient export of fresh fruit by air to Europe, and it is essential that fruits mature simultaneously, requiring common treatment across plants within the plot.

    “Plots have to be harvested on a single day for efficient export of the fresh fruit by air to Europe. It is essential, therefore, that the fruits mature simultaneously, which requires common treatment across plants within the plot.”

  • The authors state this as an assumption.

    Farmer i increases or decreases his expectation of the productivity of fertilizer at x_k,t according to whether realized profit minus his expectation is positive or negative.

    “Farmer i increases (decreases) his expectation of the productivity of fertilizer at x k,t according to whether y k,t +9 w k,t -g i,t ( x k,t ) is positive (negative).”

  • The authors state this as an assumption.

    Within farmer i's information neighborhood there is a set of H farmers using input level x_k,t sufficiently recently and sufficiently geographically close to k that they have common objective growing conditions.

    “Suppose that within farmer i 0 s information neighborhood there is a set of H farmers using input level x k,t sufficiently recently (say, at t -1 ) and sufficiently geographically close to k that they have common objective growing conditions.”

  • The paper relies on this without support.

    The information conveyed by an information neighbor's planting depends on the number of plants involved.

    “The information conveyed by an information neighbor's planting depends on the number of plants involved, so we weight each observation by Plants k,τ .”

  • The paper relies on this without support.

    The change in beliefs associated with a piece of information depends on the stock of previous experiments that farmer i has observed.

    “The change in beliefs associated with a piece of information depends on the stock of previous experiments that i has observed, so we normalize by the total number of plants i has observed from the beginning of our data until t 1 .”

  • The authors state this as an assumption.

    One fifth of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t+5 through t+9.

    “ψ ( . ) reflects the idea that 1 / 5 of the relevant information about the eventual harvest of a period t planting is revealed in each of the periods t +5 through t +9 .”

  • The authors state this as an assumption.

    The weighting scheme used to reflect gradual revelation of information is ad hoc.

    “Obviously this weighting scheme is ad hoc ; we examine the robustness of our results to variations in the weighting scheme in Section 6 .”

  • The authors state this as an assumption.

    Farmers are Bayesian learners.

    “so we assume that farmers are Bayesian learners”

  • The authors state this as an assumption.

    In base specifications, information about the eventual size of the harvest of a period t planting began to be revealed after flowering in period t+5, with the outcome fully known at harvest in t+9.

    “Recall that in our base specifications we assumed that information about the eventual size of the harvest of a period t planting began to be revealed after flowering in period t +5 , with the outcome fully known at harvest in t +9 .”

  • The paper relies on this without support.

    The characteristics of the maize-cassava production function are well-known to farmers in these villages and there should be no uncertainty about the role of labor in maize-cassava production.

    “The characteristics of the maizecassava production function are well-known to farmers in these villages and there should be no uncertainty about the role of labor in the production of maize-cassava.”

  • The authors state this as an assumption.

    Good productivity in period 1 is not too bad news about likely productivity in period 2; a sufficient condition is that µ(.|a) first-order stochastically dominates µ(.|b) for a > b.

    “In addition, we need to assume that good productivity in period 1 is not too bad news about the likely productivity in period 2 : a sufficient condition is that µ ( . | a ) fi rst-order stochastically dominates µ ( . | b ) for a &gt; b .”

  • The authors state this as an assumption.

    g1(H) ≥ g1(L).

    “As in the myopic model, this requires g 1 ( H ) ≥ g 1 ( L ) .”

  • The paper relies on this without support.

    The change in growing conditions from period 1 to period 2 is not so strong that yields are doubled.

    “If we do not expect the change in growing conditions from period 1 to period 2 to be so strong that yields are doubled”

  • The paper relies on this without support.

    Each neighbor's planting using a given input level provides the farmer with some information about that level's productivity.

    “Each neighbor's planting from t 0 up to t 1 -1 using a given input level provides the farmer with some information about that level's productivity.”

  • The paper relies on this without support.

    Some information in a planting τ that is 1 to 4 periods before t0 would be revealed after t0 and be news from the perspective of time t1.

    “Some information in the τ planting would be revealed after t 0 and be news from the perspective of time t 1 .”

  • The paper relies on this without support.

    Some of the information in a planting τ that is 1 to 4 periods before t0 should be reflected in the t0 planting but not all of it would be revealed at t0.

    “Some of the information they contain should be reflected in the t 0 planting but it would not all be revealed at t 0 .”
28 of 28

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The study uses spatial fixed effects to control for local-level unobservables.

    “We follow the literature (Conley and Udry 2010; Goldstein and Udry 2008; Magruder 2011) and use spatial fixed effects to control for local level unobservables.”

    From Environmental and gender impacts of land tenure regularization in Africa: Pilot evidence from Rwanda (2014) · cites this paper as Conley and Udry 2010

    The Introduction (p. 4) says the study uses geographic and soil data to control for spatially correlated confounders, and Controls for Similar Growing Conditions (p. 23) describes an index based on nearby plots. The Results (p. 26) describe spatial standard errors, not spatial fixed effects. Thus the source supports controlling for local conditions, but not the attributed method of spatial fixed effects. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

  • The instrumental variable (Early Copilot Adopter Exposure) satisfies the relevance condition.

    “We believe the first component satisfies the relevance condition because prior research shows that individuals learn from peers through social interactions and shared experience (Foster and Rosenzweig 1995, Conley and Udry 2010, Bollinger and Gillingham 2012, Bailey et al. 2022).”

    From The Impact of Generative AI on Collaborative Open-Source Software Development: Evidence from GitHub Copilot&amp;nbsp; (2024) · cites this paper as Conley and Udry 2010

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Fertilizer application window ends in two different periods

    Minor

    The production timeline says fertilizer is applied from t +1 through t +4, but the later description says the window runs from t + 1 through t + 1.

1 of 1

Cites

Flagged first, then by citations

4 of the 111 references couldn’t be matched to a known paper.

Research works printed but not linked: 1

  1. Moffitt Robert A, 2001, Social dynamics, ed. Steven N. Durlauf and H. Peyton Young, 45

Cited by

Sorted by citations

Related

Authors
Conley, T. G., & Udry, C. R.
Affiliations
  • University of Chicago (US)
  • Yale University (US)
ORCID
Venue
American Economic Review
Publisher
American Economic Association
Published
2010-03-01
Volume
100
Issue
1
Pages
35-69
ISSN
0002-8282 (ISSN-L)
DOI
10.1257/aer.100.1.35
OpenAlex
W2005029869
Semantic Scholar
045f9f5baee3c1f4e6089e4b073dd84d6624c77a
Cited by
2,180
References
111
Citations by year
2026: 143 · 2025: 121 · 2024: 98 · 2023: 135 · 2022: 154 · 2021: 153 · 2020: 132 · 2019: 126 · 2018: 134 · 2017: 117 · 2016: 109 · 2015: 118 · 2014: 130 · 2013: 115 · 2012: 94
FWCI
390.45
Type
journal article
Language
en
Open access
closed
Locations
  • American Economic Review (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • VTechWorks (Virginia Tech) (submittedVersion)
Topics
  • Life Sciences › Agricultural and Biological Sciences › General Agricultural and Biological Sciences › Agricultural Innovations and Practices
  • Social Sciences › Decision Sciences › Management Science and Operations Research › Innovation Diffusion and Forecasting
  • Social Sciences › Political Science and International Relations › Policy Transfer and Learning
Keywords
social learning, pineapple, farming experience, Ghana

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-08.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-08.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review