Search

• Root and tuber crops hedge against climate shocks, crop failures and food insecurity. • Extended harvests are a challenge for survey data collection on crop production. • An experiment evaluates competing survey methods on cassava production measurement. • Diary keeping with phone calls, or 6-month recall (as second-best), is recommended. • There is evidence of gross overestimation in publicized cassava yields in Malawi. Root and tuber crops, such as cassava, could be planted to hedge against climate shocks, seasonal crop failures and food insecurity during the lean season. Since their harvests occur over extended periods and often in small quantities, they present a serious challenge for household and farm surveys aiming to collect reliable information on crop production based on recall. To document the relative accuracy of recall-based approaches to survey data collection on cassava production vis-à-vis diary-based techniques, a survey experiment was implemented in Malawi over a 12-month period. The sampled cassava-producing households were randomly assigned to one of four treatments, including (1) daily diary-keeping, with semi-weekly supervision visits (diary-visit); (2) daily diary-keeping, with semi-weekly supervisory phone calls (diary-phone); (3) two six-month recall interviews, with six months in between; and (4) a single 12-month recall interview. We find that compared to diary-visit, the household-level annual cassava production is 295 kg higher under diary-phone. This effect corresponds to 28 percent of the average diary-visit annual production estimate. Since unavoidable, albeit limited, lapses in diary keeping over a 12-month period may have led both diary variants to underestimate true production, higher annual cassava production estimate obtained under diary-phone implies that this treatment is closer to the true estimate. Although the difference between the estimates based on six-month recall and diary-visit is statistically insignificant, 12-month recall, on average, underestimates annual production (i) by 516 kg with respect to diary-phone (corresponding to 37 percent of the diary-phone average) and (ii) by 221 kg with respect to diary-visit (corresponding to 21 percent of the diary-visit average). While the recall-based approaches both record production estimates lower than the diary-phone, six-month recall does so to a lesser extent. And supported by a crop cutting operation in which all sampled households participated irrespective of their assigned survey treatment, the analysis demonstrates likely gross overestimation in competing international and ministerial statistics on cassava yields in Malawi. For improved microdata on root and tuber crop production, the findings lend support to the adoption of (i) diary-keeping with phone calls, particularly if deployed in a broader mobile phone–based survey, or (ii) six-month recall, as a second-best alternative. The adoption of these practices can then facilitate a renewed look at the role of cassava farming in poverty, food security and agricultural production. [ABSTRACT FROM AUTHOR]

• Survey question recall length is associated with measurement error in agricultural survey data. • Measurement error affects key agricultural output, input, and productivity variables. • Recall length effects are sizeable at 2–5 percent of the value of the variable per additional month. • The recall effect affects the reliability of key metrics, including SDG indicators. • Investment in better measurement and conscientious survey design are critical. This paper assesses the relationship between the length of recall and nonrandom error in agricultural survey data. Using data from LSMS-ISA surveys in Malawi and Tanzania, we show that key input and output variables are systematically related to the length of the recall period, indicating the presence of nonrandom measurement error. With longer recall periods, farmers report higher quantities of harvest, labor and fertilizer inputs. Farmers list fewer plots as the recall period increases. We argue that it is plausible that farmers over-estimate plot-level outcomes – harvest, labor and fertilizer inputs – while it is also plausible that they forget some of their more marginal plots as their memory decays due to longer recall periods. We find evidence of measurement error related to the length of recall also in common measures of agricultural productivity. The size of the recall effect typically varies between 2 and 5 percent per additional month of recall length, making its impact on the reliability of key agricultural indicators, including SDG indicator 2.3, economically significant. With data reliability affecting policy effectiveness, improving agricultural survey data quality remains an important concern, especially in light of the ambitious 2030 agenda reflected in the Sustainable Development Goals. Mainstreaming objective measures where possible and reducing the risk of recall error through shorter recall periods appear to be promising avenues to improve the quality of key variables in agricultural surveys. [ABSTRACT FROM AUTHOR]