Underground Knowledge: Estimating the Impacts of Soil Information Transfers Through Experimental Auctions†

Abstract

Soil degradation, a serious environmental problem in many developing countries, often necessitates the use of fertilizers to improve crop yields. However, smallholders usually do not have sufficient information about their soil nutrient levels to make profit‐maximizing decisions about fertilizer usage. We conducted two‐round experimental auctions to determine whether providing information and fertilizer recommendations from inexpensive soil testing kits to farmers in western Kenya affected their behavior and ability to optimize their input choices. We auctioned organic and inorganic agricultural inputs, dividing farmers into information treatments, and analyzed the data using double and triple difference estimations. We find that providing soil information has significant effects on farmers' willingness to pay for inputs. We then use Monte Carlo simulations to show that there is potential for high net benefits to farmers from individualized soil tests. These results suggest that soil testing can be a cost‐effective method to increase food security in the region.

Chatting at Church: Information Diffusion through Religious Networks

“…Decreasing arable land size due to population pressures has led to more intensive cropping cycles with fewer fallow periods, draining the soils of necessary nutrients (Muyanga and Jayne, 2014). In soil samples collected as part of this study from farm households in western Kenya (and described in detail in Murphy et al (2020)), most plots had soils with critically low levels of nitrogen and phosphorus, and a significant number had critically low organic matter levels as well. Therefore, significant benefits exist if farmers can increase their knowledge of effective soil management practices with the aim of rehabilitating their soils, increasing yields and income, and becoming more food secure.…”

“…While many attend their RI multiple times a week, at the median, individuals attend their RI on a weekly basis. A plurality of individuals attend a Pentecostal or Charismatic church, while Anglicans are 1 See Murphy et al (2020) for additional information about the sample. We sampled data from 18 villages, but combined data from two villages given their close proximity, which gave us 17 distinct village clusters for our analysis.…”

Show and Tell: Causal Impacts of Field Days on Farmer Learning for Organic Inputs in Kenya

“…To frame our analysis of the effects of FFDs on farmer WTP, we use a model after Johnson and Myatt (2006), Liaukonyte, Streletskaya, and Kaiser (2015) and Murphy et al (2019) to analyze the effect of information on consumer demand. In our context, we have a new input k with profitability π k (E k ), where for simplicity we assume π k is primarily a function of the unit increase in yields for a unit increase of the input applied per hectare (or agronomic efficiency) E (Vanlauwe et al, 2011).…”

“…In practice, this experimental auction had two auction rounds, but we only use bids from the first round for this analysis. The sample drops outliers in bids so as to be consistent withMurphy et al (2019). Bids are dropped from the analysis when the change in those bids for an individual between auction rounds is in the top or bottom 1% of the sample.…”

“…We measured levels of nitrate, phosphate, potassium, sulphur, and active carbon present in the soil using SoilDoc, which adapts standard wet chemistry analysis of soil samples into a low-cost and portable kit. SeeMurphy et al (2019) for additional information.…”