Are Farmers with Postharvest Loss Acting Rationally?

Rice leaking out of holes in gunny sacks at village milling facility near Chennai, India. Credit: ADMI/K. Wozniak

Rice leaking out of holes in gunny sacks at village milling facility near Chennai, India.
Credit: ADMI/K. Wozniak

by Steve Sonka, Ag Economist

Events vs. Structure

External observers of postharvest loss in developing countries typically have little difficulty identifying loss and waste.  Indeed each winter the ADM Institute contributes to the travel funding for University of Illinois business and agricultural engineering students to study in rural India.  The resulting student blogs (see one called Losses Along the Way) and pictures document the existence of loss upstream in the food chain and waste at retail and consumer levels. However, years ago, an agricultural economist, T.W. Schulze, won the Nobel Prize for his pioneering work indicating that smallholder farmers are economically rational – at least to the same extent as senior managers of multinational firms in the West.

Although initially seeming to be contradictory, the sentences of the prior paragraph actually underscore the need for a systems perspective in addressing and designing interventions to reduce postharvest loss. The external observers and our Illinois students are responding to the “events” they see. The smallholder farmers and the managers of the associated food chains make decisions and operate in response to the “structure” of their food chain and surrounding culture.

The Meaning of a ‘Systems Perspective’

The word systems pervades discussions of agriculture and food domains. The apparent meaning of systems relates to physical components of a larger whole. It is easily understood that a supply chain for tomatoes operates as an interlinked system of physical stages to move product from the field to the village or to an urban market. A less visible aspect of systems relates to the behaviors that ultimately determine performance and the even less visible factors that determine the dominant pattern of behaviors.

Blog_140407_2

Figure 2: The Three Levels of a Systems Hierarchy

Figure 2 illustrates the commonly used hierarchy to understand and improve systems performance. The figure depicts three levels which are further described relative to the example of soybean production in Mato Grosso. As the noted in that example, observance of an event can trigger a logical response – slow down the harvest process. However that response may be inconsistent or only partially successful in the context of the patterns and structures that powerfully shape system performance.

Systems Example: Soybeans in Matto Grasso

  • Events – The Mato Grosso region of Brazil has experienced phenomenal growth in soybean production and its production systems employ world-class technologies.  Yet during harvest, even the casual observer notes considerable grain lying along the road.  This grain has fallen off the trucks transporting the grain from the field to market.  The postharvest loss associated with grain lying on the side of the road is a readily observable event.
  • Patterns – In Mato Grosso, a safrinha crop of maize or sometimes cotton is grown after the soybean crop.  Because of the rainfall patterns in the region, there is intense pressure to plant the safrinha crop before the rainy season ends.  Managers and workers in the region operate under intense pressure to “get the harvesting done” to allow planting to occur.  The recent emergence of this phenomenon has imposed a new pattern in this farming system.
  • Structure – A key contributor to the harvest time pressures in Mato Grosso is the relative absence of local storage.  Instead the bulk of the crop is immediately transported very long distances to port facilities after harvest.  More local storage would reduce that pressure.  However, until very recently, the policies of important government loan programs strongly favored purchase of equipment over grain storage facilities.  That government policy is an important factor determining the structure of harvest systems in Mato Grosso.

It, of course, is reasonable to emphasize events – that’s what draws our attention to the topic of interest. However, intervention not only needs to be aware of patterns and structure, they need to anticipate and take actions informed by the knowledge of how patterns and structure are likely to impact the results of the intervention.  Patterns and structure most often are portrayed as factors that retard the desired change to the occurrence of the events of concern.  However, they also can be key leverage points that accelerate the desired change.

Planning Interventions

One outcome of the growth in our understanding of the importance of a systems perspective is the phenomenon known as “paralysis by analysis”. Understanding the systems dynamics of a complex food chain is difficult. Attempting to simultaneously change all the components of an operating food chain typically is not a good idea. The better practice is to ensure recognition of the key patterns and structural factors as a means to shape the targeted response. Waiting until all the factors of the system can be changed seldom is productive.

Careful needs assessment during the design phase is essential and must extend beyond identifying direct impediments. It is important to determine why they exist as well. Exploration of tacit relationships, intentionally including a wide range of participants of the food chain, can add insights regarding why events occur in the manner they do. In essence, conduct of the needs assessment and the broader design phase needs to be done in a manner that aggressively exploits peripheral vision. Also although structural effects often are presented as impediments to innovation, in some instances structure can provide leverage points to accelerate change.

Steve Sonka is an ag economist at the University of Illinois and the former Director of the ADM Institute. More expert perspective posts by Sonka can be found on the Preventing Postharvest Loss blog.

Share this via
Facebook Twitter Email

, , , ,

No comments yet.

Leave a Reply