Systems approaches to value chains and diseases

Views:
 
     
 

Presentation Description

Presented by Prof. Karl M. Rich and Kanar Hamza from the Norwegian Institute of International Affairs at WorldFish headquarters in Penang, Malaysia, on the 7th of February, 2013.

Comments

Presentation Transcript

Systems approaches to value chains and diseases:

Systems approaches to value chains and diseases Prof. Karl M. Rich and Kanar Hamza Norwegian Institute of International Affairs Invited lecture, WorldFish Centre, 7 February 2013

Outline :

Outline Motivation – overcome gaps in qualitative value chain analyses Introduction to quantitative tools for ex-ante VC assessment – a role for system dynamics Examples

Gaps in qualitative VCA:

Gaps in qualitative VCA A major gap in VCA: understanding the impact of VC investments The general performance of a chain The ability to evaluate ex-ante between different options Value chain analysis does a very nice job of telling stories, of describing the chain and things that influence it. But it is less good on measurement.

Gaps in qualitative VCA:

Gaps in qualitative VCA An example: suppose tomorrow you were given US$20 million to improve an existing agricultural value chain Given the tools you have so far, could you evaluate how best to use that money?

Gaps in qualitative VCA:

Gaps in qualitative VCA A second example: how do we reconcile the different value systems and constraints at different nodes of the value chain? Different actors have different incentives – can we develop tools that allow us to unpack those incentives, incorporate feedbacks between different actors, and identify win-win solutions?

Gaps in qualitative VCA:

Gaps in qualitative VCA In this lecture, we will look at quantitative tools to better evaluate value chains. This remains a very rich research field – not much has been done here and many tools have not yet been fully explored. A lot to be done at NUPI in 2013 (work with ILRI and FAO)

PowerPoint Presentation:

We will mainly focus on tools that look at the value chain (or supply chain) as an entity. Other more traditional tools could have application that are not discussed in detailed here (e.g., SAMs). Quantitative approaches to VC modeling

PowerPoint Presentation:

Numerous examples of agricultural supply chains in the logistics/operations research literature: Van der Vorst et al. (2000, Eur. J. OR) Minegishi & Thiel (2000, Sim . Prac . & Theory) Trienekens & Hvolby (2001, Prod. & Plant Control) Gigler et al. (2002, Eur. J. OR) Georgiadis et al. (2005, J. Food Eng.) Fiala (2005, Omega) Meijer et al. (2005, working paper, Wageningen U.) Main focus is on micro (firm) level strategies (reducing costs, lead times, inventories) Quantitative approaches to VC modeling

PowerPoint Presentation:

System dynamics (SD) models present a means to capture sector/macro level interventions in supply chains: Assess investment options Analyze dynamic feedbacks to determine sustainability Applications of SD models in agribusiness: Mowat et al. (1997): R&D strategies in persimmons Cloutier & Sonka (1998): SD model of coordination between producers/processors in the hog supply chain Fisher et al. (2000): SD model of adoption/diffusion of precision agriculture Ross (2005): assessment of entrepreneurial innovations in the hog supply chain. Quantitative approaches to VC modeling

Quantitative approaches: SD methods:

What is system dynamics? System dynamics (or SD for short) is “a method to enhance learning in complex systems” ( Sterman 2000: 4). There is increased recognition that systems thinking is an essential part of decision-making, given the numerous interactions and interconnections present in any system of actors. Quantitative approaches: SD methods

Quantitative approaches: SD methods:

System dynamics models are simulation approaches that help decision-makers understand the impacts and consequences of system changes, given interactions and feedbacks between actors. The approach is grounded in the theory of non-linear dynamics and feedback control from engineering, physics, and mathematics Applications are numerous, including technical, biological, and economic systems, as well as the study of human behavior, organizational reform, and policy processes Consequently (and necessarily), SD models are interdisciplinary in nature, both in design and interpretation. Quantitative approaches: SD methods

Central concepts of SD:

Central concepts of SD Stocks (accumulation) Flows (change overtime – rate/time unit) Feedback loops (circular causality)

Equivalent representative of stock & flow :

Equivalent representative of stock & flow Source: Sterman (2000)

Quantitative approaches: SD methods:

An important concept of SD models is the notion of feedback Feedback is “the process wherein one component of the model initiates changes in other components, and those modifications lead to further changes in the component that set the process in motion” ( McGarvey and Hannon 2004: 6). More simply, feedback determines the dynamic process of a system and how things evolve over time. Quantitative approaches: SD methods

Quantitative approaches: SD methods:

Feedback can be positive or negative. Positive feedback is self-reinforcing. It amplifies what happens in the system. For instance, consider the feedback loop between chickens and eggs below. As chickens lay more eggs, this increases the number of chickens, which increases the number of eggs later, and so on. Eggs Chickens + + Quantitative approaches: SD methods

Quantitative approaches: SD methods:

Negative feedback is self-correcting. It counteracts and opposes change the system. Consider the feedback loop between chickens and road crossings below. As the number of chickens rises, this increases the number of attempted road crossings (remember what happened when the chicken tried to cross the road!). This subsequently reduces the number of chickens . Chickens Road crossings + - Quantitative approaches: SD methods

Quantitative approaches: SD methods:

In any system, there will be a combination of positive and negative feedbacks How a system behaves will depend on this interaction, much of which will be complex and not intuitive (thus the need for models!) Note that the words “positive” and “negative” do NOT conjure any value judgment about the nature of feedback (i.e., good or bad). Rather, they refer to it being self-reinforcing and self-correcting, respectively. Quantitative approaches: SD methods

PowerPoint Presentation:

An example will illustrate why this concept of feedback is so important. In the SD and management literature, a modeling tool called “the beer game” was developed to understand decision-making in a supply chain of interacting actors. Quantitative approaches: SD methods

Beer game:

Beer game The idea of the beer game: Each brewery has four sectors: retailer, wholesaler, distributor, and factory One person manages each sector Every week, customers demand beer from the retailer, who fills orders from inventory Retailers must place orders from wholesalers to replenish stocks. Wholesalers must order beer from distributors Distributors order beer from the factory.

PowerPoint Presentation:

Producer Distributor Wholesaler Retailer Consumer Consumer orders Retailer orders Wholesaler orders Distributor orders

Beer game:

Each actor in the chain wants to minimize total costs, based on inventory costs and stockout costs (i.e., costs for having a backlog of orders). Goal: keep inventories as low as possible while avoiding backlogs. Note that there is a delay between placing orders and receiving them. Consumer demand is known and simple, with an increase during the course of the game. Beer game

Beer game:

What are some of the lessons from the game? First, most people who play this do pretty badly, despite its apparent simplicity! Why? People forget about the supply line, in terms of the delays inherent in orders and the feedbacks that are implicit . Instead, we react to what we see happening right away ! Metaphor for livestock in developing countries Beer game

Quantitative approaches: SD methods:

Let’s take another example from the livestock sector, specifically livestock production. The production of livestock itself is a complex system based on a combination of biological parameters and market interactions that influence decision-making. Quantitative approaches: SD methods

PowerPoint Presentation:

Consider first the production cycle for livestock. Animals go through different stages of life Birth Growing Maturity Slaughter/death Quantitative approaches: SD methods

Quantitative approaches: SD methods:

We can represent this as a simple flow chart reflecting these stages in livestock production Gestation Young animals Mature animals Sold animals Quantitative approaches: SD methods

PowerPoint Presentation:

Some animals are also used for breeding stock to produce animals for the future Gestation Young animals Mature animals Sold animals Breeding stock Quantitative approaches: SD methods

PowerPoint Presentation:

We can further label these processes between stages of production: Gestation Young animals Mature animals Sold animals Breeding stock Birth Maturing Aging Breeding Aging Quantitative approaches: SD methods

PowerPoint Presentation:

What can we say already about this system? It is dynamic : it takes time to get from one stage to another Relationships are non-linear External influences can (and will) have important impacts on how this system evolves over time. Quantitative approaches: SD methods

PowerPoint Presentation:

Let’s think about some of those external influences and parameters to this model; what should we consider? Technical parameters (breeding rates, litter sizes, litters per year) Time lags between stages Demand for livestock (animals and meat) Inputs (and input prices) for feed, supplies, etc. Quantitative approaches: SD methods

PowerPoint Presentation:

So, given this system, what would be the effect of an increase in the price of feed? Quantitative approaches: SD methods

PowerPoint Presentation:

Gestation Young animals Mature animals Sold animals Breeding stock Birth Maturing Aging Breeding Aging Quantitative approaches: SD methods

PowerPoint Presentation:

As the discussion showed, it’s complicated! Increased feed will make production more expensive, which in turn will influence behavior as to how much breeding stock to keep (vs. sell). This in turn will influence breeding stock in future periods, which will subsequently affect how much livestock is produced. These “feedback” effects are a powerful and not always well understood characteristic of complex systems that SD models can help us analyze. Quantitative approaches: SD methods

PowerPoint Presentation:

Let’s bring in the value chain concept here. How could SD tools be used analyze investment options, ex-ante ? Quantitative approaches: SD methods

Portrayal of a generic SD-VCA model :

Portrayal of a generic SD-VCA model

Example: Disaggregation of the production and financial block.:

Example: Disaggregation of the production and financial block.

PowerPoint Presentation:

Example : An evaluation of commodity-based trade in Namibia ( Naziri , Rich, and Bennett 2012) Research question : what are the net benefits in developing commercialized livestock value chains in communal production areas? Model examines interface between production and processing, including feedbacks. Quantitative approaches: SD methods

Motivation of developing an integrated aquatic disease management model:

Motivation of developing an integrated aquatic disease management model We motivate a generic framework that can be used more broadly in analyzing the impacts of alternative aquatic disease control strategies.

Aquatic disease management: sea lice in salmon farms (portrayal of model blocks):

Aquatic disease m anagement: sea lice in salmon farms (portrayal of model blocks)

Aging & growth sector :

Aging & growth sector

Epidemiology sector:

Epidemiology sector

Policy sector:

Policy sector

Economic sector:

Economic sector

Model highlights:

Model highlights The results of the model showed that early control treatment events: (1) improve the efficiency of controlling lice population, (2) enhance fish growth, and (3) reduce lice control costs. The results showed that launching treatment events before reaching thresholds greatly facilitate to prevent high peaks of lice population and hence breaking lice population cycles.

Summary:

Summary Quantitative analysis of VCs is a fruitful area of future research Key: envision the VC as a system of interacting actors – other available techniques need to keep this in mind. Also metrics for performance needed

authorStream Live Help