41 Value chain modelling

Economist classify fisheries as primary industries, highlighting their supporting role for numerous secondary (e.g., seafood processors) and tertiary industries (e.g., hotels and restaurants) in the economy.^{[1] [2]} Extractive industries are generally limited at creating value. For example, the direct economic contribution of fisheries to national economies around the world ranges only between 0.5% and 2.5% of their Gross Domestic Product (GDP).^{[3] [4]}.

However, fishing requires inputs from other industries to operate (e.g., boat building industry and fishing net manufacturers),^[5] and as fish, marine invertebrates and macroalgae move along supply chains, they are transformed into accessible seafood products tailored to meet the needs of consumers.^[6] Thus, fisheries have “upstream” (i.e., prior to fishing) and “downstream”(i.e., post-harvesting) economic effects, which are commonly characterized using input-output models to estimate multipliers (see the External bio-economic models chapter). These multipliers are factors used for approximating the extent of the contribution of an economic sector to a nation’s economy.

The EwE Value Chain plug-in is a powerful tool for characterizing the downstream economic effects of the fishing fleets.^[7] The economic agents or components within the value chain are referred to as enterprises described via a set of common attributes (see Table 1) and segregated based on their function within the value chain. The value chain model is coupled to ecosystem model through the fishing fleets. These are considered as producers, given that they are the main source of raw materials for the seafood supply chain. Non-extractive activities using marine living resources (e.g., ecotourism or non-retaining recreational fisheries) and some types of aquaculture enterprises can also be considered as producers. Enterprises that receive marine living resources from producers and transform them into seafood are classified as processors. These typically include fish cutters and filleters, canneries, seafood freezing facilities, or fishmeal processing plants.^{[8] [9] [10]} However, some aquaculture operations can also be classified as processors if they receive feed or seeds from other processors or producers within the system.^[11]

Table 1. Input parameters used to characterize seafood value chains in EwE’s value chain plug-in.

The remaining enterprises in the seafood value chain are classified as either distributors or sellers, depending on their functional roles. Distributors typically include middlemen linking producers and processors, enterprises specialized on transporting seafood from processors to sellers, or seafood exporters linking producers with foreign markets. Sellers, on the other hand, connect producers, processors and distributors with consumers, and these typically include seafood wholesalers, supermarkets, small municipal markets, street vendors and restaurants.

It is important to highlight that the information required to populate the value chain within EwE, such as an enterprise’s management costs or its number of female workers (Table 1) must be expressed in a per tonne basis (e.g., $ per tonne or jobs per tonne). Moreover, the forms within the plug-in (Ecopath > Output > Tools > Value chain) provide ample freedom regarding what information to include. If available data for characterizing enterprises is highly aggregated, it is still possible to populate the value chain by including a single item in the cost structure. The same is the case for the employees, workers, and dependents if data is not segregated by sex.

Links between enterprises are characterized by tracking losses in weight and gains in value are estimated using the ratios between the weight of products leaving an enterprise and the weight of inputs it used to create them. These values are provided in live weight equivalents (e.g., the weight of fish in the can) and not the total weights (e.g., weight of cans including fish, liquids, and tin). Gains in value are estimated in a similar manner, using the ratios between the value of products and inputs flowing through each enterprise.

Calculations for estimating (i) revenue, (ii) profit, (iii) contributions to GDP, and (iv) employment for all enterprises described within value chain are expressed with the following equations,

[latex]L_c=W_{p,c}\cdot \prod\limits_{e=1}^c (\frac {W_{i,e}}{W_{p,c}} ) \tag{1}[/latex]

where L_c is the live weight equivalent for a given value enterprise for which the value chain holds enterprises from the first (a producer) to the last element in the chain (c),  W_p,e is the weight of products for the enterprise (e), and W_i,e is the weight of input (“raw material”) for the same enterprise.

[latex]R_p=W_p \cdot (R_a+R_e+R_i+R_s) \tag{2}[/latex]

where R_p is the overall production revenue for the enterprise. Revenues from subsidies (U) are calculated from,

[latex]U=W_p (U_e+U_o) \tag{3}[/latex]

[latex]\text{Total revenue } (R)=R_p + U \tag{4}[/latex]

[latex]\text{Cost of input and operation } (I)=W_p (I_c+I_e+I_i+I_s+C_m+C_l+C_c)\tag{5}[/latex]

[latex]\text{Cost of observers } (O) = W_p (C_o \cdot O_r)\tag{6}[/latex]

[latex]\text{Taxation costs } (T) = W_p (T_e+T_x+T_p+T_v+T_i+T_l)\tag{7}[/latex]

[latex]P_w = \begin{equation} \left\{ \begin{array}{cc} W_p \cdot (P_s + P_h) , \text{ if using a wage system} \\ W_p \cdot V_{f,s} (S_s+S_h) , \text{ if using a share system} \end{array} \right. \end{equation} \tag{8}[/latex]

where V_f,s is the value of the product (by fleet and by species) per unit weight.

[latex]P_o = \begin{equation} \left\{ \begin{array}{cc} W_p \cdot (P_f + P_m) , \text{ if using a wage system} \\ W_p \cdot V_{f,s} (S_f+S_m) , \text{ if using a share system} \end{array} \right. \end{equation} \tag{9}[/latex]

[latex]\text{Total costs } (C)=I+O+T+P_w+P_S \tag{10}[/latex]

We calculate the number of jobs for workers (J_w) and owners (J_o), and the total number of jobs from the sum of J_w and J_o.

[latex]J_w=W_p (J_s+J_h)\tag{11}[/latex]

[latex]J_o=W_p (J_f+J_m)\tag{12}[/latex]

Further the numbers of dependents of workers (D_w) and owners (D_o) is calculated from

[latex]D_w=W_p (D_s \cdot J_s + D_h \cdot J+h) \tag{13}[/latex]

[latex]D_o=W_p (D_f \cdot J_f + D_m \cdot J_m) \tag{14}[/latex]

which can be summed to give the total number of dependents, D = D_w + D_o.

For producers, it is assumed that the number of jobs is proportional to effort, while their income depends on the catch value of the catches.

The socio-economic indicators (i.e., i-iv, see above) can be summarized by: (a) functional group, (b) fishing fleet, and (c) any enterprise within the value chain, or (d) for the whole fisheries sector. Moreover, the contributions to the GDP and employment can be divided between activities taking place at sea and on land to estimate income and employment multipliers (e.g., how many jobs (or $) are made on land for each job (or $) made at sea) by fleet, functional groups and across the fisheries system.

At the Ecopath stage, the value chain models can be used to address multiple questions related to characterizing the economic network. For example: Which fishing fleets are the most important contributors to national employment (at sea and on-land)? Are the income multipliers similar among functional groups and fishing fleets? How big is the fishing industry in economic terms? Do mackerels contribute more to a country’s GDP when fished by purse seiners or by gillnetters? Are canneries and seafood freezing plants paying similar wages to the women they employ? How many people are employed in export-driven seafood supply chains in comparison to those selling locally? Are people earning annual salaries above the minimum wage across all enterprises in the seafood value chain?^{[12] [13] [14]}

Value chains are good tools for highlight the roles played by marginalized groups within the fisheries sector (e.g., women, small-scale fishers, subsistence fishers) in a systematic manner. For example, characterizing seafood value chains in Peru under this approach revealed that supply chains starting with small-scale fishers were the main contributors to employment and GDP across the fisheries sector, although being responsible for only 15% of the country’s catch.^[15] Additionally, the implementation of this approach in Baía Formosa (Brazil) allowed the users to quantify the indirect contribution of subsistence fisheries to local economies.^[16]

Furthermore, understanding value chains’ flows and structures opens new opportunities for addressing “what-if” questions regarding the end use of marine living resources, and how these affect their potential contribution to the economy, employment and job security (a known limitation of input-output models)^[17]. For example, a study revealed that a transition from the fishmeal-dominated status quo to a hypothetical scenario where all anchoveta (Engraulis ringens) landed in Peru were used for canning; would result in a 53% reduction in the country’s fishmeal production, a 21% increase in fisheries sectors’ profitability, a 183% increase in job creation, and 179 times more seafood production.^[18] Moreover, national and provincial value chains can be compared to highlight how the same functional groups and fishing fleets can have dissimilar employment and income multipliers at different spatial scales (revealing the local importance of certain functional groups or fishing fleets).^[19]

At the Ecosim model stage, value chain models can be used for equilibrium analyses. For example, Christensen et al. ^[20] sought to assess the maximum sustainable yield (MSY) by setting a constant fishing effort over a 25 year-long Ecosim run, letting the system reach a steady state and then repeating the run with a new fishing effort level. They explored a wide range of effort levels (from no exploitation to overexploitation) on a theoretical fleet targeting tuna. In each step revenue, fishing costs, income, and employment for the fleets and the entire supply chain, were registered. This approach allowed researchers to: (i) highlight trade-offs between fishing fleets (e.g., tuna fleet vs mackerel fleet), (ii) showcase how fishing costs are non-linear (although commonly assumed to be so in equilibrium analysis [6]), and (iii) highlight the strengths of using MSY instead of the Maximum Economic Yield (MEY) as a descriptor for the maximum socio-economic benefits that can be achieved by a fishery.^[21]

Moreover, value chain models can also be used together with Ecosim and Ecospace for studies involving hindcasts and forecasts. This tool is robust for expressing the socio-economic outcome of “what-if” scenarios (from climate change to fisheries policies), and can be used for testing the effects of broader economic policies (e.g., the introduction of new taxes or the elimination of fuel subsidies). Moreover, granted that it allows users to define the cost-income structure of the fleets in a more detailed manner (Ecopath > Output > Tools > Value chain > Components > Producers), even if only producer data is available, the Value Chain plug-in can strengthen estimates for the fishing profitability over time, as well as simulate the evolution of particular variable cost (e.g., fuel costs) across scenarios.

Yet, it is important to note that in these instances, food web modelling outputs (i.e., catches per fleet per functional group) will enter the value chain model without affecting its base parametrization. Thus, some value chain outputs might be misleading given that value chain parameters are not equally stable over time. For example, processing yields (e.g., the difference between the total weight of tuna entering a cannery and the weight of tuna in the cans coming out of it) tend to be quite stable over time, unless important changes in processing technology come to play. Alternatively, if a processing industry (e.g., reduction industry) is growing in a country, then the number of workers or owners per tonne of processed fish will vary substantially between years. Thus, if the processing capacity is in excess in the year corresponding to the base value chain model, then projections on the contribution this industry for total employment might be overestimated in future years (as the processing plant could still take more fish without having to adjust its labour force).

Given this issue, it is key to ground model outputs with data when using the value chain models, particularly for medium to long-term projections. First and foremost, users should highlight, when reporting results, that uncertainty in model outputs increases substantially over time. The classic economic assumptions of ceteris paribus will be damaging to fisheries system if the resulting management advice is given based exclusively on socio-economic indicators using a single set of value chain parameters. Just as one should avoid giving fisheries management advice using only Ecopath models, one should also avoid using static value chains. Solutions for these limitations can be undertaken by:

1. Developing multiple value chain models for consecutive years (e.g., 5 or 10 years) and running Ecosim over each value chain parametrization. The outputs could be plotted together to highlight the consequences of parameter and structural uncertainty in the socio-economic projections, or
2. Constructing a base value chain model for a 5-year or 10-year period (by averaging annual parameter values over time) and then using the coupled value chain-Ecosim runs only to study changes over that same period or a future period of the same length.

The latter solution is certainly a less preferable one. However, access to data might make it difficult to develop multiple consecutive full value chain models. Notwithstanding, adopting a scenario approach to value chain outputs is still useful, particularly by forcing users to explicitly state their hypothesis on what is assumed to remain constant over time and why.

Nonetheless, EwE’s Value Chain plug-in grants unique capabilities for directly estimating the income and employment multipliers of specific marine living resources (i.e., through EwE functional groups), producers (i.e., fishing fleets) and the fisheries sector. Moreover, it provides modelers with a platform to synthesize large amounts of socio-economic knowledge of fisheries systems to provide a description of its whole economic subsystem in a succinct but comprehensive manner, much like what an Ecopath model does for an ecosystem’s food web.

The process of constructing a value chain model requires developing a working hypothesis of its structure and understanding what information is available, how it is stored, who has access to it, and who collects and updates it. This process is quick to reveal areas (e.g., enterprises, items within their cost structure or segments of the value chain) with information deficits, that can be used to prioritize research and monitoring efforts. Moreover, if conducted in a participatory and inclusive manner, this process can help strengthen seafood traceability,^[22] improve managers’ understanding of leverage points along the supply chain, and be used to simulate the consequences of potential governmental interventions (both at sea and on land) on the national economy. For example, value chain models in EwE can be used together with the Management Strategy Evaluation (MSE) Module or the CEFAS MSE plug-in (See the MSE CEFAS tutorial)^[23] to simulate the effects of implementing alternative management procedures (e.g., harvest control rules) on individual or multiple functional groups using indicators that describe the ecological, economic and social components of fisheries systems. This allows modelers the capacity to directly quantify trade-offs amongst management objectives and harvest strategies, while highlighting how the costs and benefits of the different management procedures are distributed among stakeholders and their enterprises within the system. ^{[24] [25] [26]}.

Explore this tool further through the value chain tutorial.