GROUNDWORK STUDIO
Groundwork is an independent research studio analysing nature- and climate risks
and providing decision-making support for corporates and public bodies.
Demystifying Indicators Part 2
From Causal Chains to Causal Networks
In Part 1, we explored a basic but essential question – What is an indicator? We introduced the idea of the causal chain as a starting point for thinking about indicators and capturing cause-effect relationships in environmental systems.
“Indicators are decision tools that exist to simplify complexity.”
In this article, we take a step deeper, exploring the Causal Network concept, as described by Niemeijer and de Groot (2006). The causal network creates a system for choosing indicators, clarifies what indicators represent, and guides the analysis of complex environmental systems.
The causal network sets an explicit request at the outset. Mapping the processes that shape the environmental issue that is particular to the organisation. Once the whole system is mapped, indicator selection becomes more purposeful, answering the question that is based on the scope of the study.
Purpose of environmental indicators
Environmental indicators reveal trends or phenomena that are not immediately detectable. They simplify complex ecological processes into signals that can be monitored and interpreted over time.
Dale and Beyler (2001) describe indicators as tools that capture aspects of ecological systems while remaining practical for monitoring and management.
Dale and Beyeler are of the opinion that indicator selection should be guided by the concept of ecological integrity. Imagine seeing it as a hierarchy
and that indicators should ideally capture processes across hierarchies to reflect structural, functional and composition of ecosystems.
Ecological indicators, which are indicators specifically representing ecosystems should do the following, according to Dale and Beyler:
- be relatively easy to measure
- be sensitive to stress on the system
- respond to stress in predictable ways
- provide early warning signals of change
- allow management actions to prevent undesirable outcomes
- integrate information about system conditions
- have known responses to disturbances, including anthropogenic pressures
- show low variability unrelated to environmental change
Why do we need causal networks?
Environmental analysis begins with a causal chain, describing cause and effect and linking human/ industry activities to an environmental outcome/ consequence. An example of this can be seen in eutrophication processes in freshwater ecosystems.
Fertilizer use à nutrient runoff à increases nitrogen content in lakes à algal blooms à organic matter decomposition à decline in oxygen content à decline in aquatic species
This clarifies cause and effect in a linear fashion.
However, environmental systems rarely follow a simple linear pathway. Activities may produce multiple environmental pressures. Ecological changes may be caused by several interacting causes, or feedback loops, between ecological processes and social processes. Processes occurring in one part of the system can have influence over others across space and time.
A single causal chain may therefore overlook important relationships or even fail to capture the complexity of the entire system and/or consequences associated with an activity.
The Causal Network Concept
A causal network brings an issue into perspective through variables that can influence it. This shows the environmental issue as a network of interacting pathways where
- nodes represent conceptual variables or indicators
- arrows represent causal relationships between them
Multiple causal chains intersect within the network. Variables may influence several other variables simultaneously, and feedback loops may emerge between ecological and social processes.
This approach better reflects the complexity of environmental systems, where ecological, economic and social dynamics are closely intertwined.
By explicitly mapping these relationships, the causal network helps analysts understand where indicators should be placed to capture key system dynamics.
Building a Causal Network
Niemeijer and de Groot propose a structured method for constructing causal networks before selecting indicators.
Step 1 — Define the Issue
The issue must be clearly defined and bounded. In many ways, this step corresponds to defining the research question.
For example, the Figure 1 is taken from Niemeijer and de Groot where the clearly defined issue in question is “ecological impacts of nitrogen fertilization on surface water ecosystems adjacent to agricultural land” (in the Netherlands).
Keeping the scope relatively narrow helps keep the causal network manageable.
Step 2 — Define Scope
The next step is to define the boundaries of the system or the scope of the study being studied.
This includes specifying:
- the ecological system under consideration
- climatic conditions
- socio-economic context
- spatial scale
- temporal scale
Boundary conditions also determine what is excluded from the analysis. For example, a study may focus on nitrogen impacts within agricultural landscapes while excluding upstream processes such as fertilizer manufacturing. In the example, the study considers in situ temperate climate conditions for the Netherlands.
Step 3 — Identify potential Indicators
Before selecting measurable indicators, the analysis begins by identifying conceptual variables that represent key processes in the system.
Examples might include:
- nutrient load
- species composition
- crop production intensity
- fertilizer application rates
At this stage, the focus is on capturing the major processes and relationships relevant to the issue.
Step 4 — Construct the Causal Network
Once the conceptual variables are identified, they can be organized into a network of cause–effect relationships.
Niemeijer and de Groot suggest grouping variables into three broad domains:
Environment-related variables such as soil, water, air and biodiversity.
Society-related variables such as policy decisions, market demand, agricultural practices or consumer behaviour.
Pressure interface where societal activities interact directly with environmental systems. Agricultural production using nitrogen fertilizer is one such interface.
The variables are then connected through directional links representing causal relationships.
Rather than forming a single chain, the variables form a network of interacting pathways.
How the Causal Network Guides Indicator Selection
Once the causal network has been constructed, indicator selection becomes a more systematic process.
Several practical considerations must then be addressed:
- the specific research question
- availability of data
- spatial and temporal scale
- monitoring capacity
- the analytical perspective being used (for example, environmental or socio-economic)
Key nodes within the causal network can then be identified. Indicators are selected to represent these nodes in a way that captures the most important processes within the system.
The starting point remains the research question. Without a clearly defined question, indicator selection risks becoming arbitrary or driven by data availability rather than conceptual relevance.
A short note on Nodes
Nodes are points in the network that are centers of activity. They can be root nodes, central nodes or end-of-chain nodes.
End-of-chain nodes have multiple incoming arrows showing long range of causes and effects. Take for example Water fauna.
Central nodes are those that typically have multiple incoming and outgoing arrows. In this example, they can be N concentrations in water or N emissions.
Root nodes – there are not so many here, but one could imagine that manure use or fertilizer use is a root node in this example. They have multiple outgoing arcs in Crop production.
In practice, not every node has a corresponding indicator. It is therefore important to identify key nodes where measurement provides most insight into system behaviour or where management interventions can influence outcomes.
Implications for Corporate Sustainability
The causal network approach also has implications beyond ecological research.
In corporate sustainability contexts, indicators are often selected to satisfy requirements for internal audits, environmental management systems or sustainability reporting. Causal networks see companies as inherently nested in physical, ecological systems where risks from loss of ecosystems are real.
However, without a clear conceptual foundation, these indicators may fail to reflect the underlying environmental processes associated with a company’s activities.
Applying a causal network approach requires that organizations:
- define system boundaries clearly
- articulate causal relationships between activities and environmental outcomes
- justify indicator selection based on conceptual relevance rather than convenience
- connect ecological indicators with social and economic decision-making
This approach increases transparency and improves the credibility of environmental reporting.
Indicators should therefore not be treated as decorative elements in sustainability reports. They are analytical tools that shape how environmental issues are understood and addressed.
Robust indicator selection strengthens:
- scientific credibility
- decision relevance
- governance transparency
- policy coherence
For example, a food processing company assessing water-related risks may construct a causal network linking sourcing of agricultural raw materials, nutrient runoff from processing, watershed conditions around the factory, and ecosystem impacts such as eutrophication.
Ultimately, indicators influence not only how environmental change is measured, but also how it is interpreted and managed.
In our next article, we look at indicators in different corporate contexts to see how they were chosen and find the causal network. We also discuss how indicators enable decision-making in corporate contexts.
Demystifying Indicators Part 2
From Causal Chains to Causal Networks
In Part 1, we explored a basic but essential question – What is an indicator? We introduced the idea of the causal chain as a starting point for thinking about indicators and capturing cause-effect relationships in environmental systems.
“Indicators are decision tools that exist to simplify complexity.”
In this article, we take a step deeper, exploring the Causal Network concept, as described by Niemeijer and de Groot (2006). The causal network creates a system for choosing indicators, clarifies what indicators represent, and guides the analysis of complex environmental systems.
The causal network sets an explicit request at the outset. Mapping the processes that shape the environmental issue that is particular to the organisation. Once the whole system is mapped, indicator selection becomes more purposeful, answering the question that is based on the scope of the study.
Purpose of environmental indicators
Environmental indicators reveal trends or phenomena that are not immediately detectable. They simplify complex ecological processes into signals that can be monitored and interpreted over time.
Dale and Beyler (2001) describe indicators as tools that capture aspects of ecological systems while remaining practical for monitoring and management.
Dale and Beyeler are of the opinion that indicator selection should be guided by the concept of ecological integrity. Imagine seeing it as a hierarchy
and that indicators should ideally capture processes across hierarchies to reflect structural, functional and composition of ecosystems.
Ecological indicators, which are indicators specifically representing ecosystems should do the following, according to Dale and Beyler:
- be relatively easy to measure
- be sensitive to stress on the system
- respond to stress in predictable ways
- provide early warning signals of change
- allow management actions to prevent undesirable outcomes
- integrate information about system conditions
- have known responses to disturbances, including anthropogenic pressures
- show low variability unrelated to environmental change
Why do we need causal networks?
Environmental analysis begins with a causal chain, describing cause and effect and linking human/ industry activities to an environmental outcome/ consequence. An example of this can be seen in eutrophication processes in freshwater ecosystems.
Fertilizer use à nutrient runoff à increases nitrogen content in lakes à algal blooms à organic matter decomposition à decline in oxygen content à decline in aquatic species
This clarifies cause and effect in a linear fashion.
However, environmental systems rarely follow a simple linear pathway. Activities may produce multiple environmental pressures. Ecological changes may be caused by several interacting causes, or feedback loops, between ecological processes and social processes. Processes occurring in one part of the system can have influence over others across space and time.
A single causal chain may therefore overlook important relationships or even fail to capture the complexity of the entire system and/or consequences associated with an activity.
The Causal Network Concept
A causal network brings an issue into perspective through variables that can influence it. This shows the environmental issue as a network of interacting pathways where
- nodes represent conceptual variables or indicators
- arrows represent causal relationships between them
Multiple causal chains intersect within the network. Variables may influence several other variables simultaneously, and feedback loops may emerge between ecological and social processes.
This approach better reflects the complexity of environmental systems, where ecological, economic and social dynamics are closely intertwined.
By explicitly mapping these relationships, the causal network helps analysts understand where indicators should be placed to capture key system dynamics.
The causal network approach proposed by Niemeijer and de Groot provides a framework for achieving this.
Figure 1: Taken from Niemeijer and de Groot (2008). A structured method for depicting causal networks
Building a Causal Network
Niemeijer and de Groot propose a structured method for constructing causal networks before selecting indicators.
Step 1 — Define the Issue
The issue must be clearly defined and bounded. In many ways, this step corresponds to defining the research question.
For example, the Figure 1 is taken from Niemeijer and de Groot where the clearly defined issue in question is “ecological impacts of nitrogen fertilization on surface water ecosystems adjacent to agricultural land” (in the Netherlands).
Keeping the scope relatively narrow helps keep the causal network manageable.
Step 2 — Define Scope
The next step is to define the boundaries of the system or the scope of the study being studied.
This includes specifying:
- the ecological system under consideration
- climatic conditions
- socio-economic context
- spatial scale
- temporal scale
Boundary conditions also determine what is excluded from the analysis. For example, a study may focus on nitrogen impacts within agricultural landscapes while excluding upstream processes such as fertilizer manufacturing. In the example, the study considers in situ temperate climate conditions for the Netherlands.
Step 3 — Identify potential Indicators
Before selecting measurable indicators, the analysis begins by identifying conceptual variables that represent key processes in the system.
Examples might include:
- nutrient load
- species composition
- crop production intensity
- fertilizer application rates
At this stage, the focus is on capturing the major processes and relationships relevant to the issue.
Step 4 — Construct the Causal Network
Once the conceptual variables are identified, they can be organized into a network of cause–effect relationships.
Niemeijer and de Groot suggest grouping variables into three broad domains:
Environment-related variables such as soil, water, air and biodiversity.
Society-related variables such as policy decisions, market demand, agricultural practices or consumer behaviour.
Pressure interface where societal activities interact directly with environmental systems. Agricultural production using nitrogen fertilizer is one such interface.
The variables are then connected through directional links representing causal relationships.
Rather than forming a single chain, the variables form a network of interacting pathways.
How the Causal Network Guides Indicator Selection
Once the causal network has been constructed, indicator selection becomes a more systematic process.
Several practical considerations must then be addressed:
- the specific research question
- availability of data
- spatial and temporal scale
- monitoring capacity
- the analytical perspective being used (for example, environmental or socio-economic)
Key nodes within the causal network can then be identified. Indicators are selected to represent these nodes in a way that captures the most important processes within the system.
The starting point remains the research question. Without a clearly defined question, indicator selection risks becoming arbitrary or driven by data availability rather than conceptual relevance.
A short note on Nodes
Nodes are points in the network that are centers of activity. They can be root nodes, central nodes or end-of-chain nodes.
End-of-chain nodes have multiple incoming arrows showing long range of causes and effects. Take for example Water fauna.
Central nodes are those that typically have multiple incoming and outgoing arrows. In this example, they can be N concentrations in water or N emissions.
Root nodes – there are not so many here, but one could imagine that manure use or fertilizer use is a root node in this example. They have multiple outgoing arcs in Crop production.
In practice, not every node has a corresponding indicator. It is therefore important to identify key nodes where measurement provides most insight into system behaviour or where management interventions can influence outcomes.
Implications for Corporate Sustainability
The causal network approach also has implications beyond ecological research.
In corporate sustainability contexts, indicators are often selected to satisfy requirements for internal audits, environmental management systems or sustainability reporting. Causal networks see companies as inherently nested in physical, ecological systems where risks from loss of ecosystems are real.
However, without a clear conceptual foundation, these indicators may fail to reflect the underlying environmental processes associated with a company’s activities.
Applying a causal network approach requires that organizations:
- define system boundaries clearly
- articulate causal relationships between activities and environmental outcomes
- justify indicator selection based on conceptual relevance rather than convenience
- connect ecological indicators with social and economic decision-making
This approach increases transparency and improves the credibility of environmental reporting.
Indicators should therefore not be treated as decorative elements in sustainability reports. They are analytical tools that shape how environmental issues are understood and addressed.
Robust indicator selection strengthens:
- scientific credibility
- decision relevance
- governance transparency
- policy coherence
For example, a food processing company assessing water-related risks may construct a causal network linking sourcing of agricultural raw materials, nutrient runoff from processing, watershed conditions around the factory, and ecosystem impacts such as eutrophication.
Ultimately, indicators influence not only how environmental change is measured, but also how it is interpreted and managed.
In our next article, we look at indicators in different corporate contexts to see how they were chosen and find the causal network. We also discuss how indicators enable decision-making in corporate contexts.
However, environmental systems rarely follow a simple linear pathway. Activities may produce multiple environmental pressures. Ecological changes may be caused by several interacting causes, or feedback loops, between ecological processes and social processes. Processes occurring in one part of the system can have influence over others across space and time.
A single causal chain may therefore overlook important relationships or even fail to capture the complexity of the entire system and/or consequences associated with an activity.
The Causal Network Concept
A causal network brings an issue into perspective through variables that can influence it. This shows the environmental issue as a network of interacting pathways where
- nodes represent conceptual variables or indicators
- arrows represent causal relationships between them
Multiple causal chains intersect within the network. Variables may influence several other variables simultaneously, and feedback loops may emerge between ecological and social processes.
This approach better reflects the complexity of environmental systems, where ecological, economic and social dynamics are closely intertwined.
By explicitly mapping these relationships, the causal network helps analysts understand where indicators should be placed to capture key system dynamics.
Demystifying Indicators Part 2
From Causal Chains to Causal Networks
In Part 1, we explored a basic but essential question – What is an indicator? We introduced the idea of the causal chain as a starting point for thinking about indicators and capturing cause-effect relationships in environmental systems.
“Indicators are decision tools that exist to simplify complexity.”
In this article, we take a step deeper, exploring the Causal Network concept, as described by Niemeijer and de Groot (2006). The causal network creates a system for choosing indicators, clarifies what indicators represent, and guides the analysis of complex environmental systems.
The causal network sets an explicit request at the outset. Mapping the processes that shape the environmental issue that is particular to the organisation. Once the whole system is mapped, indicator selection becomes more purposeful, answering the question that is based on the scope of the study.
Purpose of environmental indicators
Environmental indicators reveal trends or phenomena that are not immediately detectable. They simplify complex ecological processes into signals that can be monitored and interpreted over time.
Dale and Beyler (2001) describe indicators as tools that capture aspects of ecological systems while remaining practical for monitoring and management.
Dale and Beyeler are of the opinion that indicator selection should be guided by the concept of ecological integrity. Imagine seeing it as a hierarchy
and that indicators should ideally capture processes across hierarchies to reflect structural, functional and composition of ecosystems.
Ecological indicators, which are indicators specifically representing ecosystems should do the following, according to Dale and Beyler:
- be relatively easy to measure
- be sensitive to stress on the system
- respond to stress in predictable ways
- provide early warning signals of change
- allow management actions to prevent undesirable outcomes
- integrate information about system conditions
- have known responses to disturbances, including anthropogenic pressures
- show low variability unrelated to environmental change
Why do we need causal networks?
Environmental analysis begins with a causal chain, describing cause and effect and linking human/ industry activities to an environmental outcome/ consequence. An example of this can be seen in eutrophication processes in freshwater ecosystems.
Fertilizer use à nutrient runoff à increases nitrogen content in lakes à algal blooms à organic matter decomposition à decline in oxygen content à decline in aquatic species
This clarifies cause and effect in a linear fashion.
groundwork studio
Groundwork is an independent research studio analysing nature- and climate-related risks to economies, organisations and communities.
We combine rigorous analysis, practical tools and cross-sector insight to support decision-making in a rapidly changing world.