Concept/Systems Thinking/No. 1011
System Boundary
A system boundary is the line a model draws between a system and its environment. In systems thinking, it sets what is included and which flows cross the edge. Ludwig von Bertalanffy’s work on open systems helped show how this choice shapes explanations.
- Evidence
- Well established
- Read
- 6 min
- Links
- 12 connections
01You've seen this when…
- in life
You budget 25 minutes for your commute because that’s how long the train takes. Walking to the station, waiting and reaching your desk bring it to 45.
- at work
Your team cuts handling time by moving checks into a customer form. The dashboard counts staff minutes; customers spend longer completing the paperwork.
- out in the world
A city approves a waste facility just beyond its border. Its own streets get fewer garbage trucks, while residents in the neighboring town get more.
02The idea
Every model leaves things out. A household budget includes rent and groceries. A delivery model includes warehouses, drivers and packages. A flood model includes some stretch of river and some period of rainfall. The system boundary marks the edge of that selection.
Inside the boundary are the things the model represents directly. Outside is its environment: everything else that can affect the system or receive its effects. Money, materials, energy and information can cross that edge. Drawing the boundary includes deciding which crossings to track.
A useful boundary answers a particular question. For scheduling a train, station-to-station travel time may be enough. For deciding whether to take a job, door-to-door travel time matters. For assessing the commute’s environmental impact, fuel production and vehicle manufacturing may also matter.
Boundaries can be physical, organizational or temporal. A factory fence, a company’s legal ownership and a five-year accounting period each create different exclusions. They can also concern people: whose effort, health or inconvenience enters the calculation?
This is one of the practical consequences of abstraction. Simplification makes a problem manageable. The boundary determines which simplifications the answer rests on.
03Why it matters
Changing the boundary can change the verdict while leaving the underlying events unchanged. An outsourced process disappears from one company’s emissions total and appears in another’s. A faster appointment system saves clinic time while increasing patients’ waiting time at home.
Four consequences deserve attention:
- Costs can cross the edge. A decision can look efficient when its burdens fall on people outside the calculation. Economists call uncompensated effects on others externalities.
- Causes can sit outside the frame. A warehouse stock shortage may begin with a supplier’s batching policy. Studying the warehouse alone limits which explanations become visible.
- Feedback can disappear. Cutting maintenance improves this quarter’s spending figure. Failures arrive later. A short time boundary can hide the feedback loop connecting today’s savings to tomorrow’s repairs.
- Responsibility can become blurred. A manager may control only one department while influencing customers, suppliers and neighboring communities. Those groups can belong in an impact assessment even when they sit beyond the manager’s authority.
The danger grows when the boundary remains implicit. Readers see a clean result and supply their own assumptions about what it covers.
04A worked example
In 2008, Timothy Searchinger and colleagues examined the climate effects of using crops for fuel. Their question extended beyond growing corn and producing ethanol: what happens when fuel production displaces food production and encourages new farmland elsewhere?
What it looks like The comparison discussed in their paper gives corn ethanol a 20% greenhouse gas saving relative to gasoline when indirect land-use change is excluded. That accounting covers the fuel’s production and use while leaving out a wider agricultural response.
What’s actually going on Demand for fuel crops can alter agricultural markets. If farming expands elsewhere, converting forests or grasslands releases stored carbon. Searchinger’s team used a worldwide agricultural model to estimate this effect. Under their assumptions, including land-use change made corn ethanol’s emissions nearly double gasoline’s over a 30-year period. Expanding the boundary changed the sign of the estimated benefit.
What would have helped Showing the narrower and broader calculations together, naming the land-use assumptions, and using sensitivity analysis to test how strongly the conclusion depended on them. An emissions claim needs enough scope information for readers to understand what has been counted.
The larger estimate was a model result, with uncertainty about yields, market responses and land conversion. It does not establish one emissions figure for every biofuel. The case shows why effects beyond the immediate production chain can be decisive—and why including those effects also introduces assumptions that need scrutiny.
05Where people trip up
- The reporting unit becomes the model’s edge. Departmental accounts are convenient, so the analysis stops at the department. Before accepting an improvement, follow one affected customer, employee or product across the handoff. Check whether work or risk has moved with it.
- The clock stops before consequences arrive. A project breaks even within the review period but leaves cleanup, replacement or support costs afterward. Choose a time horizon that captures the consequences relevant to the decision. Record what remains beyond it.
- An outside factor is treated as fixed. A price, demand level or supplier response enters the model as a constant, even though the proposed decision might change it. This is where partial versus general equilibrium becomes useful: a wider analysis follows how other parts of the economy respond.
- The boundary changes during the comparison. One option is assessed from purchase to disposal; another is assessed only during use. Put both options on the same footing. Compare their included stages, affected groups and time periods before comparing totals.
A quick check is to write a boundary statement before doing the calculation: the question being answered, what is inside, the time horizon, and the important exchanges with the outside. Then ask someone affected by the decision to identify the largest omission.
Keep the distinction between a model and the world visible. Map versus territory applies here: the edge of the diagram belongs to the representation.
06Where it doesn’t help to keep expanding
A boundary that includes everything defeats the purpose of modeling. More scope brings more data needs, uncertain relationships and opportunities for error. A local air-quality question can reasonably focus on local exposure, even when a separate climate assessment needs a wider frame.
Expand the boundary when an omitted consequence could change the decision, a dependency could change the forecast, or an affected group could change the assessment of fairness. Otherwise, document the exclusion and proceed.
Simple models can handle important crossings as inputs and outputs. The practical goal is enough scope to answer the question reliably.
07Roots
In a 1950 paper, Austrian biologist Ludwig von Bertalanffy examined a puzzle at the meeting point of biology and physics. A living organism maintains its organization while continually exchanging material with its surroundings. A cell takes in nutrients and releases waste; its contents keep changing while recognizable structure persists.
Physics already used boundaries to define systems. Bertalanffy helped make open systems central to a broader systems theory: understanding an organism required accounting for exchanges across its boundary. The question was how living systems could remain in a steady state through ongoing flows.
The idea traveled into management and public decision-making. In The Systems Approach in 1968, philosopher C. West Churchman pressed questions about a system’s objectives, environment and beneficiaries. Choosing the frame became part of deciding whose problems an analysis addressed. Later systems writers, including Donella Meadows, made boundary choices a practical concern for understanding flows, feedback and policy consequences.
08How solid is this?
A standard concept in systems theory and modeling: boundaries determine which processes and effects enter an analysis. The right scope depends on the question, and broader boundaries introduce additional data needs and uncertainty.
09Connections
- Helps counter Unintended Consequences
- Part ofAbstraction, Map vs. Territory
- See also Chesterton’s Fence, Emergence, Local vs. Global Optima, Second-Order Thinking, Feedback Loops, Stocks and Flows, Externality, Sensitivity Analysis, Partial vs. General Equilibrium
+ 2 more in the list
10Origin and sources
System boundaries predate general systems theory in physical modeling. Ludwig von Bertalanffy’s work on open systems, including his 1950 paper and 1968 synthesis, helped establish their importance in general systems theory.
- [1]von Bertalanffy, L. (1950). The Theory of Open Systems in Physics and Biology. Science, 111(2872), 23–29.
- [2]Churchman, C. W. (1968). The Systems Approach. Delacorte Press.
- [3]Meadows, D. H. (2008). Thinking in Systems: A Primer. Chelsea Green Publishing.
- [4]Searchinger, T., Heimlich, R., Houghton, R. A., Dong, F., Elobeid, A., Fabiosa, J., Tokgoz, S., Hayes, D., & Yu, T.-H. (2008). Use of U.S. Croplands for Biofuels Increases Greenhouse Gases Through Emissions from Land-Use Change. Science, 319(5867), 1238–1240.
Suggest an edit· Updated 2026-10-02