Systems Dimensions: Throughput, Membership, and Conflict

This is post 5 of 23 in a series on Systems Thinking: Managing Chaos and Complexity by Jamshid Gharajedaghi (ISBN: 978-0-7506-7973-2).

In the last post, we wrapped up systems principles with emergent properties. Now we’re into Chapter 3, where Gharajedaghi breaks down the five dimensions of any social system. This post covers the first three: throughput, membership, and conflict management.

The Five Dimensions

Before getting into the details, here’s the big picture. Gharajedaghi says every social system operates across five dimensions:

1. Wealth - producing goods and services, distributing them fairly
2. Truth - generating and spreading knowledge
3. Beauty - the emotional side, the meaningfulness of what people do
4. Values - the rules and norms that govern relationships
5. Power - governance, authority, legitimacy

Different thinkers have picked their favorite and called it the root cause of everything. Marx said economy. Weber said power. Religious thinkers point to values. Gharajedaghi says you need all five. They interact. You can’t understand one without the others.

There’s a fun story about how Gharajedaghi and Ackoff argued for days about whether “power” and “beauty” belonged on the list. Ackoff had missed power. Gharajedaghi had missed beauty. Each couldn’t see the thing they were swimming in. As Gharajedaghi puts it: “Fish don’t realize that something called water exists.”

Throughput: How the Work Actually Gets Done

Throughput is the business dimension. It’s the chain of activities that turns inputs into outputs and gets them to people who need them.

For a company, that means marketing, selling, processing orders, purchasing, producing, shipping, and accounting. For a school, it’s selecting students, scheduling courses, teaching, and issuing certifications. Different context, same idea.

Gharajedaghi lays out four elements for designing a good throughput system.

The Process Model

You need a model of how the process works. Not just a flowchart, though that’s a start. Something that captures the dynamics: time lags, bottlenecks, feedback loops, buffers and queues.

He uses a telephone company as an example. Map out the key variables: customer base, installations, repairs, preventive maintenance, plant capacity. Then connect them to show how they affect each other. More customers means more installations. More installations means more repairs. More repairs eat into resources for new installations.

The point is to see these relationships before they surprise you.

Critical Properties

Every throughput process has critical properties you need to manage: time, cost, flexibility, and quality.

Here’s the catch. These properties are interdependent. You can cut costs by sacrificing quality. You can speed things up by throwing more money at it. But if you only focus on one, you’ll mess up the rest. The goal is to improve all four at the same time.

Measurement and Diagnostics

You need to monitor all critical variables together, in real time, in the same view. Not separate reports for separate departments. One picture.

Gharajedaghi makes a great point about plateaus. When your improvement efforts stop producing results, the current design has maxed out. You’ve used up all the slack. Tweaking won’t help. You need to redesign the whole process.

I’ve seen this happen a lot. Teams squeeze a process that’s fundamentally limited until they’re exhausted. The answer is to step back and rethink.

Read-Only Memory

Simple but smart. Record every assumption, expectation, and design change in a log that can’t be altered. Not so you can blame people. So you can learn.

When something goes wrong six months from now, you want to see what you were thinking when you made that decision. If people edit the history, the learning disappears.

Target Costing

This flips traditional pricing on its head. Old way: add up your costs, tack on a margin, that’s your price. The assumption is costs are fixed and prices are what you control.

Target costing says the opposite. The market sets the price. So you work backwards. Start with what customers will pay, subtract your margin, and that’s your cost target. Design the product to hit that number.

The Lexus story makes this real. When Toyota decided to compete with Mercedes-Benz, they figured out about $20,000 of a Mercedes’ $65,000 price tag was brand premium. They targeted $40,000 for the Lexus. After subtracting distribution costs and profit margin, the design team got a $26,000 target cost. The message: build it for this amount, or step aside.

Cost becomes a design constraint, not an afterthought.

Membership: People Aren’t Machine Parts

Machines get assembled once. You bolt the parts together and you’re done. Organizations aren’t like that. Keeping people integrated into a whole is a constant, ongoing struggle.

Gharajedaghi points to something real here. We want to be individuals, but we also desperately want to belong. An exciting project, a shared mission, a group that gets it. Beauty, he says, is the ultimate agent of social integration. Commitment depends on how much excitement and meaning the organization generates.

Key insight: the units of a social system aren’t individuals. They’re roles. The same person behaves differently in different settings. Your brilliant VP might be a terrible CEO. Your amazing friend might be a frustrating coworker. The role shapes the behavior.

This means effective membership requires three things:

1. A role - something meaningful to do
2. A sense of belonging - feeling like part of the group
3. A commitment to participate - actively helping shape the future

When someone feels like their contributions don’t matter, or when they feel powerless to make a difference, indifference sets in. They check out. They’re physically present but mentally gone.

There’s also a chain-link effect. Incompatibility among members causes the stronger ones to regress to the level of the weakest.

For membership to work, you need two background systems. An exchange system where people gain something by contributing. And a threat system where behaviors that harm the whole are not tolerated. Carrot and stick. But grounded in the idea that the organization exists to serve its members just as much as its members serve it.

Conflict Management: Stop Pretending Conflicts Go Away

In a system full of people with their own goals, conflict is guaranteed. The question isn’t how to prevent it. It’s how to deal with it.

Gharajedaghi describes four types of relationships:

• Cooperation - same goals, same methods
• Competition - different methods, same higher goal
• Coalition - different goals, temporarily working together against a common obstacle
• Conflict - different goals, each side reducing value for the other

And four ways to handle conflict:

• Solve it - pick a winner. Someone wins, someone loses.
• Resolve it - compromise. Nobody’s thrilled, but everyone can live with it.
• Absolve it - ignore it and hope it goes away. (Spoiler: it rarely does.)
• Dissolve it - redesign the system so the conflict no longer exists.

Dissolving is the systems thinking move. Instead of picking sides or splitting the difference, you change the game entirely. You find a frame where opposing forces become complementary.

From Lose/Lose to Win/Win

Gharajedaghi says that in today’s complex world, it’s easier for groups to prevent each other from winning than to actually win themselves. The result is a lose/lose environment where everyone blocks everyone else and nobody moves forward.

But once people realize they’re in a lose/lose situation, that awareness itself can shift the game toward win/win. It’s the prisoner’s dilemma logic. When players understand that trying to screw the other side will screw themselves too, cooperation suddenly makes sense.

Turning Conflict into Competition

Remember, competition is just a conflict of means, not ends. Both sides want the same thing but disagree on how to get there.

So the move is: find a higher-level goal that both sides share. Their conflicting ends become competing means under that shared objective. You haven’t eliminated the disagreement. You’ve reframed it so both sides push toward the same destination through different paths.

If you can’t find a shared objective even at the highest level, the conflict can’t be dissolved within your current worldview. You need a completely new way of thinking.

What I Take from This

The throughput section is solid practical thinking. Target costing alone is worth the read. But the membership and conflict sections hit harder because they’re about people. And people are where systems get messy.

The biggest takeaway for me: conflict isn’t a bug, it’s a feature. Trying to eliminate it is a waste of time. The real skill is learning to dissolve it by finding higher ground where opposing forces become complementary. That’s not compromise. That’s design.

Previous: Systems Principles - Emergent Properties

Next: Systems Dimensions - Decision Systems and Learning