Paradigm S Curves

by John Moser

People work in a mindset commonly explained by fear of change, resistance to change, or appeal to tradition:  the new and unexpected disturbs people, and so they fight it.  Biologically, integrating new knowledge requires a lot of mental energy; throwing an emotional tantrum when something conflicts with what you already know employs considerably less effort.  It’s a useful trait, and only burdens us with resistance to new facts and corrections to our misconceptions as an unfortunate side-effect.

This provides larger challenges when both sides have correct information.  Situations change:  what’s correct in one situation fails miserably in another, and people rely on what’s worked for 30 or 40 years instead of seeing the imminent failure ahead.

We call this phenomena paradigm, and it behaves in the simple model of an S-curve:

Paradigm S-curve

Paradigm S-Curve.  From Dealing With Dilemmas by Frank Buytendijk.

This curve shows time or cost investment on the horizontal, while the vertical shows productive output.  It illustrates a prior process or paradigm providing steady return for additional invested cost, right up until the same amount of input starts producing less output.

For example:  you may need 1 new worker to produce 1,000 widgets per hour, up to about 9,000,000 widgets; once you reach 10,000,000 widgets per hour, you’re now employing need 1 new worker per 500 per hour; and at 20,000,000 per hour, it’s 10 new workers per 1 per hour.  It scaled linearly up to only about 9,000,000 per hour, then started requiring more and more effort to squeeze out small additional output.

In our example, the right-most curve shows that a new process to manufacture widgets would have originally produced only 7,000,000 widgets each hour with the same people producing 9,000,000 widgets per hour; and then improvements to the new process eventually allowed the same workers producing 10,000,000 widgets per hour to produce 15,000,000 widgets per hour.

Workers Old Process New Process
5,000 5,000,000 2,000,000
6,000 6,000,000 3,500,000
7,000 7,000,000 4,500,000
8,000 8,000,000 6,000,000
9,000 9,000,000 7,000,000
11,000 10,000,000 15,000,000

If this looks like some kind of logical sleight-of-hand, it’s because it is.  S-curves often illustrate two different problems simultaneously.

The way I originally explained it, you’d think our new process would perform poorly if we’re producing under 10,000,000 widgets per hour.  In reality, it’s our old process which behaves this way:  it’s mature, and works as well as it’s going to work, and it doesn’t work very well when you start getting up into numbers like 10,000,000 widgets per hour.  That process is an old, failing paradigm no longer capable of keeping up with our new demands.  This is where economic scarcity comes from.

The new paradigm requires all kinds of engineering effort.  Maybe we need advanced nanomaterials, and our processes produce a 1% yield:  we do the work to produce 100 tonnes of carbon fiber, but only about 1 tonne of that is usable.  Building widgets out of carbon fiber costs a lot more than building them out of aluminum because we’re investing 50 times the labor in making carbon fiber than we are in making aluminum.  Carbon fiber may also take longer to shape than aluminum.

When we improve the process, we get nearly 100% yield in carbon fiber manufacture.  Carbon fiber now costs half as much as aluminum—it requires half the labor—and takes less than twice as long to form into widgets than aluminum.  Our ability to produce and shape carbon fiber widgets scales up much higher than aluminum before we hit engineering bottlenecks, so we can hire more people and produce more widgets using this new carbon fiber process.

Workers Old Process New Process

(10 years ago)

New Process


5,000 5,000,000 2,500,000 7,000,000
6,000 6,000,000 3,000,000 8,000,000
7,000 7,000,000 3,500,000 9,500,000
8,000 8,000,000 4,000,000 11,000,000
9,000 9,000,000 4,500,000 12,000,000
11,000 10,000,000 5,000,000 15,000,000

You may notice that carbon fiber processing actually costs less.  Not only can we scale it up without increasing costs—which would keep prices the same, but allow more production—but we can even make more of it with less labor, bringing costs down.  The $75 widget is now a $50 widget, and the retail price can move down to reflect this.

For a person who’s worked in widget manufacture his whole life, old, aluminum widgets may be the best.  In his time, carbon fiber may have been poorly-made, expensive to work with, and of low quality; while aluminum was well-engineered, cheap to manufacture, and of consistently high quality.  At that time, carbon fiber widget processes are simply wrong:  they’re expensive and produce inferior results.  Today, carbon fiber widget processes are better than aluminum, producing higher-quality, more consistent results at lower cost; aluminum process is just inferior.

These paradigm shifts occur constantly.  They occur in everything.  Paradigm shifts apply to economics even as far as public policy:  welfare systems, minimum wages, and tax brackets aren’t just often products of inexperience, but products of necessity.  By the same token, you can’t eliminate an old system without replacing it with a new one:  it’s either unnecessary because of the modern systems in place, or it’s less-optimal than a new system which most likely wouldn’t have worked 40 years ago.

If the whole system is non-functional and you have nothing to replace it with, you have a crisis.