Private growth sounds private.
A company expands.
A factory opens.
A data center gets built.
Investment arrives. Jobs are promised. Economic activity increases.
But large-scale private growth rarely remains private for long.
Eventually, it reaches the systems underneath it.
Electricity.
Water.
Roads.
Transmission.
Emergency services.
Housing.
Public administration.
And once growth requires those shared systems to expand, a different question appears:
Who pays for the infrastructure that makes private growth possible?
This is why private growth keeps creating public infrastructure costs.
Not necessarily because corporations are cheating.
Because private expansion and shared infrastructure operate under different economic rules.
The Part of Growth We Usually Don’t Count
When a large company announces a new project, the visible investment is easy to measure.
There is a construction budget.
There are jobs.
There may be new buildings, equipment, payroll, and tax revenue.
But the project also creates new demands on systems that already exist.
A large industrial facility may require more road capacity.
A major housing development may require expanded water and sewer systems.
A hyperscale data center may require enormous amounts of electricity, new generation, transmission upgrades, substations, or other grid infrastructure.
The investment therefore has two sides:
the infrastructure the company builds for itself, and the infrastructure everyone else must expand around it.
Those are not always financed the same way.
The AI Data Center Makes the Mechanism Easy to See
AI provides an unusually clear example because computing infrastructure can concentrate enormous electricity demand in a single location.
The company may build the data center.
But electricity does not appear at the property line by magic.
Power must be generated.
It must be transmitted.
Local infrastructure must be capable of delivering it.
The wider grid must remain reliable while supplying everyone who was already connected.
Suddenly, what looked like a private construction project becomes a shared infrastructure question.
The important issue is not whether AI is good or bad.
It is not whether data centers should exist.
The structural question is simpler:
When one participant creates an unusually large new requirement for a shared system, how much of the resulting cost should that participant absorb?
The Cost-Allocation Problem
Shared infrastructure creates an unusual economic problem.
Once built, infrastructure often benefits more than one user.
A stronger electrical grid may serve a data center, but it may also improve capacity available to other customers.
A new road may primarily support an industrial development while remaining open to everyone.
A water-system expansion may accommodate new commercial demand while becoming part of the community’s permanent infrastructure.
That makes cost allocation complicated.
Who caused the expense?
Who benefits from the asset?
Who should finance it?
Who carries the risk if projected growth never arrives?
There is no universal answer.
But there is a predictable incentive.
Every participant would prefer someone else to absorb as much of the enabling cost as possible.
Why Costs Naturally Try to Spread
Imagine a company needs $1 billion of additional infrastructure to make a project viable.
If the company pays the entire billion dollars, the project becomes less profitable.
If a utility pays part of it and recovers the investment broadly over time, the project’s economics improve.
If government contributes through infrastructure programs, financing, land, tax treatment, or other incentives, the project’s economics improve again.
None of this automatically means something improper occurred.
It means the parties have different incentives.
The company wants the lowest possible cost of expansion.
The locality wants investment and jobs.
The utility wants sufficient capacity and a financially viable system.
Existing customers want reliable service without paying more because somebody else arrived.
Each position makes sense from inside its own role.
The conflict appears when those incentives meet.
The Five-Layer Map
Use the familiar functional hierarchy:
Deciders → Creators → Operators → Enforcers → Everyone Else
- Deciders determine which kinds of growth will be encouraged and which trade-offs are acceptable.
- Creators design tax structures, utility rules, development agreements, rate structures, and infrastructure plans.
- Operators make the expansion function in practice.
- Enforcers administer the resulting rules, rates, permits, and requirements.
- Everyone Else experiences whatever costs and benefits eventually emerge.
The critical point is that the private investment decision occurs near the top of this chain while many of its secondary effects can travel much farther down.
That creates an externalization opportunity.
When Growth Becomes Cost Externalization
There is an important distinction here.
Public infrastructure supporting private activity is not automatically a subsidy.
Societies build shared infrastructure precisely because shared systems can make everyone more productive.
The problem begins when the economics become asymmetric:
the private participant captures a disproportionate share of the upside while unrelated participants absorb a disproportionate share of the enabling cost or downside risk.
That can happen through higher rates.
It can happen through taxes.
It can happen through public financing.
It can happen when infrastructure is built for projected demand that later disappears.
It can also happen through opportunity cost: capacity, land, water, capital, or administrative attention committed to one use cannot simultaneously be committed somewhere else.
The cost does not have to arrive as a bill labeled:
Corporate Expansion Fee.
Architecture is usually subtler than that.
Why the Current AI Debate Is So Useful
The rapid expansion of AI infrastructure is forcing this normally obscure question into view.
Data centers can be unusually large electricity customers.
That means utilities, regulators, governments, and technology companies have to decide explicitly who pays for new generation and grid upgrades—and who carries the risk if infrastructure is built around demand forecasts that later change.
That debate is valuable because it exposes the underlying mechanic.
The issue isn’t merely electricity.
It’s cost allocation inside shared systems.
And the same architecture appears elsewhere.
Why “Economic Development” Doesn’t Settle the Question
Large investments can create real benefits.
Jobs matter.
Tax revenue matters.
New infrastructure can benefit communities.
Technological development can produce enormous downstream value.
None of that answers the cost-allocation question.
“This creates economic growth” and “someone else should pay part of the enabling cost” are separate propositions.
They are frequently bundled together.
Structural literacy requires pulling them apart.
A project can be socially valuable while still having a poorly designed cost structure.
A public investment can be justified while still disproportionately benefiting private interests.
A private company can legitimately need shared infrastructure without automatically being entitled to have everyone else finance it.
Those possibilities can all be true at once.
The Clarifying Insight
Private growth keeps creating public infrastructure costs because large-scale growth eventually encounters systems no company owns by itself.
At that boundary, private ambition meets shared capacity.
Then the real negotiation begins:
Who captures the upside?
Who pays for the capacity?
Who carries the risk?
Those three questions tell you considerably more than the ribbon-cutting ceremony.
The point is not to oppose growth.
It is to see the entire transaction.
Want the larger map?
The free Vampire Playbook explains how costs, risks, rewards, and consequences move through modern systems—and why they so often land in different places.