The “Free Market” Internet Myth: Why Switzerland Can Go 25 Gbit While the US Lags

A Friday night when the speed doesn’t match the bill

Picture this: you pay for “gigabit internet,” you run a speed test, and the numbers look fine—until 8 PM. Then the same household bandwidth that felt endless turns into buffering videos and stalled downloads. That mismatch between marketing and lived experience is the real story behind the puzzle: why Switzerland can offer very high-speed, symmetric, dedicated fiber connections (often cited around 25 Gbit/s) while the United States commonly delivers lower speeds, fewer choices, and network sharing that collapses during peak hours.

People often point at politics: “America believes in free markets; Switzerland regulates.” But the deeper issue isn’t “freedom vs control.” It’s what happens when an industry has a natural monopoly—a situation where competition over the infrastructure doesn’t make economic sense.

And when you get that wrong, you don’t get “more choice.” You often get less.

The core concept: what is a “natural monopoly”?

A natural monopoly is an industry where building the underlying infrastructure costs so much that it’s inefficient for multiple companies to duplicate it, while the cost of serving one additional customer is relatively low.

A classic analogy is water pipes: digging multiple sets of pipes down the same street is wildly expensive, disrupts neighborhoods, and still ends up delivering water to the same households. Electricity distribution works similarly in many places: it’s cheaper to build the grid once, then let different providers compete in how they use it.

Fiber internet fits this pattern more than most people think. Laying fiber optics involves expensive civil work (digging, permitting, trenching, restoring streets) and long-lived infrastructure planning. Meanwhile, sending data to one more customer across already-installed fiber is comparatively cheap.

So the “smart” model is usually:
- build the infrastructure once (as a shared asset),
- let services compete over top.

When a country instead competes by duplicating the infrastructure, it can waste money without improving consumer outcomes.

The Swiss approach: infrastructure as a shared backbone

In Switzerland, the story that often emerges is that fiber is treated less like a collection of isolated businesses and more like a shared backbone. In that model, multiple internet service providers (ISPs) can offer plans to end users while using the same underlying fiber access path.

Two technical details matter here:

1) Symmetric speed

Symmetric means download and upload speeds are the same (or close). If a plan advertises 25 Gbit/s, symmetric means uploads can also reach that ceiling—not just downloads.

That’s crucial for modern workloads: video calls, cloud backups, and hosting content all need strong upstream capacity.

2) Dedicated vs shared capacity

A dedicated connection (or dedicated capacity along the access segment) means your household’s throughput isn’t directly divided with many neighbors on the same portion of the network.

A shared connection means your “gigabit” (or “25”) is a pool used by multiple homes, typically using a shared architecture like point-to-multipoint (more on that below).

When capacity is shared, peak-time congestion becomes predictable. It’s not a mystery; it’s the consequence of how bandwidth is allocated.

Why “competition” can backfire: overbuild and wasted trenches

Now let’s talk about the version of competition that sounds good on paper: allow multiple companies to dig, build, and deploy their own fiber everywhere.

In Germany’s described model, that leads to overbuild—multiple fiber networks running in parallel, sometimes close enough that they feel like redundant lines across the same physical space.

Overbuild tends to create a paradox:
- It increases the number of networks.
- It also increases construction costs, permitting complexity, and street disruption.
- But it doesn’t necessarily increase effective consumer choice, because the expensive duplication can slow down coverage and upgrades.

Think of it like three grocery delivery services each insisting on opening their own warehouse in every neighborhood street. You might get more “options,” but you also get higher costs, slower rollout, and still the same delivery bottlenecks.

In infrastructure terms, overbuild is a costly way to “buy competition” when the real efficiency comes from sharing the hard parts.

The American pattern: protected territories and shared access

The United States often looks like a free-market champion—lots of companies, lots of brands, lots of marketing.

But the described on-the-ground reality frequently becomes territorial monopolies: in a given neighborhood, one incumbent provider dominates because it installed (or acquired) the physical network first. Other providers then struggle to enter because duplicating fiber access is expensive and slow.

When you hear “competition” but there’s no real choice between fiber providers, that situation behaves like a cartel. Consumers can’t easily switch providers for better price, performance, or service terms.

The peak-time problem: shared architecture

A big reason advertised speeds may disappoint is shared architecture.

One common design concept is point-to-multipoint (P2MP)—a network layout where one sending point serves multiple receiving points. In plain language, it’s like one microwave tower broadcasting to multiple households: the more households share the same transmission resources, the more contention appears when everyone transmits at once.

So when the network is engineered with shared segments, your plan’s “up to 1 Gbit/s” can collapse under evening load.

This is why you can see behavior like: at 8 PM the same connection that “tested” fast in the afternoon becomes much slower in practice. It’s not always a failure of the provider; it’s the design tradeoff they used to deliver service profitably.

The other barrier: private bottlenecks

Even when a competitor wants to enter, they often hit a physical reality: the fiber terminates at a point of presence (PoP)—a central hub where network lines from many homes converge.

If that PoP is controlled by the incumbent, the competitor can face a coordination wall: permission, access arrangements, and installation constraints. In effect, the network’s “front door” can be locked.

Competition becomes less about “free markets” and more about who controls the critical node.

The tricky truth: both “more regulation” and “less regulation” can fail

The most confusing part of the story is that regulatory philosophy alone doesn’t decide the outcome.

• Germany can be “over-regulated,” yet still organize competition in a way that produces overbuild and inefficiency.
• The United States can be “deregulated” in spirit, yet still produce monopolistic outcomes through the natural monopoly dynamics of fiber.
• Switzerland can be regulated, but the regulation can be aligned with efficiency—enabling shared access to infrastructure rather than forcing duplicate construction.

So the real variable is not how much regulation exists. It’s what the rules optimize:
- Do they push for duplication of infrastructure? (Often wasteful.)
- Or do they enforce open/shared access so multiple service providers can compete without duplicating the streetscape?

That’s the difference between “regulated competition” and “competition as duplication.”

A mental model that ties it together

Here’s a way to visualize the infrastructure economy behind these outcomes.

Step 1: The expensive layer

Trenching, ducts, rights-of-way, permitting, restoration—this is the expensive layer.

Because it’s expensive, it naturally tends toward one-time investment. That’s the natural monopoly.

Step 2: The cheap layer

Once fiber is in place, running additional traffic and offering service plans to customers is comparatively cheaper.

That’s where competition can thrive: multiple ISPs can compete on pricing, packaging, and customer experience.

When a country lets companies fight for the expensive layer, you get overbuild, delays, and congestion between rivals. When a country enables competition at the cheaper layer, you get more rapid rollout and better performance.

Switzerland’s “hyper-speed at reasonable prices” narrative makes sense under this lens: service competition grows on top of shared fiber access.

The US narrative—lower speeds, fewer providers, and peak-time sharing—often makes sense under a different engineering and market structure: protected territories plus shared capacity at the access level.

Why speed numbers sound like marketing but behave like engineering

People often argue about the exact advertised figure—25 Gbit/s here, 1 Gbit/s there. But the important distinction is how throughput is provisioned.

A fast plan on paper can still behave poorly if:
- capacity is shared among many households,
- upstream and downstream are balanced differently than users expect,
- or the network has insufficient headroom during peak demand.

Conversely, an extremely high plan can feel smooth if the system is engineered for dedicated access and symmetric bandwidth.

That’s why questions like “Why does my gigabit slow down at night?” are fundamentally engineering questions—because the “gigabit” is only guaranteed under certain network conditions.

And those conditions are shaped by policy decisions made decades earlier: who was allowed to build where, when, and under what access rules.

Conclusion: the internet isn’t a normal market

The puzzle of Switzerland versus the United States isn’t best explained by a simple slogan like “free markets win” or “regulation works.” It’s explained by the structure of the business: fiber access is a natural monopoly, and the country that organizes competition around sharing the expensive infrastructure can achieve faster rollout and higher practical performance.

When competition targets the costly backbone (overbuild) or gets blocked by controlled bottlenecks and territorial incumbency, the result can be fewer real choices and shared capacity that degrades during peak hours.

In short: the internet’s speed story is a story about incentives, infrastructure economics, and how rules determine where competition happens—not about whether capitalism is “pure” or “constrained.”