These days, arguments over Internet policies like universal broadband or net neutrality get all of the publicity, and they are important. But these showy sides of Internet policy may not be the most important ones. Behind the scenes there’s a bigger issue whose outcome could completely change the nature of the Internet and how you get access to content, services, and other users.

That issue is known as transit pricing.

Internet transit is a fee-based service that allows an ISP, a CDN (content delivery network), or even an enterprise to purchase the right to pass Internet traffic along a specific route. Often this is a route to a specific access point where customers or other ISPs are available. It avoids having traffic take circuitous multi-hop routes that increase delay, performance variability, and packet loss.

Recent research has shown that there are enormous variations in the transit pricing available worldwide.

According to TeleGeography’s IP Transit Pricing Service, transit prices in North America and Europe have fallen about 20 percent per year for the last three years, and pricing in all areas tends to be comparable. This is not surprising, given the advances in optics and competition. Transit prices are higher in areas like Asia, Africa, and Latin America, which is also no surprise.

What is troubling is that price reductions in transit have been about 20 percent annually for the last three years, when cost of bandwidth has fallen between 50 percent and 70 percent.

What’s probably happening is that big ISPs with transit services are holding their transit prices high to reap some profit from the Internet. That works because transit prices aren’t as visible as consumer broadband prices.

But the impacts of transit congestion sure are.

If you have 100-Mbit/s Internet and share a 100-Mbit/s transit Internet connection with a million of your friends, how much bandwidth does each of you have? With this kind of transit congestion, your actual performance at 100 Mbit/s might be no better than at 10 Mbit/s, in fact.

When you pick an ISP based on access speed, you might find yourself with one that actually gives you worse performance than a competitor because it's cut back on its transit expenditures.

Regulators don’t talk about transit bandwidth issues much these days. In the mid-to-late 1990s there was a surge of interest in making Internet peering and transit work properly, and at the time, the Federal Communications Commission (FCC) and others were looking into the issue, but no regulations came out of that. The problem now is that we’re focusing on a national broadband plan that, by definition, looks at Internet access as though it were the only factor in performance and cost of service. It’s not.

The transit pricing issue is a clear demonstration that we need a better overall understanding of the business of the Internet, and how business decisions made by ISPs and content providers will affect our experience.

The Internet’s business complexity encourages providers to play a shell game with us, hiding performance limits in one place while they give us “faster” connections elsewhere.

The FCC is, in theory, committed to basing its national broadband policy on hard statistics, but do we know what statistics they’re gathering? After all, transit pricing is an issue that was looked at by regulators and then dropped, and it now seems it was important after all.

How many other hidden issues like it threaten our Internet experience? Let’s not wait until we’re disappointed with our service to learn the answer to that one!

— Tom Nolle, software engineer and founder of CIMI Corp.