The telecom world is getting stranger every day. The rates it advertises for long-distance calls are getting smaller and smaller, but the prices that telecommunications companies are paying to buy each other keep getting bigger and bigger.

First Sprint, the nation's third-largest long-distance company, advertises nickel nights, setting off a round of price cuts. Then last week, MCI WorldCom, the No. 2 long-distance company, turns around and agrees to buy Sprint in the biggest takeover in history: $115 billion. Or, if you prefer, 2.3 trillion minutes of nickel calls. If all goes as planned, the deal will trigger the biggest single management payday in history, with Sprint Chairman William Esrey picking up about 13.5 billion nickels, or $673 million for his stock and options.

Meanwhile, your bill for local phone service is getting as long as your phone book, filling up with incomprehensible, ever-rising fees--including two little jewels known to bill mavens as Pixies and Slicks--that can more than double your basic bill. So your stated phone rate stays the same, but your bill keeps going up. As for long distance, forget about it. You need a spreadsheet program, 16 databases and a doctorate in numerology to figure out which plan is best for you.

Which raises the following questions: If talk is cheap and getting cheaper, why did Sprint fetch the highest takeover price since God created money? Why can't you understand your phone bills? And what are all those telecom regulatory hearings, lawsuits and takeovers about? It's Confusion Central--almost enough to make you wish that Ma Bell, the old AT&T monopoly, were still around. At least then, you had some idea of what was going on--or thought you did.

What's causing all this turmoil and churn? In a word, competition, much of it an unintended result of the Telecommunications Act of 1996. That act was supposed to unleash the Baby Bells and the big long-distance companies to raid each other's markets. That hasn't happened. What has happened is that a host of new competitors, armed with deep pockets and spiffy new technologies, are competing for telecom business. And telecom business, which includes much of the Internet, is an industry worth hundreds of billions of dollars and is growing faster than a teenager's phone habit.

The game here is to stake a claim on delivering "broadband" communications of phone service, long distance, wireless, data, Internet access, cable TV, interactive movies and anything you can shove into someone's home or office with a wire or a wireless signal. Instead of settlers in wagons, this land rush consists of companies trying to lock up access to customers. Everyone is so desperate--and the technological future uncertain--that what once was a crummy cable company called Tele-Communications Inc. and a mediocre long-distance company called MCI became "vital strategic assets" that AT&T and WorldCom decided they had to buy at almost any price. AT&T is betting that its newly acquired cable-TV properties will become a digital superhighway into millions of homes.

Will it work? Who knows? MCI WorldCom wants Sprint's huge wireless phone business so it can offer the full array of broadband services, and is betting on a separate piece of Sprint technology called Multichannel Multipoint Distribution Service that, in theory at least, will allow WorldCom to reach millions of customers without having to lay wires. Qwest and WorldCom paid high prices for not terribly wonderful outfits like US West and Frontier, the former Rochester Telephone. Why bother with them? Because that way, Qwest and WorldCom will have traffic to route over the fiber highways they're busy building.

The biggest single cause of all this turmoil is the advent of "packet switching" technology, which allows a single circuit to carry more than 100 conversations at once. With the old technology, "circuit switching," a single conversation took up an entire circuit. By increasing capacity a zillionfold, the new technology has made long distance incredibly cheap. The betting is that this huge surge in capacity will let companies sell long distance the way they now sell Internet service: a sizable monthly fee for access, but little or nothing for each use. Good news for heavy users, whose total bills will drop. Bad news for light users.

America being what it is, these amazing changes have allowed a whole lot of people to pile up huge fortunes in a virtual eye blink. These include folks that most people outside Wall Street or the telecom business have never heard of: Gary Winnick of Los Angeles, Phil Anschutz of Denver, Bernie Ebbers of Jackson, Miss. They're the hitters behind Global Crossing, Qwest and WorldCom, respectively. Their secret? That if you can get a high-enough stock price and access to new technology, being a newbie can be lots more fun than being an old incumbent company such as a Baby Bell, AT&T or MCI. A company almost no one has heard of--RCN--has raised almost $5 billion in its two-year existence, including a splashy $1.65 billion put up last week by America's second-richest person, Paul Allen, for a 22 percent stake. While the Winnicks, Anschutzes and Ebberses have made fiber fortunes, plenty of individual investors have also done very nicely. They include my wife and me, whose holdings in Global Crossing, RCN and a company related to RCN totaled $51,000 at the end of last week, a significant investment by our standards.

But while the telecom business is changing at Internet speed, the regulatory system is still based on the century-old phone monopoly that existed before AT&T broke up into separate long-distance and local-service companies in 1984. Lush long-distance profits subsidized local rates. Why not? AT&T owned both the long-distance and the local businesses.

In the post-Ma Bell world, there's a much more complicated system designed to accomplish the same thing, along with having urban and suburban customers subsidize rural customers and having businesses subsidize everybody. It's very well intentioned and is based on the idea that phone service is something that should be accessible to everyone. But that's all starting to crumble now, because distinctions between local- and long-distance service have blurred, and companies are building their own lines to individual customers in lucrative markets, allowing them to bypass the fees that regulators have given local companies permission to collect.

"The difference between local and long-distance is a regulatory effect," says John Petrillo, AT&T's top strategist. Phone calls "don't know when they cross a regulatory border," he adds. "We're all changing quickly into 'all-distance' companies that operate across these borders."

"The regulatory system is based on 'Alice in Wonderland' economics, crafted in the age of monopolies," adds John Nakahata, a former chief of staff at the Federal Communications Commission who is now a partner in the Washington law firm of Harris, Wiltshire & Grannis. "But in the age of competition, you have to get out of Wonderland and into the real world."

The FCC is making a valiant attempt to hold the current system together, as it's legally charged to do. But it's like trying to stuff a 42-inch waist into 36-inch shorts. It kind of works, if you don't look too close, but there are still all sorts of unsightly bulges.

The underlying problem: Even though local phone companies make nice profits, basic local-phone service is wildly unprofitable. Local companies' profits come from the access charges that long-distance carriers pay them, the FCC-mandated fees that customers pay them and lucrative offerings like call waiting, voice mail and caller ID. This is bad economics but good politics, because no one wants to be responsible for raising local rates. Hence two FCC-mandated charges--the Slick and the Pixie--that put money in local companies' pockets but that no one has to take the blame for. The Slick--or SLIC, which stands for Subscriber Line Charge--and the Pixie--or PICC, for Presubscribed Interexchange Carrier Charge--total $4.54 a month in a one-line household, and more than $13 for two lines. That keeps the stated basic rate low, lets the company blame the FCC for the charge and lets the FCC tell irate customers that the federal government doesn't get any of the money. Anyone want to play pass the buck?

Then there's the Universal Service Fund, collected by long-distance companies to subsidize "high-cost" phone systems, mostly rural. Some carriers charge a flat fee, some charge a percentage of the bill. This fee, like the Pixie, used to be hidden. But last year, after the fee was jacked up to include a new subsidy to help schools, libraries and rural hospitals connect to the Internet, AT&T decided to break the charge out on its bills. While it was at it, AT&T also decided to charge customers for other USF fees that it had previously absorbed. The ensuing uproar caused the FCC to cut back the Internet-connection subsidy.

Okay. That's the past, what's the future? It seems inevitable that competition will win out over regulation. Wireless service will take more and more business. Because wireless service is inherently less reliable than wires, quality is likely to decline. Customers in desirable, densely settled areas will be a lot worse off. Heavy users will get breaks on rates, low users will get clobbered by increasingly high fixed rates. This is one of the unfortunate ways that competition works--you get losers as well as winners.

"It's going to be like it was in the 19th century," predicts Robert Rosenberg, president of Insight Research of Parsippany, N.J. "If you're the right person in the right place, you'll have everything top tier. If you're not, God help you."

Things may not turn out to be quite that apocalyptic. But everything has costs attached to it. If long-distance rates drop and subsidies shrink, someone has to bear the burden. Even in the Brave New World of Telecom, there's no such thing as a free lunch. Or a free phone call.

Allan Sloan is Newsweek's Wall Street correspondent. His e-mail address is sloan@panix.com