It's the free market's Orwellian answer to the digital divide: A plucky web start-up offers, say, the cash-strapped Jones family free internet access. In exchange, family members allow the company to track them as they trundle around the net. Mom shops for a new shade of lipstick; little Johnny downloads an 'N Sync song; Dad researches that weird rash on his thigh-and all the while, the internet service provider slices, dices and sells the data. Call it keeping up with the Joneses.

With all the attention focused on bridging the digital divide, new-era thinkers have overlooked the commercial and technological trends that have given way to a new privacy rift that separates a tiny minority of well-off, privacy-concerned net users from a much larger, less prosperous and less informed set of consumers. As more low-income Americans begin to realize the benefits of going online, an increasing number are trading their privacy for free internet access and) sometimes free personal computers. Kmart's BlueLight.com, Freeinternet.com and the eerily named WorldSpy (worldspy.com) now rank among the country's largest free ISPs NetZero (netzero.com) which claims to have more than four million subscribers, is the biggest of them. One clue as to how much NetZero's revenue model depends on sharing user data with other companies: Its IPO filing with the Securities and Exchange Commission in July 1999 listed the possibility of government regulation on privacy among the "risk factors" that might adversely affect its business.

What's wrong with allowing low-income families to trade something of value-their privacy-for goods and services? It's not simply that someone's watching, but rather that companies can use the information they collect to manipulate consumers. The more personal data e-tailers gather, the easier it becomes to engage in price discrimination. Under such a scheme, consumers might receive different price quotes on goods depending on where they live or how much they earn.

Meanwhile, just as market influences and new technologies are conspiring to limit privacy for people of modest means, equally powerful forces are creating a robust market in which well-heeled consumers are buying incognito status on the net. A handful of specialized companies such as Zero-Knowledge Systems and Privada are vying to offer these services for as much as $60 a year. As this pay-for-privacy market develops, people who can't afford the additional cost of anonymity are forced to wander the internet in full-disclosure mode.

At the beginning of the new millennium, the entire notion that the internet is The Great Equalizer-allowing users to throw off the burden of racial, religious or economic discrimination-seems so... 1994. Still, what's most alarming about a future in which privacy stratifies along income lines is that companies have an incentive to recreate the biases that many of us encounter in our daily lives. In the United States, differential pricing is perfectly legal unless it's specifically intended to harm consumers or inhibit competition. But with the exception of airline flights-where you're likely to be rubbing elbows with travelers who paid $400 more for a similar seat-differential pricing formats are difficult for businesses to maintain in the physical marketplace. In 1996, for instance, Denise Katzman, a feisty New York film producer, filed a class-action lawsuit against Victoria's Secret after she discovered that the lingerie company sent out virtually identical catalogs to male and female subscribers, but offered men deeper discounts on some items. Katzman stumbled on the price differences only because a male coworker had left his catalog lying around the office. Victoria's Secret said it wasn't engaging in gender discrimination; it merely wanted to encourage men to buy frillies for their significant others. As a result of an out-of-court settlement with the Federal Trade Commission, the company agreed to stop its gender-based pricing.

Such discoveries are harder to make on the internet. Consumers visiting the same website have little way of knowing who's getting what price or why. In one of the net's rare documented cases of differential pricing, University of Maryland economist Joseph P. Bailey observed that, in 1998, Books.com had been employing a pricing scheme in which some shoppers had received deeper discounts based on their buying behavior. Books.com showed consumers one price for a particular book, then presented them with two options: Either check out immediately or press a button labeled "compare prices"; those who opted to compare prices with Amazon.com and Barnes&Noble.com could be quoted an even lower price. (Books.com has since been sold to Barnes&Noble.com, and there's no evidence that the company still carries out its old pricing system.) The scheme was a crude foreshadowing of things to come. "The dynamically rendered pricing example of Books.com may seem benign or even beneficial to consumers at first," Bailey wrote in a recent paper on internet price discrimination. "However, internet retailers may use complex consumer information, along with shopping behavior, to go beyond a binary separation of consumers who compare prices and those who do not. The ability of computers to store and process large amounts of information may make other forms of price discrimination possible, and very profitable, for internet retailers. Jay Stanley, an analyst with Forrester Research, notes that, as far as the poor are concerned, "a teen profiled as an inner-city minority will be offered the worst prices for online goods-because he's unlikely to be an affluent, high-value customer-and dumbed-down content, because his demographic isn't seen as sophisticated."

Zero-Knowledge president and cofounder Austin Hill doesn't foresee an online privacy gulf between rich and poor. Rather, he believes that ISPS will soon realize the value of privacy and start to bundle cloaking features like his into their internet services. "If privacy is only going to be for people who can afford it, we will have failed our mission," he says. "Our goal is to be as ubiquitous as AOL." Until then, real privacy costs extra ."The risk," says Deirdre Mulligan, a privacy advocate with Washington's Center for Democracy & Technology, "is that privacy could be affordable to some, but out of reach for many."

It's not surprising that privacy advocates didn't anticipate this burgeoning digital disparity. Much of their activism has focused on. privacy violations against middle-and upper-income consumers. The rationale? The well-to-do have-and spend-more money than their less prosperous counterparts, and, as a result, corporations expend more time and capital seeking out their personal data. Fair enough, but low-income consumers are still important to many companies. Consider a joint 1999 study between Pricewaterhouse Coopers and the Initiative for a Competitive Inner City, which showed that inner-city consumers have $85 billion in annual spending power. The bulk of this money is likely to go toward basic consumer goods, such as food, beverages and personal-care items-the very products that Madison Avenue markets most aggressively. In a new era of easily gleaned digitized information, it's becoming more cost-effective to commission a marketing study that reveals why a low-income family might choose Coke over Pepsi, Bic over Gillette, or Pampers over Huggies.

Privacy advocates are at odds over whether to take up this new cause. In part, they see any action as potentially patronizing toward the poor. There's another twist, too, according to Dr. Ann Cavoukian, the Ontario government's information and privacy commissioner. "The poor are disproportionately disadvantaged in this area, but it's better than the status quo, in which no one benefits when companies use their personal information." Though privacy remains a much-discussed topic in Washington, no clear solutions are in sight. It would be wise to set limitations on how companies use consumer data once it's obtained. Whatever the case, federal regulators, privacy advocates and corporations should consider the changing nature of internet privacy. As it stands, it is lower-income families-not the middle class or the affluent-that have the most to lose.