Is a government Web site more like a bus or a park? How you answer this question is likely to determine how you feel about the idea of government agencies selling advertising space on their official Web sites. This practice is likely to become more common as municipalities hunt for cash to bankroll large investments in information technology. A company called govAds, for example, is already marketing its advertising-placement services to revenue-hungry governments.

This development leaves some people feeling queasy. To them, a government Web site is like a park, a public space that shouldn't be marred by clever click-through ads featuring images of supine Pamela Anderson look-alikes. Others, though, are not troubled by this prospect. These people point to city buses literally wrapped in advertising and nod approvingly. As long as they don't have to pay for it, they don't care.

The debate is new to the Internet but familiar offline. Many are dismayed by the kudzu-like spread of advertising into the public sector. Schools accept advertising revenue through soft drink contracts and cable television broadcasts. Publicly owned stadiums bear the names of corporate sponsors. Even the presidential debates were sponsored by Anheuser-Busch to the tune of $550,000. It was just a matter of time before governments were approached regarding their Web sites. So what should it be, park or bus?

The promise that the Internet would improve government services is repeated so frequently it has become something of a public-sector catechism. First, the argument goes, information technology will make provision of public services dramatically more efficient. Citizens will go online to register vehicles, obtain fishing permits, renew driver's licenses, apply for zoning variances, pay parking tickets, etc. Second, information-hungry folks will be able to delve deep into issues of great concern by reading reports, council meeting minutes, new proposals and the like. More practical-minded Internet users will be able to access the garbage pickup schedule, athletic-field reservation information and tax rules on their home computers.

Finally, elected officials will communicate directly with their constituents. This interactivity, the gospel goes, will break down the vast chasm that separates the governors from the governed. A new golden age of democracy will be born!

This may all be entirely true. The problem is that making it happen will cost a great deal of money, money that most governments do not have. Maybe someday our more traditional interfaces with government -- bureaucrats behind frosted-glass windows -- will be replaced by kiosks at the local supermarket. Potential long-run savings, however, provide little help in making contemporary budgets add up. Right now, governments have to pay for both the high-tech and the low-tech facilities. That is not cheap.

There are three ways to raise the money. One is to pay for the technology investment out of general revenues. This may work for the limited number of governments that are flush with cash. Elsewhere, however, and assuming that citizens are not eager to pay more taxes, this option is going to encounter significant opposition.

A second approach would be to charge user fees for all Web-based transactions. This has some obvious appeal. It would tax only those directly benefiting from the provision of the service, and it would make it easy to gauge "market demand" for that service. On the other hand, it would price significant portions of the population out of the market for a public good. This is especially troublesome inasmuch as it is likely to reinforce the much-discussed "digital divide." There has already been some grousing that community groups conducting business via Usenet groups systematically exclude certain parts of the community.

That leaves the advertising solution. The public has already expressed some dissatisfaction with the ubiquity of advertising in our lives. Critics claim that by allowing commercialism to seep into educational institutions we are effectively commoditizing our children. The 1996 Olympics in Atlanta were reviled for their omnipresent advertising. Denverites are rallying to preserve "Mile High Stadium" as negotiations with sponsors proceed to name the Broncos' new home.

Still, there is ample reason to shrug at the prospect of Internet ads on government Web sites and say, "A little more can't hurt." After all, the aesthetic displeasure created by the placement of crass come-ons on the same page as schedules of city council meetings is not too big a price to pay.

But there are additional issues to consider. First, the endorsement problem. Advertising on a government Web site may carry a tacit endorsement that should not be sold to the highest bidder. Is it possible to place a banner ad for, say, a hotel on the state tourism site without implicitly endorsing that hotel to prospective visitors? Not surprisingly, govAds' CEO Timothy R. Bartlett says yes. People are used to ads on Web sites, he says, and don't infer an endorsement.

And indeed the bus analogy supports him. Just because Los Angeles buses drive around wrapped in plastic technicolor Yahoo advertisements, does that mean the government is making a portal recommendation? Most would say no.

More problematic is what might be called the Rudy problem. Anyone who spends more than a little time on the Web knows the two strategies that seem to guide almost all advertising campaigns there: Create a banner that looks like a Windows dialog box or try to be as shocking as possible. Assuming that most people (outside of Palm Beach County) are not fooled by the fake dialog box, the shocking ads are more of an issue.

Someone is going to have to make decisions regarding appropriate content. Who is that going to be? New York City Mayor Rudolph Giuliani famously had a series of advertisements for New York magazine removed from city buses because they made fun of him. He even went to court to protect his claimed right to do so. (He was soundly rebuffed.)

But placement of ads on Web sites would invite this type of activity on a daily basis. There is simply no way around it. Unless we are willing to accept any and all advertisements on the Web, there will have to be judges of propriety. Bartlett argues there is no censorship problem because his company will make the decisions based on publicly available, contractually set criteria. This helps, but leaves the door open to controversy. Bartlett notes that government clients have the right to veto approved advertisements. Thus a cyber-savvy Giuliani could reject a mocking advertisement.

Interestingly, govAds touts the right of governments to limit advertising as an advantage. In a document prepared by the company's chief lawyer, govAds tries to soothe potential clients' fears of First Amendment lawsuits. He cites a court decision regarding a Texas city's Web site in which, he says, the judge essentially concluded that a Web site is a bus. And, insofar as it was a "nonpublic forum" (for example, unlike a park) reasonable policies regarding acceptable advertisements, such as those that do not discriminate against particular viewpoints, were just fine. Giuliani's problem was his unreasonableness. Are we clear on that?

Advertising on government Web sites is probably not the most pernicious commercialization of public space. For one thing, it may not work. Skepticism regarding the profitability of ventures that rely on Web-based advertising revenues is quite intense. Bartlett does, however, make a convincing case that he's got a winning formula in his ability to target eyeballs for advertisers. More important, his client list is growing.

Still as more of our lives are spent on-line, our concern with the sale of public cyberspace is likely to grow. Because there is essentially infinite room to expand the Internet, there is no need to protect government Web sites as the Grand Canyon or Everglades of the Web. In this case, effectively designating Web sites as "parks" is more a matter of preserving a mindset, a belief that some things are above commerce.

Talk is cheap. But someone has to pay for the servers and software that will make e-government work. If you're not willing to pay for them, maybe Motel 6 will be.