With a gravity usually reserved for quotes from Harry Truman, this full-page ad reported the chairman as saying, "This isn't getting us where we want to go." And then the ad continued, "It makes no difference whether these problems are due to labor issues, weather or heavy traffic ... ."

My first thought in reading this line was that there was an obvious possible cause that had been omitted: bad management. When, for example, our delayed Sao Paulo flight finally arrived at JFK Airport, we were forced to queue up with 60 other passengers as two agents wrote out tickets by hand for new connecting flights. I doubt the pilots were responsible for United's failure to prepare in advance for these known missed connections.

But more striking was that the chairman could write as if he were really unsure of the cause. Weather, congestion, labor delays - who could tell? We don't have the data to accurately separate the cause from this list. Or at least, the public (or the Federal Aviation Administration) doesn't have sufficient data to know which of these three causes accounts for the disaster that United Airlines has become.

We live in an age when Kmart can tell how many Huggies were sold in its stores across the country 10 minutes ago, but public officials, and hence the public, can't really say why airplanes don't fly. That's not because the data doesn't exist somewhere within the government - it does, spread across a gaggle of different computer systems, no doubt accessible through superhuman diligence - it's just not easily accessible to us. Thus is it easy for chairmen to speak as if the facts were unknown.

This picture points to the next great hope for the information revolution: that we might be able to learn as much about governments and business as they have learned about us. That this might be the end of their effective privacy, just as it has effectively been the end of ours.

The market has provided an excellent incentive for businesses to learn about their customers. Coders have responded to this incentive with technologies for business-to-consumer monitoring. These technologies have produced an extraordinary ability to gather data about the "c" in b-to-c. This capacity in turn has panicked privacy advocates, who are worried that the ability of Little Brother, as The Lexus and the Olive Tree author Thomas Friedman put it, to monitor and penetrate our lives will far exceed the danger posed by Big Brother. Friedman's warning predated Carnivore, the government's system for sifting e-mail, which the government tells us is actually finicky rather than ravenous. But whether it is commerce or the FBI, the focus of the coders so far has been on technologies for monitoring us.

An extraordinary opportunity exists for using these technologies to gather data not about c but about b and g - using the Net, that is, to more efficiently gather and make data available about business and government, those massively powerful, if not always transparently accountable, institutions.

Some of this, of course, is already going on. The BigCompanySucks sites are an important channel for criticism of particular companies. Election information sites are beginning to do the same for government. But both are just a start. If just part of the energy now devoted to monitoring consumers were devoted to more efficiently gathering data from business or government, the potential would be enormous. If it were simple to say why flights on United were so delayed, it would be hard to feign poignancy while muddying the waters of accountability.

The point is not that we need more regulation, or even that we need to gather more data from business. Business already gives up an extraordinary amount of data; most of it is simply lost in the bowels of bureaucracy. The point instead is about efficiency - more from less - and about the benefits that efficiency produces. The government already collects the data I'm describing - whether in safety reports or customer complaints about service. My point is that it could collect that data much more cheaply, and in a manner that would make it easier for consumers and citizens to access and use. And it could do this if it deployed open systems for gathering and storing that data - architectures for b-to-g and g-to-c, that is - that mirrored the architectures of the Net.

We won't get to this more-efficient information architecture, however, without creating some pressure for change within the government. And this points to an increasingly pressing need for the next administration: to establish a chief information officer. This CIO, like any good CIO in business, would be responsible not only for policies that protect privacy, but also for policies that would make information management more efficient and transparent. The objective would be to design systems for managing the government's demand for information more cheaply, not through absurd paperwork reduction act disclaimers, but through architectures that coded better efficiency into the information system.

Information regulations are regulations. And following the best of the Reagan administration, we should focus on ways to make information regulation more efficient. Administrations should start with an estimate of the cost to business and consumers from gathering and accessing the data demanded by the government. CIOs would then be tested by whether they have improved information efficiency. Some simple rules would get the process started: No paper form without a Web-based equivalent, no data-gathering system without a transparent link to nonproprietary protocols such as XML. The aim would be for the government to create pressure for business and governmental transparency, as market incentives have created pressure for the transparency of consumers.

All this points to an underappreciated aspect of the privacy debate. For many, the solution to the privacy problem is technologies that either keep data hidden or that obscure data in the first place. These solutions increase the cost of gathering data, and in turn reduce the demand for surveillance.

No doubt this is often an excellent solution to consumer privacy concerns. But as a general solution, it is more luddite than lucid. Data is power, whether in the hands of consumers or business. Better informed consumers make more-efficient markets; more-transparent government makes less-powerful special interests. The solution to the dangers of data is not obscurity; rather, as author David Brin well argues, it is architectures that permit us to better control how data gets distributed. A proper information architecture would give consumers more control over the data that others collect about them. But it would also enable consumers to better collect data about their government or about actors in the market. Let us know as much about them as they have learned about us, and we'll better decide, in the words of Mr. Goodwin, "where," or with whom, "we want to go."