Paul Misener
Vice President for Global Public Policy
Amazon.com
"Internet Public Policy: Finally Getting Beyond the Metaphors"
March 1, 2001 

Paul Misener: Thank you.

It’s a great pleasure to be back in Charlottesville; I’ve been away too long.

Two decades ago, I used to visit here frequently. My older brother Chris is a "Double Wahoo." He received his B.S. in chemistry in 1980 and PhD in chemical engineering a few years later. His wife Linda also is a Double-hoo, with undergraduate and graduate degrees in mathematics. It was always a pleasure to see them here.

And I came very close to attending UVa myself. So close that, after being accepted here in the spring of 1981, I made a deposit to reserve a dorm room (in Metcalf Hall, I believe) and found a roommate. But, for a variety of reasons – including, I’m sure, a desire to chart a slightly different course from my brother – I changed my mind and attended Princeton, instead.

Now, I’m not much of a movie buff, but I think I’ve seen just about every film in which a person travels back through time in order to change the present. The Terminator series and Dennis Quaid’s Frequency, are examples of the genre. So this week I got to wondering what part of the present would change if I returned to 1981 and stuck with my initial decision to attend UVa.

Well, for one thing, Amazon.com would be profitable.

And who knows? A cheap Chevrolet might be called a "Tiger," and the world’s best golfer might be named "Cavalier Woods."

But in spite of my questionable judgment 20 years ago, I’m smart enough to know to commence speeches at the University of Virginia with the words of Thomas Jefferson.

I didn’t select, however, one of the well-known passages having to do with "unalienable rights," and such. Indeed, in what might initially appear to be a nefarious act of a vindictive Ivy-Leaguer, I chose the quote that garnered Jefferson more ridicule in his day than any other.

I hope to show you, however, that this quote was worthy of Jefferson in a way that only the benefit of hindsight can prove.

Way back in 1783, atop a high rock that now bears his name, in the southeast corner of a county that also now bears his name, Jefferson gazed down upon the confluence of the Shenandoah and Potomac rivers and pronounced the scene, quote, "[w]orth a voyage across the Atlantic."

Yet, at that time, the place had only recently become known as Harpers Ferry, replacing the unflattering yet descriptive moniker, "The Hole." And the vista, though pleasant, was hardly worth a roundtrip flight from Heathrow to Dulles, much less several month’s of seasickness and peril. Not surprisingly, Jefferson’s assessment was widely mocked on both sides of the Atlantic. In a relatively understated commentary many years later, John Quincy Adams said that, quote, "those who first see [Harpers Ferry] after reading Mr. Jefferson’s description are usually disappointed."

But as he did so many times throughout his life, Jefferson saw clearly what others couldn’t even imagine. At Harpers Ferry, this Founding Father must have seen beyond, in his words, "one of the most stupendous scenes in Nature," and foreseen the commerce and communications that soon would flourish there. Ironically, these future developments are the reason that Jefferson’s Rock, Harpers Ferry, and the rest of Jefferson County no longer are part of the Commonwealth of Virginia.

* * * * *

But you’ve come to hear a speech about Internet Public Policy. So let me begin by discussing an era of rapidly growing networks, competing new technologies, and boundless opportunities for commerce.

I’m referring, of course, to Virginia in the mid-1830s.

The rapidly growing networks were run by companies called "Chesapeake & Ohio" and "Baltimore & Ohio"; the competing new technologies were the canal and the railroad; and the opportunities for commerce focused on communications and trade with the fertile Ohio Valley and the Old Northwest.

As likely foreseen by Jefferson, one of the most important nodes of Virginia’s network was Harpers Ferry. About three miles outside of this town, I own a farmhouse that was built in 1830. It was three years old when the C&O canal arrived from Washington, DC, and it was four when the locals welcomed the B&O railroad from Baltimore.

When my West Virginia farmhouse was built, of course, it was in Virginia, not West Virginia. The reason why my farm, Harpers Ferry, and the rest of Jefferson County now are part of West Virginia is not as obvious as it may seem.

At the time of John Brown’s famous raid of 1859, and commencement of the Civil War a year and a half later, the three counties now known as West Virginia’s "Eastern Panhandle" were aligned geographically, economically, culturally, and politically much more with the plantation counties of the southeastern Old Dominion, than with the transmontane counties that today comprise the remainder of West Virginia.

And, yet, Jefferson, Berkeley, and Morgan counties were included in the newly formed state of West Virginia in 1863.

Why?

Because of the networks.

Quite simply, the C&O canal and, even more importantly, the B&O railroad were far too important to the people and economy of western Virginia for the founders of the new state to countenance Confederate control. Indeed, without control of these networks, the economy of Wheeling – then the second largest city in all of Virginia and the center of the statehood movement – was at risk. So, with the help of the federal Army of the Potomac, and against the wishes of the vast majority of local citizens, West Virginia laid a specious claim to the counties of the Eastern Panhandle.

Well, the importance of networks to the people of Virginia is just as clear today. An oft-cited statistic notes that one-half of the world’s Internet traffic flows through the Commonwealth. So, rebellion or not, governments still need to do what is appropriate to protect the use of networks.

This is what public policy in the Internet Era is all about.

Some time ago, America Online’s Steve Case, the most prominent of Virginia’s high tech CEOs, opined that the future of the Internet, quote, "will be determined more by policy choices than by technology choices." In essence, said Mr. Case, policymakers, not technologists, are now running the Internet.

As any of you who deal with government are aware, this is cause for concern.

But how hard can this policymaking be? After all, as we’ve just noted, governments have been grappling with Virginia’s networks for well over a century. Why not just keep up the good work, as it were?

Well, I believe there is a huge and pervasive problem in online public policy making. This problem must be addressed soon, or serious mistakes will be made over the next few years.

This problem, simply stated, is the inability or unwillingness of many public policy makers to recognize the real differences and real similarities between online and offline activities.

Let’s look at this problem in detail.

Online vs. Offline Differences

First of all, in spite of all the metaphors we often use (the information superhighway allegory being the most common) there are real and important differences between the Internet world and the physical world.

Take Amazon.com’s business, for example, which is retailing.

In a small, isolated town, a single store can hold a near monopoly over sales to the town’s consumers. Although the mere threat of another store opening somewhat regulates prices and service levels, the economy works far better when two stores actually are open in the town. If one retailer raises his prices, he knows he will lose sales to the other; and if one retailer preemptively cuts her prices, the other must follow or lose sales.

The system isn’t perfect, however, for two principal reasons. The first is information: consumers may find it too difficult to know what the prices or service levels are like at the other store, and therefore may unwittingly pay a higher price or accept worse service. The second is friction: almost invariably, one store will be located more conveniently to some of the town’s consumers, and the other store more conveniently to other consumers.

As the number of stores in the town increases, the effects of information (or the lack thereof) and friction decrease. So when there are, say, four or five similar retailers serving a town, there will better information flow and less friction. It just keeps getting better and better as more stores are added.

Great stuff, huh?

Well, except that experience tells us there are never more than a few stores serving a small town or small section of a larger town or city. So, in the offline retail world, there always will be a shortage of information and an abundance of friction.

But this is not the case in the world of online retail. And this is a very important difference for public policy makers to recognize.

In the world of online retail, individual consumers are far more empowered, because knowledge about prices and services is free and easy to obtain, and because there is virtually no friction for consumers to move among thousands of different stores.

Because consumers are so empowered, many traditional consumer protection laws that apply to offline retail are unnecessary, at best. If consumers don’t like the way they are treated at one Internet store, they can – and will – migrate elsewhere. Consumers online are not beholden to the one store in town, or even to the closest of four or five stores.

So what’s the importance of this difference? Quite simply, there is little or no need for government to prohibit or dictate certain online retail behaviors; consumers won’t put up with being mistreated and, because they are so empowered, they need not.

Thus, in spite of the analogies we all use, there are real and important public policy differences between online and offline retail.

Online vs. Offline Similarities

But what about the similarities between online and offline?

In spite of the many obvious (and some subtle) differences between these worlds, they are similar in many important respects.

One key similarity is something many people incorrectly believe to be a difference: the use of computers. Virtually all retailers – online and offline – are heavy users of computers for marketing and sales purposes. Yet, because online retail involves computer use by consumers, some policy makers are led to believe that special rules need to be applied to the online retail industry in order to protect consumer information. With all due respect, I think this belief is misguided.

Let’s examine the issue closely. A consumer doesn’t personally use a computer when she writes a check; when her "customer loyalty" card gets scanned at the grocery; when she buys a tank of gas on credit; or when she fills out a warranty card for a new electric heater. Nonetheless, her information is soon thereafter collected and shared among vast computer networks.

The electric heater example is very fresh in my mind. A few weeks ago, I bought a new $15 space heater for my farmhouse and, inside the box, was a yellow, multifold business reply mail card that said, in big letters, "Important! Fill out and return this card to register your product." Oddly enough, to, quote, "register my product," they need to know my household income; where I took my last vacation; if I read the Bible; and whether anyone in my household has prostate problems.

But I’m safe, right? I didn’t use a computer! It’s not that scary newfangled Internet, after all! It’s just a cheap paper card that I filled out in pencil, right?

Wrong.

The fact is that, when I mail that card, a typist somewhere is going to enter my information into a huge database that will be used to market a wide variety of products – way, way beyond electric heaters – to me and my family.

Oh, yes, and although I might have been concerned about buying that electric heater over the Internet because an online retailer would then know where I live and that I’ve purchased a heater, I have, by filling out that little yellow card, told the heater company about my finances, travel history, religious beliefs, and family health! And there’s no law to prevent the heater company from sharing this information with others.

The message here is simple: consumers and policy makers need to get beyond the analogies and pay close attention to what the online world is, and what it is not. Failure to account for differences and similarities will not solve any existing problems, and easily could create new ones.

Key Public Policy Issues

For the remainder of my remarks, I would like to focus on two public policy issues that exemplify the importance of understanding and acting upon the differences and similarities between the online and offline worlds. These issues, which promise to be the most important e-commerce and Internet public policy topics for 2001, are online sales taxation and privacy.

Internet Sales Tax

Let’s look at the online sales tax issue first. This is an issue where the differences between the online and offline retail worlds are critical, yet frequently missed. The key question is whether or not online retailers, such as Amazon.com, can and should be required to collect taxes on sales to customers in states where the retailer has no direct connection, known as "nexus." Again, and this is important, the question is whether or not online retailers can and should be required to collect taxes on sales to customers in states where the retailer has no direct connection.

An example here may help.

Amazon.com is headquartered in Washington State; that is our principal place of business, our sales organization is located there, and there we have tax nexus. We have no facilities in Wyoming, however, and lack tax nexus there. So when we sell an item – say a frying pan – to a customer in Washington State, we collect sales tax for the government in Olympia. But when we sell a frying pan to a customer in Wyoming, we cannot be required to collect sales tax, and we don’t.

The reason for this situation is that the United States Supreme Court reaffirmed almost a decade ago that forcing remote sellers to collect and remit taxes on sales to customers in other states is an unconstitutional burden on interstate commerce. At the time the Court first ruled on this burden, there were some 2300 jurisdictions (states, counties, cities, townships, etc.) collecting sales tax. 2300. It is interesting to note that currently, there are over 7600 jurisdictions collecting, and another 23,000 that could! If anything, the U.S. sales tax system – if it can be called a system at all – is getting more, not less, burdensome.

However, the Supreme Court also said that, at some point in the future, Congress could determine that the tax collection system had been adequately simplified, and that the unconstitutional burden no longer existed. Sales tax simplification efforts putatively designed to remove this burden will be at the heart of the Internet tax debate this year in Congress.

Note that sales tax is not at all addressed by the three year Internet tax moratorium adopted by Congress in the fall of 1998. Indeed, the Internet Tax Freedom Act merely blocks "new, multiple, and discriminatory taxes" online. Although it’s a good law because it prevents the local dogcatcher from levying a tax on the Internet traffic that runs through a cable in his jurisdiction, it has nothing directly to do with the collection of sales tax. Of course, efforts to extend the moratorium will be linked politically to the sales tax issue.

Let me emphasize here that Amazon.com is not anti-tax. We like schools, police cars, and firefighters as much as anyone else, and in most places these are supported by sales taxes.

But Amazon.com is anti-burden. The difficulty and expense of collecting sales taxes nationwide under the current crazy quilt of tax definitions, policies, and rates are untenable and obviously unconstitutional. We are, therefore, working to ensure that state tax codes eventually are simplified to the point where there no longer is an unconstitutional burden.

What would it take to achieve true simplification? Well, although a uniform nationwide sales tax rate has been suggested, that would not be necessary. Certainly, such an uniform rate would make collection easier, but it would be very difficult to achieve politically.

The most important aspects of simplification are readily achievable. For instance, the burden of determining tax jurisdiction must and can be lifted. Let me explain.

In one infamous example, a single zip code in Denver, CO actually has five different sales tax rates, ranging from about 4.5% to over 7%. We wouldn’t mind collecting either rate (or some other rate, for that matter), but we need to know which rate to collect! If we under-collect, we subject ourselves to liability from local tax authorities. And if we over-collect, we subject ourselves to class-action lawsuits from our customers.

So, you say, why don’t we just ask our customers? There are two problems with this approach. First, any such dialog would seriously diminish the customer experience at our website and, frankly, scare customers away. Imagine having the clerk at the local brick and mortar store demand to know your sales tax jurisdiction before he will tally your purchase. Indeed, because brick and mortar stores collect tax at a rate determined by the store’s location, not where the customer lives, consumers can make purchases in brick and mortar stores without being asked about their tax jurisdiction. At least the same convenience should apply to online retail purchases.

The second problem is that customers often don’t know in which sales tax jurisdiction they live. So even if we ask them, they may not be able to tell us. In sum, if there is no way for us to determine a customer’s tax jurisdiction from his or her basic mailing address, there may be no way for us to know or for them to tell us.

Accordingly, easily ascertainable tax jurisdiction rules must be adopted. As I mentioned before, a single nationwide rate would be nice, but one rate per state would be acceptable, and Amazon.com would support legislation establishing that goal. We could, however, even live with a single rate per zip code, because we already have zip codes for delivery and billing purposes and, most importantly, consumers know their zip code.

Another facet of the unconstitutional burden that must be addressed is reasonable compensation to retailers. Well over 99% of Amazon.com’s transactions are consummated with a credit card. And, because all of our sales are remote, we get charged by credit card companies the very highest service rate, known as, quote, "card not present." Were we suddenly required to act as sales tax collection agents for states and localities, we would have to pay the credit card companies, out of our own pocket, this very highest credit card fee percentage on the tax we collect and remit. A back of the envelope calculation would put this new out-of-pocket price of sales tax collection at about $7 million a year. And this expense is in addition to the obvious cost of infrastructure and personnel to manage the collection. We believe we deserve reasonable compensation for the expense of serving as tax collectors.

There are a few other key aspects to the unconstitutional burden that must be removed before Congress can allow states and localities to require out-of-state sellers to collect and remit sales taxes. Importantly, removing these burdens is an attainable goal. The bar is not too high. And Congress must be satisfied that the burden actually has been lifted, at the peril of reversal by the Supreme Court.

But in addition to removing the current unconstitutional burden, there is one aspect of the online retail taxation question that is even more fundamental: why should this tax be collected? Even if, at some point in the future the state taxation system is simplified to the point that there no longer is an unconstitutional burden on out-of-state retailers, it is not clear that collecting this tax would be the right public policy choice.

After all, in spite of the doomsday predictions of some state and local officials, e-commerce has not had an adverse effect on state and local government revenues. To the contrary, the vast majority of state and local governments are running surpluses, and the tremendous efficiencies of the Information Age are largely to credit.

Moreover, our tax codes are full of behavior incentives that favor socially beneficial activities over other activities. Because online retail is consumer-friendly, environment-friendly, and led by American companies, it is a perfect candidate for favorable tax treatment. So-called sin taxes are applied to discourage the use of alcohol, tobacco, and gasoline; why not leave the blessing of Internet retail alone?

Finally, even if at some point, say five or ten years from now, it can be shown that state and local governments are suffering from lost taxes due to online retail, and even if at some point years from now it makes sense to tax online retail in spite of its obvious benefits to society, and even if at some point many years from now the tax codes of thousands of jurisdictions in this country can be simplified to the point that collection would not be an unconstitutional burden, online transactions should never be taxed at the same rate as offline.

The reason is simple: online transactions use fewer state and local services than offline transactions.

Consider a consumer in her home. She needs to buy a cordless drill, a CD, a book, and a bottle of shampoo. To do this, she can spend two hours getting in her car, driving to, and shopping at, four different stores all across town. In addition to wasting her time and societal productivity, she has caused pollution, helped create traffic jams, and increased the probability of traffic accidents and, accordingly, has increased the need for police and fire services. She also has patronized stores that require significant police and fire protection.

Alternatively, she could have stayed in her home and purchased these four items in the span of about five minutes on Amazon.com. To be sure, some state and local services still are necessary: her home still requires police and fire protection, and the delivery truck will pollute and help cause traffic jams.

The point is not that her online transaction caused no burdens on her state and locality. Rather, the point is that online retail transactions place fewer burdens on states and localities. And, as a result, online transactions – to the extent they should and can be taxed at all – should be taxed at a lower rate. To do otherwise would over-collect taxes and would unfairly punish online consumers. Again, this is an instance where it is important for policy makers to recognize the differences between online and offline activities and act accordingly.

Internet Privacy

Well, enough for now about tax. Let’s talk about online privacy. This is an issue where the similarities between the online and offline retail worlds are critical, yet frequently overlooked.

There is a belief, perpetuated by some policy makers and certainly the media, that because of online activities, personal privacy is at serious risk. There also are those polls, which all seem to say that some 75% of Americans are concerned about their privacy online. And finally there are the horror stories: some guy you’ve never met is using your credit card to rack up thousands of dollars of purchases.

Well, the message being received in Washington is: "Quick! Do something!"

But we’re in Charlottesville, the home of Jefferson, where reason should triumph over hysteria, so let’s talk some sense here.

First of all, personal privacy simply is not at serious risk due to rational online activities, any more than it is at risk due to rational offline activities.

Let me explain by way of analogy to personal safety in Washington, DC, a city where I spend a lot of my time. The vast majority of Washington’s shopping sites – generally the stores and outdoor markets – are perfectly safe places to shop. On the other hand, there are some dark alleys where one can buy a variety of products, both legal and illegal, that are extraordinarily dangerous. Common sense says that everyone should shop only at the safe places. The same common sense applies online.

Similarly, those surveys amaze me. Where do they find those people in the 25% who are completely unconcerned about privacy online? I’m very surprised that less than 100% of people are concerned about online privacy, for the very same reasons that 100% of people should be concerned about their personal safety in Washington, DC.

Moreover, people should prefer more privacy over less, better service over worse, more personalization over little, and lower prices over higher. In other words, the problem with these polls is that they are incredibly oversimplified. All other things being equal, of course we all want more privacy! And of course we want better service! And of course we want more personalized service. And of course we want lower prices! It’s simply a matter of balancing these qualities.

Finally, what about those horror stories? Some guy gets a hold of your credit card number and charges his new teak deck furniture, his phone calls to Paris, and a long night at Caesar’s Palace. Could it happen? Sure. Does it happen? Very, very rarely. Will you have to foot the bill if it does? No, by law it’s limited to fifty dollars, and now some credit card companies don’t require you to pay even a dime.

But a key question for us is: does this happen only in the online world?

Absolutely not!

One of my favorite Dilbert comic strips depicts Dilbert out on a dinner date. As he hands the plainly dressed waitress his credit card, he’s telling his date about how he refuses to put his credit card number on the Internet. He believes that it’s just too risky. At that moment, unnoticed by Dilbert, his waitress returns with his credit card, but this time she’s dressed in a fur coat and wears a diamond necklace.

It really is amazing that some folks are scared to use a secure Internet connection to send credit card information for a purchase on a reputable website, yet don’t think twice about handing their card to some anonymous waiter in a restaurant. And before you point out that the waiter is just one person, while the Internet could be used to distribute your credit card number to millions of people, remember that our unknown waiter almost certainly has an email account and could do the very same thing.

So, it’s unfortunate that these irrational fears are helping to drive the current online privacy hysteria. And, just as bad, they are making it difficult to see and understand reasonable privacy concerns.

Are there reasonable online privacy concerns? Absolutely. Just like there are reasonable personal safety concerns when, say, shopping on the streets of Washington, DC. The same common sense pertains to all shopping: there are many very safe places to shop, and there are others that are much less safe.

One very safe place to shop is at Amazon.com. We have long prided ourselves on our customer-centricity and, on behalf of our customers, last fall updated our privacy policy. We had adopted our first privacy policy long before most online retailers and, over time, as our business grew from an online bookstore to what we call Earth’s Biggest Selection – which includes tools, cosmetics, toys, and even automobiles – we occasionally added provisions to the policy, but not in a particularly well organized fashion. So, last spring, we began an effort to update the policy to make it better organized and more accessible to our customers.

When we were ready to post the updated policy on our site, we took the highly unusual step of notifying by e-mail all of our then-current customers, who then numbered 23 million. Indeed, even though the usual practice of Web sites is to hide or at least downplay changes to privacy policies, we proactively notified all of our customers. We are proud of the substance of the updated policy, and wanted to make sure our customers knew about it.

Well, the reaction in the press was extraordinary.

Extraordinarily bad, that is.

Somehow, many of the stories got it exactly backwards. Instead of lauding our updated policy in which we clarify and further limit our use of customer information, many press accounts lambasted us for deciding to sell customer information for the first time.

Let me set the record straight: under our old privacy policy, we reserved the option to sell customer information to third parties at will. This, of course, is standard practice among magazines, credit card companies and, as we have seen, heater manufacturers, so it was not unusual for us to reserve this option. Also, it was customer-friendly – and not legally required – for us to state that we might exercise this option at some point in the future. Even better, under the old policy we had given all of our customers the opportunity – through an email address on our site called never@amazon.com – to preclude the sale of their information to third parties such as telemarketers.

It turns out that we actually never, ever exercised the option of selling customer information that we had reserved for ourselves. Moreover, in our new policy, we pledge never to do so except in the limited context of the sale of a business unit or the entire company. We state emphatically that we are not in the business of selling customer information. In other words, we specifically limited – not expanded – our options.

In this hype-charged environment, what should policy makers do about online privacy? Well, like doctors, policy makers should first do no harm. The best course of action would be to fully understand the similarities of online and offline commerce before acting. And then, to the extent any privacy legislation is necessary, it should apply equally to both the online and offline worlds. To do otherwise would mislead consumers and, sadly, provide greater "protection" only to those fortunate enough to be online. The elderly, poor, or uneducated folks not online would receive none of the putative benefits of this law.

Many folks in Washington, DC, are saying that online privacy legislation in the 107th Congress is a forgone conclusion because the issue, quote, "polls off the charts," and that it’s one of the handful of visible issues on which Democrats and Republicans really see eye-to-eye. Well, there is some truth and some fiction here, but one thing is clear: in Washington, if you keep saying something long enough, it becomes reality. And this certainly would not be the first time that politicians championed a solution in search of a problem.

Believe it or not, however, there actually is at least one fair reason to support federal online privacy legislation. This is the threat of state-by-state law on the subject. It would be very difficult, if not impossible, for a single nationwide website to abide by 51 different sets of rules affecting the content of the site. Notwithstanding this fact, last year there were well over 100 state bills introduced nationwide that addressed online privacy issues.

Thus, online entities may be in a position to support federal legislation designed to preempt state and local laws. And such preemption would be an absolutely necessary component of any other federal privacy legislation. Here is another instance in which the unique nature of online activities precludes a solution – here, state-by-state rules – that might be fine if it were applied to offline activities.

As you may have heard, privacy bills tend to touch on two or more of the four fair information privacy principles, which are notice, choice, access, and security. Of these, the only obvious candidate for consideration is a requirement that online sites provide users some notice of the site’s privacy practices. With such notice, consumers can decide what balance of privacy features they desire, and the Federal Trade Commission already is empowered to force websites to live up to their promises. And because the FTC has this authority, private rights of action must not arise from new legislation. Federal regulators can handle these issues in a much more rational and uniform way than class action plaintiffs’ attorneys.

Other Issues

As I’m sure has become obvious, I could continue talking for hours. And, let me assure you, there are hours worth of other public policy issues to discuss. Before I close, however, I’ll just briefly mention a few of them.

Business Method Patents

So-called "business method patents" were validated by the United States Supreme Court in the State Street Bank case in the early 1990s. Amazon.com holds several of these patents and, in spite of the fact that the patents are legally valid, has received considerable criticism as a result. This criticism caused us to think a lot about the current patent system and whether it works well in the Internet era.

We concluded that business method and software patents are different than traditional patents and, accordingly, deserve disparate treatment. Again, where there are real differences, there should be different law.

In the first place, it makes little sense to protect inventions like our "One-Click"™ technology for twenty years; we believe as little as three to five years would be sufficient. In addition, because the techniques and technology behind many of these inventions is changing very rapidly, we recognize that it may difficult for patent examiners to keep up. Thus, we support a pre-grant public comment period for U.S. patents that would allow outside experts to help examiners discover any prior art. We also support full funding for the U.S. Patent and Trademark Office, which currently serves as a cash cow for the government, taking in considerably more fee money than it is allowed to spend.

Obviously, the adoption of such changes would not be in our direct best interest. As holders of business method patents and inventors of other methods, we selfishly might want the patents to last as long as possible and be easily granted. But, taking a very long- term view to the health of the Internet, my boss and I actually have lobbied for these changes. In addition, he has invested heavily in a company, called "BountyQuest," that allows people to post rewards for prior art information on business method and software inventions.

Incidentally, two prominent members of Congress already have introduced legislation on business method patents. I would expect there to be some debate on this issue in the next two years.

Online public policy issues are not limited to the United States. For example, the European Union and its member nations are grappling with the issue of taxing digital downloads such as electronically delivered books, music, and video.

EU members currently apply value-added tax to two categories of commodities: goods, on the one hand, and "non-goods" – a.k.a. "services" – on the other. In some countries, the difference in VAT rates between the categories can be striking. As the EU decision currently stands, digital downloads are treated as "non-goods" even though, it’s obvious to so many of us, that they are clear substitutes for physical books, CDs, video tapes, and DVDs, all of which are taxed as goods. Nowhere is the problem with this approach more evident than in the UK, where paper books are tax-free goods, while e-books sold online are taxed at a 17.5% non-goods VAT rate.

Here, again, it’s a case of where the similarities of offline and online retail need to be recognized and taken into account.

Hack Attacks

Finally, on an issue that received a lot of media coverage one year ago, public policy makers can be expected to continue to address the problem of hacker attacks on e-tailers and other online entities. In Amazon.com’s case, for about an hour on February 8, 2000, a large amount of so-called "junk traffic" was directed to our Web site. This junk traffic degraded the technical quality of service at the site, and somewhat hindered our customers’ ability to visit and shop. Obviously, no matter how much attention we dedicate to protecting our site, such attacks cannot be prevented without the cooperation of other online entities or the participation of government and, thus, this issue appears ripe for legislation.

Everything Else

Unfortunately (or perhaps fortunately for those of you with plans this evening), we cannot today discuss all the other public policy issues on the table, but let me list a few: domain names and trademarks; foreign government content regulation; product label tampering; unsolicited e-mail or "spam";

database copyright protection;

third party liability; the digital divide; and instant messaging. It’s enough to keep a legion of policy wonks like me very busy.

Conclusion

So, in conclusion, I think it’s very important that policy makers worldwide take the time to recognize the real differences and similarities between online and offline activities, and act accordingly. We need to, at long last, get beyond the metaphors. I urge you in academia, industry, and elsewhere to be precise. Don’t let similarities between the online and offline worlds conceal the differences, or let differences cause us to overlook the similarities.

Thank you for your attention. I welcome your questions.