Broken Symmetry

Patent Reform idea: Bifurcate the Patent System?

[This is a cross-post from Patent Prospector where I'm guest blogging.  Check out the comments and replies from Patent Hawk over there!]

Thanks in part to his entry on Entrepreneurship on August 17, Patent Hawk has invited me to guest blog at the Patent Prospector.  As a longtime fan of his blog, I'm honored to have the chance to share some thoughts with him and you, his readers.

Patent Hawk and I both believe in the value of the patent system for the United States.  More specifically, I don't think it absurd to suggest that the phenomenal industrial growth we've seen in the United States since the 19th century has been due in part to the incentives created by the patent system.  Where Patent Hawk and I may disagree, however, is on the question of whether uniformly stronger patents are good for our economy.  In particular, I believe that the recent Supreme Court cases and PTO rulemaking cutting back at the scope of patent protection may be beneficial for some high-growth markets in our economy.

It is a rule of law under our constitution that Congress shall have the power to "promote the progress" of science and technology by granting patents to inventors.  To paraphrase in economic terminoloy, Congress can, as necessary, grant patents as an incentive to develop new technology.

The main reason I support patent reform is this: Some new technologies get developed without the extra incentive of patents.  For these "fully-incented" technologies, the patent system can be more like a tax than a credit.  In the worst case scenario, the inventor who already had incentives to develop new technology does not apply for patents, but then has to pay the other later inventors or copyists whose only incentive was patents.

The messes of the patent system can thus be tied to the system's lack of economic sophistication.  For the most part (Hatch-Waxman is an important exception), it treats all technologies the same.  Perhaps the most important and effective step we could take in patent reform is to recognize that inventors have different incentives in different technological markets.

Having identified the mess, it's easy to spot the problem that created it: technology markets are as diverse as technology itself.  It's tough to distinguish a "fully-incented" technology market from one that would benefit the added incentive of patents.  We have one patent system for all technology because we're unable to draw the line.

Notice how this helps explain why patent reform legislation has been so difficult to pass.  You're either for the patent system or against depending on what side of the line you're on.  Pharmaceutical companies are generally on one side; consumer-Internet companies are generally on the other.  If we were able to reliably divide the "fully-incented" technologies from the rest, we'd make almost everybody happy.  Pharmaceutical companies get stronger patents; consumer-Internet companies get weaker ones.

How to draw the line?  In economic terms, I think the venture capital market defines reasonably well the "fully-incented" technologies.  If an inventor is able to go out and get funding for her idea on the private market, then additional government incentives to invent are generally irrelevant.  It's true that some VCs like to see their portfolio companies apply for patents, but this is because in a world with the patent system it's a tax you have to pay (see above).  Patent protection is at best a high-order term in the expansion that VCs use to decide whether to invest in a new company.

But although precise economically, "subject to possible venture capital investment" is not a great legal category.  As an imperfect substitute, I suggest the following: reverse-engineerability.

Technologies that are easy to reverse-engineer need patent protection in the sense that without it, copyists are far too likely to encompass part or all of the value created by the inventor.  Technologies that are difficult to reverse-engineer, by contrast, may be marketed and sold for some time before a trade secret theft or independent invention permits others to enter in competition.

Readers, what are your thoughts?  Could a patent system that denied patentability to technology that could not be reverse-engineered ever work?  As a practical matter, this could be implemented simply by amending 35 USC Section 101.

How hiearchies develop in social networks

Hundreds of years ago in England, when real property changed hands, it changed hands through a ceremony called the livery of seisin.  The ceremony was necessary for its evidentiary value.  Before the existence of written title and recording offices, disputes over ownership were settled by eyewitness testimony.  Incidentally, this is probably also why you hear the minister ask the crowd whether there is anyone present who has reason to the object to a marriage -- there were no marriage certificates until relatively recently.  Similarly, until the resume came into common use (probably within the last 100 years), work experience was based almost entirely on eyewitness testimony (i.e., references).  In general, one had to talk to the people who had known a person to learn even basic information about that person's status (property ownership, marital status, employment).  Even after the paper recording systems had been established, the non-trivial costs of the system were born by the searchers, so most searchers were businesses that could estimate in dollars and cents how much that information was worth.  Socially, it is still the case (except in extreme circumstances) that nobody is going to do a marriage certificate search to determine whether someone is married, much less if someone owns property.

The way in which status information is stored and made available within a community has social consequences.  One such consequence is the relative flatness of hierarchies in large (offline) social communities.  Think of the difference between how you pick your friends and how you hire people.  In choosing who to hang out with, decisions are made almost entirely through references within your network.  The balance of information comes through the external status signals that someone sends (the first impressions).  But external signals are often unreliable ("don't judge a book by its cover").  There simply isn't enough information for strangers to be placed within any hierarchy.  As a result, social hierarchies are flatter in large (e.g., urban) communities in which people move around often and are constantly being exposed to new people.

Social hierarchies do exist, however, in smaller (e.g., rural or church) communities.  And they exist also in corporations and government institutions.  What these groups have in common is improved storage and availability of status information.  Rural communities have no storage or availability advantages over urban communities (i.e., still memory and word of mouth).  But they have the advantage of time.  Corporations and government institutions have resumes and titles.

Until recently, the Internet has been more like a large urban community than a small town or a corporation.  Social hierarchies, to the extent they exist, have developed organically from public-facing value.  For example, active bloggers from relatively less well-known academic institutions have developed some of the largest followings.

But the Internet is changing in this regard.  More and more offline status information is going online. The result is a de facto audit of personal status information.  Already, some professional status information is available free.  Eventually, title searching, birth, death, and marriage certificates will also be free.

The development of social hierarchies on the Internet looks, unfortunately, to be an inevitable result.

The credit market debacle

Major financial scandals seem to transpire once every three to five years.  The cycle seems to go in four phases: (1) price crash; (2) journalist frenzy, usually ending with the revelation of scheme to "defraud" investors; (3) government investigation and prosecution; (4) private lawsuits and reorganization.  The quotes are there, of course, because for most people, most of the time what looks culpable in one context (e.g., when indictments are being handed down) looks merely aggressive in another (e.g., when everyone else is doing it and you're afraid of being at a competitive disadvantage if you don't).

We seem to be nearing the tail end of the cycle for options backdating scandals.  I'm thinking we're just now getting into phase 2 with the credit markets.  My prediction is that some major fraud in how collateralized debt obligations (CDOs) were packaged and repackaged into securities is going to be "discovered," and one of the major investment banks is going to be left holding the bag.  Like the risk profile for LTCM assets, the risk profile for CDOs was determined based on market data from the past.  And if there is one place where Shakespeare's dictum ("what is past is prologue") fails, it's the publicly traded markets.

Things may not end as badly as they did for Arthur Andersen.  Almost nobody's retirement is going up in smoke this time around.  But nobody's pension disappeared because of options backdating either, and there was nonetheless plenty of of public political activity in response.

UPDATE: looks like the SEC is ready to skip to phase 3, and the law firms to phase 4!

UPDATE 2: We're now definitively in phase 3.  Check out this photo.

Series FF stock and the Friedman-Savage Model for Risk Profiles

The Friedman-Savage model was developed to explain an economic paradox: assuming rationality, why would the same person buy both insurance and lottery tickets?  The answer that Friedman and Savage proposed is a double inflection utility curve:

The x-axis here corresponds to the size of the payout for a particular activity, the y-axis to the utility actually gained from the payout.  (Click on the graph for a larger version.)  Imagine that right now you have wealth = zB, and thus have utility = B.  The uncertain future forces you to take two different gambles: one with a payout somewhere between A and B and another with a payout between B and C.  The expected value of the payout is shown by the chords E and E'.  Although the expected value is everywhere lower than the actual utility gained from the payout between points A and B, a risk-averse person will prefer the certain result of E.  Thus, a risk-averse person would choose paying an insurance premium costing E over an uncertain result between A and B.  Likewise, a risk-preferring person would prefer the certain result of E' (which might be obtained, for example, by squandering income on lottery tickets) to the uncertain result between B and C.

The need for Series FF stock, which was pioneered at The Founders Fund where I work, can be explained by assuming that entrepreneurs (in particular, company founders) bear a similar risk profile.  Series FF stock can be issued to founders in place of common stock when the company is founded.  But unlike common stock, at later rounds of financing, a portion of the Series FF stock can be sold (subject to board of directors approval) to a later-round investor, and converted into whatever series of preferred stock is being issued at that round of financing.  For example, if a company founder held a 25% stake in the company through Series FF stock at formation, that company founder could one or two years later choose to sell 10% of that stake (or 2.5% of the company) to a Series A investor.  The company founder would get cash equivalent to the 2.5% stake at the Series A valuation.  The investor would get a 2.5% stake in the company in Series A stock.  Why should this be preferred to the status quo, in which company founders (usually) keep their common stock until the company is either acquired or goes public?

Entrepreneurs are extremely leveraged in their startups, and take a relatively small salary for the amount of value they create.  What they get in exchange is the payout on their equity when the company hits liquidity.  Although the timeline from seed stage to liquidity is shortening in some areas, the period is still on the order of 10 years.  This presents a problem from the point of view of maximizing the value of a company when, 5 years in, the entrepreneur is under pressure from personal creditors at the same time she is entertaining acquisition offers from a larger company.  In that situation, the entrepreneur has a strong personal incentive to sell the company at a low (but certain) price now, rather than holding out for a better offer in the uncertain future.

Referring to the curve, if the average person (including the average entrepreneur) tends to be risk-averse at lower payouts and risk-preferring at higher payouts, then it makes sense for investors to compensate for that by (1) monitoring management through board seats, thereby avoiding the excessive risk-taking between points B and C that is probably more common at later-stage companies, and (2) offering some liquidity, thereby avoiding the excessive risk-aversion that is probably more common at early-stage companies.

Most venture capital investors are already doing (1).  Series FF is the first formal, concerted effort that I know of to do (2).

What does the Facebook News Feed have to do with the Price System?

In The Road to Serfdom, F.A. Hayek observes that the price system:

enables entrepreneurs, by watching the movement of comparatively few prices, as an engineer watches the hands of a few dials, to adjust their activities to those of their fellows

Hayek probably has the train engineer in mind, trains seeming a favorite source of examples for mid-20th century economists.  The point is that divisions of labor, which permit more fine-grained and complicated activities within an institution (or the economy at large), require quicker feedback, which the price system permits.  This key strength of the price system is achieved through its compact representation of aggregated information about value.

But the compactness of price information also creates a key weakeness:  price collapses the measurement of value, which in general might have as many independent dimensions as the number of market participants, into a single dimension.  The price system thus tends to work best in markets where the measurements of value have lower dimensionality -- i.e., where different market participants are more likely to measure value using similar input data, with similar means for processing that input.

The price system and its single dimension is at one end of a spectrum with the bartering system at the other.  In the bartering system, there is no aggregation of information about value, and hence no information is lost in the collapse of the real multi-dimensional phenomenon of value into the artificial, single dimension of price.  The bartering system works poorly, however, when the goods are "lower-dimensional" in how their value is measured by market participants.  The extra transactions costs (of time) associated with the bartering system are hardly worth the information preserved.  For example, every holder of wheat does not have a unique perspective on how much wheat (much less that particular grain of wheat) is worth.  Yet something like the opposite is true of transactions involving human capital, such as the market for a potential spouse.

What if there were a more multi-dimensional means for aggregating information about value?  What would that look like?  My suggestion is that it would look something like the Facebook News Feed.

For non-Facebook users, the Facebook News Feed provides status updates and links to multimedia that provide new information about your friends on Facebook.  New information about myself that I add to my profile is automatically pushed out to all my friends, usually showing up as a single line on their News Feed.  Now the value of a friendship with me must be among the most difficult things to measure in a single-dimension.  Yet the value of my friendship, like the value of wheat, is going to depend in part on how many others are offering you the chance to know someone like me.  If all of your friends love to go kite surfing, you may be indifferent to whether I too enjoy kite surfing.  On the other hand, if it turns out that I'm your only friend who shares your taste in medieval English legal history, we may be spending more time together soon.

Notice how the Facebook News Feed falls somewhere in between a pure barter system and a pure price system.  Unlike a pure barter system, I don't have to transact bilaterally with everyone who might be interested in what I have to offer.  Unlike a pure price system, I don't have to submit to my value being collapsed into an almost inarticulate number.

Notice also how the News Feed system requires market participants to internalize the positive and negative effects of their status updates.  For example, by posting photos of yourself at a party, you may get more party invitations, but fewer job offers.  This is how the News Feed is distinct in an important from other social networks, such as Wikipedia.  Professor Sunstein had an insightful post related to this about six months ago.

Now what if instead of collecting information about the relative value of various friends, the news feed collected information about the relative value of various projects within a large organization, or the relative value of different programs within government?

Reducing administrative costs, federal adminstrative agencies, and globalisation

Rees-Mogg and Davidson point out in The Sovereign Individual that the Internet is producing declining returns to violence.  What they mean is that it is harder for governments to control undesirable citizens in a world with the Internet.  Putting aside the problems that the possibility of nuclear holocaust present for their argument (which I may address in a later post), I find this argument believable.  At the margins, some rich or talented people will opt out of citizenship in large nation-states in favor of smaller jurisdictions that demand less in taxes.  A direct consequence is that nation-states will more often be in competition for the "best" (i.e., richest or most productive) citizens.  An indirect consequence is that more economically efficient governments will have a competitive advantage with respect to less economically efficent governments that offer a comparable package of "public goods" in exchange for the same taxes.

As a patriotic American it's natural then to wonder, what aspects of American government could be made more efficient?  Note that the question is not how could we spend less.  For purposes of this analysis, we should assume that we'll come to the same substantive decisions regarding what services (or disservices) the governmment will offer, regardless of how efficent or inefficient the process of decisionmaking and administration of these services.  For example, ending the war in Iraq would save trillions.  But the substantive decision of whether to stay or leave turns on substantive political questions (e.g., "Will Iraq collapse into anarchy?" "Would it collapse into anarchy anyway?") rather than the efficiency of the process that results in the decision, or the efficiency of the process within the institution that oversees the war.

The largest cluster of procedural inefficiencies that I see in the federal government right now are the administrative costs associated with the executive branch.  Looking at the United States constitution, one might imagine the executive branch to be the most efficient arm of federal government.  Even without the idiosyncrasies of the current Bush administration, administrative agencies are the most responsible for something like the exact opposite being true.  Although the legislative and judicial branches are slow and often deadlocked in controversy, they are so by design.  Moreover, they remain transparent throughout their long, drawn-out process.  By contrast, administrative agencies, many of which hold as much substantive power over the public as the legislative and judicial branches, are inaccessible to the public, and run by agents relatively well-insulated from political pressures.

As the United States loses its political market share because of the globalisation of the world economy, there is a relatively easy strategy that we might consider deploying in order to stem the outflow of money and talent: bring federal administrative agencies onto the Internet.  We still have the chance to be a first-mover in this regard, although other countries (such as South Korea) have been quicker to get people Internet access.  Pioneering web-enabled government would gives us an edge in maintaining our position within the global economy.

I will post more on how administrative agencies and corporations might use the Internet to improve institutional efficiency over the next few weeks.