over 8 years on 2010-08-22


in early august i gave a rant at the new york tech meetup responding to clay shirky's new book "cognitive surplus" and generally mouthing off about the power and simplicity of neoclassical economics to explain the changes we are witnessing as a society and help chart the course forward.

with regards to shirky, the four/five line version of my objection is that his book has at its core two central pillars with which i take fundamental issue (along with many other economists looking at the studies he uses to make his points). first, he believes that we as a society have this thing he calls 'free time' (we may have more time, but neoclassical economics would suggest that there is no such thing as 'free' time… just as there is no free lunch). second, he believes that there is a difference between "extrinsic" and "intrinsic" motivations, that can influence each-other in non-fungible ways (hotly contested, and i come down firmly on the other side of the coin).

the real point here, upon which i want to elaborate, is that increasingly it is 100% clear that we face a world with multiple forms of capital. i believe strongly that these forms of capital are exchangeable / tradable and that exchange rates between them are improving with dramatic implications... - so, here is the expanded version of what i believe is going on from the bottom up, the neo-classical version as of august 2010 ~~ thoughts and comments much appreciated:


1. three and only three fundamental things are changing in the world: the cost of storage (remembering things), bandwidth (communicating), and compute (matching things efficiently /discovering new things from memory / etc). at an absolute baseline these three changes should (and do) dramatically lower barriers of exchange leading to more specialization, higher productivity, and more societal wealth.... but i do not believe that we are seeing any other core change in human behavior/disposition/etc.

2. while common wisdom is that these changes should drive huge economic expansion, traditional measures of the economy suggest that we are facing sluggish growth at best: there is a lot of excitement about these three changes, but traditional models suggest that real income in the us hasn't risen almost at all in 30 years, gdp growth has been very sluggish, and the economy of late is totally in the toilet… despite these clear stats, i think i would be in good company making the assertion that it is very hard to argue that the average household in the us isn't better off today than it was 30 years ago, just as is is overly simplistic to say that the total economic impact of google as a service is captured in its ~$200b market capitalization (or some derivative there of). so there is a disconnect here in what is going on, and what we are measuring.

3. so, where is all the fantom growth? it is all non-currency forms of capital: the simplistic view of our economy is that it is currency denominated (and that is what traditional measures of the economy look at). the real answer is that modern-currency is just the most liquid form of capital that human beings earn (think m1). capital also takes on social, political, intellectual, etc. forms which we earn in exchange for our time (our only finite resource) and efforts... the point is that the 'growth' that we are "missing" over the last few decades can be accounted for in the growing percentage of non-traditional currency denominated transaction.

the cash exchange part of our economy is growing rather slowly, but there is a ton of growth occurring in non-cash forms. the mix of capital people are earning is shifting away from cash.

4. why are we seeing a mix shift away from the cash economy? because non-cash capital exchange rates are improving: one possible argument is that these multiple forms of 'capital' each have their own diminishing marginal return associated with them… you might say, we just don't want cash anymore and so we want to earn other types of desecrate capital. i find this totally un-compelling… it is like a little child saying their dinner stomach is full, but their dessert stomach is still hungry. this might be where shirky fits - i say bollox.

rather, i would argue it is because the exchange rate between manifestations of capital used to be just terrible, non-cash forms of currency (like perhaps social capital with friends) wasn't easily transferable into other forms of capital without huge tariffs and fees. so, people earned in the most fungible form of capital, even at a personal utility discount… this is where technology is having the biggest impact -- tech is improving exchange rates between capitals, so people can actually be far more optimized in their lives and get to way higher net-utility outcomes.

5. for example, pretend you are don-draper and you go on vacation to cabo in the 1960s vs. 2015: don shows up on vacation in the 1960s and no one knows him, the only form of capital he can have is the cash in his wallet and the signaling that he has more capital in the form of his dress, demeanor, and dashing good looks… if don shows up on the same vacation in 2015, cash still talks, but upon arrival and facebook 'check-in' friends of people with whom he has a social balance sheet back home and happen to be in town know he has arrived, the owner of the hotel he wants to stay at instantly knows that he is an advertising genius and will gladly trade dinner for some wisdom, etc. don's non-cash forms of capital in 2015 are tradable at low cost and friction in a way they were not 50 years ago.

6. -- by the way, non-cash denominated economic activity is how the world used to work: in the last few hundred years we have used currency as an intermediate step between exchange -- it has some really powerful properties that make that a very good idea in a lot of cases (and also quite useful if you are trying to operate a modern nation-state funded with taxes and need to control the means of exchange), but if you go back further into history you will see that less and less exchange was currency denominated. the point, my bet is that as a percentage of total economic activity, the world hit a peak in terms of the percentage of transaction that was done in a 'cash' form sometime in the 1990s.

7. so, the economy is growing very very quickly in a holistic sense - we are exchanging more and building more utility - it is just in forms that we aren't used to quantifying, and that has some big implications: i will develop the above more later, but if i have you convinced that we are watching a huge mix shift in the forms people earn capital in, great -- if not, humor me for a bit and let's play out consequences and see if they fit. some implications:

-- a. the traditional "un-employment" measure will never come back below 10% and will probably continue to rise, this will be ok we are already running out of cash denominated human intelligence jobs… efficiencies from technology mean they will never come back… this is ok - people can still be highly productive, successful, and happy without traditional employment

-- b. in fact, long term, no one will be traditionally "employed" cost of communicating & cost of memory moving towards zero means that people will be able to switch on/off their cash employment with very little friction, and employers will be able to instantly poll a large market of demonstrably qualified people to perform any hit (human intelligence task) at the market place. anyone who is employed full time will be so only as a financial hedge (by employee or employer)…

-- c. practically speaking -- face-turk: so, facebook "answers" is cool and all, but what is going to be way cooler is mashing mechanical turk with facebook… just like you specify advertisements by segment, i really want to specify questions to the wider audience and put a bounty on them. a clear win, and even more of a win in mobile application form, so i can answer questions for cash on the fly,… this will be far more valuable when the answers to my questions aren't shared as well -- if i pay for information, i want to be able to harvest the value of that information, which means it can't be broadly distributed

-- d. and the tax-base of any nation state is seriously threatened: when non-cash exchanges becomes easy (which it is), then the government looses it's easy ability to levy taxes. the taxable base of the economy (the cash part) becomes a smaller and smaller part of the whole, so taxes go up dramatically on that part… that only causes further flight away from the cash economy. long term, the shift away from currency denominated transaction will threaten the viability of nation-states

-- e. traditional personal privacy is over, because the economic models don't support it anymore: this one is scary to me emotionally, but undeniable and overall a good thing… unless you really fear authoritarian world government (see implications of d)…. you can't store value or trade value in non-cash forms if you keep your information to yourself / if you are private. so, where historically being public only paid if you were a celebrity, now it pays for everyone -- that means people will trade away all their privacy of their own volition.

-- f. practically speaking -- facebook location: perfect win. exactly the right way to go with structured easily parse-able data that creates tradable value. people say that they don't want to 'check in', but ultimately if you don't check in you can't trade the economic value pent up in your location in time. meaning, don draper doesn't get the benefit of his location at any specific if he doesn't declare it. publishing information is how you create value from your information stream, if you don't - that is fine, but you are not able to transact along that axis of value

-- g. so, privacy isn't dead, but it is only for the super rich: if you don't want to declare your location / otherwise are private, that is totally your prerogative, but it is economic value you are leaving on the table/ not putting into a monetizable form. only rich people will be able to afford to do that… so, privacy is absolutely becoming a luxury good.

-- h. and information in general is a monetizable premium good: the value of information is a function of scarcity (which is a function of speed and verity -- since all information is probabilistic and diffuses along a value curve). knowing something that many other people know is worthless in a frictionless society. knowing something that no one else knows is ever more valuable (because the knowledge can be leveraged on an expanding horizon). the trick if you are in the information business, like any business, is to get information at lower prices than you can sell it. there are a few strategies here, but companies like facebook are really in the information arbitrage business… they get people to give them information at low prices (largely information that is very low value to individuals alone) and then aggregate it and sell it to others in a very low friction environment. they actually deal at the very low end of the spectrum, but they find consumers/buyers of information (people reading) for all the sellers (people posting). this new market is god for the sellers and the buyers (and facebook takes a cut in the middle)… but this is another form of an economy that fits along side the cash economy.

-- i. as is refining information to make it higher octane: most people don't consider the full end to end cost of information production through consumption in time, money, etc. to some degree twitter might have understood this (or stumbled into the reality) with the false limit of 140 characters (put the burden on the producer not the consumer to pack value tightly), which makes sense when 1 pushes to 100… good editing of information to make it faster more efficient to consume (utils/$/hour of ingest) and fact checking (higher probability that a given bit of information is correct) can have a huge multiplicative value.

-- j. nlp vs. structure: this will be another post soon, but there are two general approaches to getting more value out of a mass of data, nlp (which is taking language and mining it for meaning) and just changing the format of information to begin with. google is generally an nlp company, facebook is a structured information company…. this deserves its own letter.ly, it will get it sooner than later.

-- k. ... and owning physical assets is will be relegated to the under-privileged: partially because as this shift continues owning cash denominated assets is going to be a huge liability from a tax perspective,.. but mostly because with lower friction/transaction costs you won't need to own things to leverage them on demand, and there shouldn't be core appreciation that you can capture as an individual by owning physical assets (because they should be fully valued and cash-economic growth rates should stay low)

-- i am going to pause here, because i doubt anyone made it this far down. (see earlier letter.ly on editing, sorry)

other codas for another day:

- how does the us postal service account for the liability of 'forever' stamps?: one would hope they aren't just writing futures contracts without the proper financial offset, somehow i suspect that is exactly what they are doing, oy.

- betting on your grades: some college kids are creating an online system to bet on their upcoming grades, seems like a great way to de-risk if your parents pay you for getting good ones.

- the straw is the elevator of spoons: nuf said

- macro problems that are information vs. non-information based: 1. education, 2. food (some), 3. finance, 4. transportation (some), 5. ?law?. things that need massive overhaul and are not information problems: 1. energy, 2. food, 3. transportation

and finally, my notes from the talk i gave earlier this month, if you care: clay shirky builds the concept of "cognitive surplus" as a core framework for understanding a broad swath of activity from lolcats to ushahiti, to fan created charities, and wikipedia and suggesting the impact of the internet in a broad sense into the future. the book is a fun and worthwhile read, but i take issue with the analysis in a few places, and i think that the interpretations matter for how we think about, react, and even shape the clear and obvious change we see around us.

clay's explanation of what he sees as "cognitive surplus" is based on two pillars:

- the first pillar is that we are coming to understand a distinction between intrinsic motivations vs. extrinsic motivations (meaning, we can design around core human generosity which is in meaningful opposition to / tension with neo-classical ego centered economics)

- the second pillar is that technology since wwii has given us a lot of "free time", and with new tools and a new understanding of motivation we now have 1t hours of "participatory time" which we can deploy to make the world better building around intrinsic motivations not extrinsic motivations

- summed up: clay's "cognitive surplus" framework is based on the idea that we have new tools, a lot of time on our hands, and can build things based on intrinsic motivation paradigms as opposed to traditional economic paradigms.

with regards to the first pillar, i believe clay's most vivid illustration is his example of a study by gneezy and rustichini of a set of daycare facilities in haifa. the simple version is that they started fining parents for late pickups of their kids, and instead of the deterrence theory meaning that they stopped picking their kids up late, they picked them up later, and kept picking them up later after the fine was lifted. shirky uses this to make the argument that the fine broke an unspoken social pact by turning the late pickup from a social to an economic sphere... the point being that the study illustrates that intrinsic and extrinsic motivations exist in separate spheres and can trample each other. ...but in the study itself the first explanation offered in the conclusion is actually a classic game theory analysis based on an assumption of perfectly rational and perfectly selfish individuals. the framework didn't change, just the payments.

the same thing is true of a second study shirky cites by edward deci which studied intrinsic vs. extrinsic motivations and showed a supposed 'dampening effect' of intrinsic motivations by external motivations as shirky explains it (this is the question, do you pay kids for good grades)... one read of the hotly contested study is that this dampening effect exists, but there is a whole literature finds in the other direction... which to his credit shirky does acknowledge, though i believe he dismisses too quickly.

in short... there is an argument that there are two distinct spheres of intrinsic and extrinsic motivations, but it is far from the only viewpoint. there is a strong camp, of which i wold consider myself a member, which believes that a unified framework based on neo-classical economics is correct

with regards to the second pillar, "free time", clay says that we have a lot of free time which we can use on these intrinsically vs. extrinsically motivated projects..... i agree that we have a lot of "time" on our hands. that has been the march of technology since we started farming. technology is actually in many ways a machine for manufacturing time, which makes a lot of sense given that our only definitively finite resource is time. whether or not that time is 'free' or not is a different story, i would argue that it is not 'free' just as there is no such thing as a 'free' lunch.

clay speaks to the point that we used to watch a lot of tv, as we watch less tv we are freeing time, which opens up cognitive surplus. totally agree that we have a lot of social hours not employed in the cash economy... in fact, we are continuing to watch work weeks erode, retirement expand, and employment numbers drop (likely not to ever recover)...

why i care / am taking 5 min of your life, we live at a time of very exciting change. some of the projects that clay outlines are indeed quite interesting... but to distinguish free time from all the other time in our lives, and to bracket intrinsically motivated projects from extrinsically motivated projects i believe misses the point and confuses not just economic but also social issues.

an alternate view:

- human beings are neo-classically rationally self maximizing, intrinsic and extrinsic motivations are fungible
- we have multiple forms of capital for which we can optimize
- those forms of capital have exchange rates between them and differential liquidity, those exchange rates change with friction
- there is no such thing as a trillion hours of 'free' time, actually there is just time and utility.

the key is that you can think about actions and reactions on traditional axis utility curves, people make good rational trades, capitalism is good.

original swl blogposts and letters 2007-2010