Where the bias is in the innovation debate

Robin Hansen made a point (in a debate on AI) that people who are thinking about the progress of innovation should take to heart.

"The literature on economic, technical, and other innovation says most value comes from many small innovations – more useful and wider scope innovations are rarer, and usually require many small supporting innovations."

People who are skeptical that there has been much innovation recently should realize that in many cases they are missing real improvements not because they aren't happening, but because they are occurring in small amounts over many different areas. Sure, they don't have flying cars*, but different processes are being optimized by engineers everyday.  There are people who blindly trust that society will keep innovating, and in some areas the institutions that typically drive innovation might be sufficiently broken that this blind trust is incorrect. However, those who believe that people will keep figuring out ways to make their lives better are more right than wrong.

*Incidentally, many of the people who use "Where are the flying cars that we imagined we'd have in the future?" as a rhetorical point about innovation in our society would also be quick to dismiss flying cars as a true technological innovation if they were actually common. After all, as the rhetoric goes, they aren't a new technology, just a combination of two existing technologies that have been around for a long time.

Important Links

1. Seth Roberts on healthcare stagnation. Life expectancy dropped by 0.1 years in the US from 2007 to 2008, and no one seems to care. This is less of a statistical fluke than people would like to think and therefore it is more important. This is especially important because demographic trends mean that healthcare spending is going to increase regardless of the spending's actual impact. HT: MR

2. An interesting perspective on Greece's economy. The discrepancy where civil workers are paid much more than private sector workers is very damaging to an economy in the long run. HT: Fistful of Euros

3.  The jobs picture is unlikely to return to the previous peak if the current expansion is the average 59 months. This isn't even accounting for the increasing population. The employment to population ratio is not quite returning to the levels it was at before women entered the workforce en masse, but it is getting close.

4. Perhaps just as important: Using the right bike seat. I've been told that being able to grip the seat with the thighs while going down bumpy hills is important, but the potential side effects of the normal bike seat are a large price to pay.

Republican Presidential Probabilities

Prediction markets are useful. They give us a way to determine what the market is pricing in about any number of events. With the 2012 campaign starting, it will be one of the easiest ways for a casual observer to avoid the noise and see who people think is actually going to win.  Intrade is pricing in a victory for the democrats in 2012 at 60%, and for Republicans at 38%.  They are also pricing in a 1.5% chance of another party winning, which considering which party is more likely to fracture may partially explain why the chances for Republicans are so low. 

More interesting than the broad probabilities is what you get when you combine the probabilities relating to individual candidates.  The three categories of "Republican primary presidential candidate," "Republican party wins the presidency" and "Individual wins the presidency" has an interesting relationship.  If every candidate has the same probability of winning the presidency after winning the primary, P(Primary winner)*P(Republican wins) = P(Individual Winner).  However, this isn't always the case and the chance that a specific Republican will win can be priced in from the other two contracts..

That table shows the probabilities for different Republican candidates and the implied probability of their chance of winning the election if they win the the primary.  The data should be taken with a grain of salt.  Ron Paul enthusiasts bidding up his contract are probably responsible for his seemingly high chances of winning a general election. Wider bid/ask spreads and illiquid markets also mean that fast moving markets such as Rick Perry's probably won't yield an accurate market view. Still, it is interesting that Sarah Palin and Michele Bachmann are priced in at levels where they would hurt the Republican's chances to win the presidency while Jon Huntsman and Rick Perry are priced at levels where they give the Republicans a better chance of winning while the rest of the candidates are clustered right around 40%.

The "chance of winning" is something people talk about during the primaries so it is nice to have a market mechanism that provides these numbers. However, if it gets too popular as a measure without the market getting more liquid then the market might end up with numbers as skewed as Ron Paul's.  

Evidence of innovation

A lot of people have been talking about the need for innovation.  It's easy to think about about inventions that haven't been made yet, and new trends that have a noticeable impact on industry sometimes get looked over by people who are worried about the lack of innovation.  The extraction of natural gas from shale is an important new technology*. Extraction of natural gas from shale has increased 14 fold over the last decade.  Before this widespread implementation, natural gas in the US was going to get very expensive. Now, projections by the EIA find that the United States' net imports of natural gas will shrink to 1% in 2035 from 11% in 2009.


This isn't necessarily the innovation that people want. At the 2011 Ira Sohn conference, Jim Chanos discussed how cheap natural gas was killing the ideas of wind and solar power as economically viable enterprises.  Natural gas will be seen as an old technology, and people complaining about the lack of technological innovation might opine about how alternative energy hopes are mere pipe dreams. If they do, it should be pointed out that these pipe dreams are being crushed by innovation in an older sector setting too a high a bar for the newer technology.

There are still concerns about this method of extracting natural gas.  James Hamilton at Econbrowser points out how these natural gas wells are getting methane into drinking water. These negative externalities need to get worked out, but overall the new production technology has been driving down the price by enough that the benefits should vastly outweigh the costs.

The chart above gives some idea of the benefits.  The orange line is crude oil, while the white line is an average of the first 12 natural gas contracts.  Natural gas has seasonal price movements, so it is better to look at the average of the contracts over a year rather than a generic front month contract.  Note how the new supply coming online from shale gas helped natural gas decouple from crude oil and prices stayed low despite the economic recovery.  

*It was developed a while ago, though the technology is still relatively new. There has been some learning by doing, and perhaps more importantly, new reserves found where the technology is applicable.

Dictatorial Incentives

When it comes to running a country, non-hereditary dictators already have among the worst incentives. They are there to get what they can from their country while they still have control. They are going to be more short term oriented than a monarchy and much less responsive to the immediate well being of the people than a democracy. When it looks like they might lose control their incentives are even worse.

The recent news about Mubarak sends a powerful message to dictators.  "Don't publicly kill your citizens. Don't enrich yourself too much at their expense.  But if you have already done these things, you might as well fight to the death of your citizens and yourselves because it isn't likely that there will be a safe way for you to give up power."

These people have done horrible things, and from Egypt's perspective this prosecution makes a lot of sense. It attempts to quell the recent surge in unrest by bringing people who have done very bad things to justice.

But if Mubarak, a leader who voluntary stepped down, is being treated like this then other dictators are going to think twice before following his lead. They have very little incentive to surrender. Assad and Gaddafi probably don't see any way out at this point*, so they have no choice but to dig in. In Yemen Ali Abdullah Saleh might have changed his mind because he wanted to keep power over the promise of immunity, but it is also possible that he didn't think that he would have real immunity.

Before extending this simplistic analysis too far it should be noted that many dictators are crazy and won't always respond to changes in incentives (and even if they were sane, one particular person will not always respond significantly to a change in incentives). Still, systemic policies that stack the deck further against a good outcome are not desirable.

We can just be thankful that none of these misaligned incentives are immediately significant in the countries run by dictators who have nuclear weapons. 

*At this point it is questionable whether they should even be offered a way out. Preventing future bloodshed needs to be weighed against bringing past wrongs to justice.

Do normal Bloomberg users not use data?

These days, people who buy new PCs running windows will be purchasing a machine running Windows 7 64-bit. I don't have strong opinions on operating systems so this seems fine, except Bloomberg's excel API does not support 64-bit applications. When asked they don't give a data at which they'll start supporting it either.  So either no one in finance is buying top end new computers that benefit from the 64-bit architecture or only a very small percentage of users actually download Bloomberg's data into excel files for more complex analysis so Bloomberg hasn't felt enough customer pressure to make sure that their product is fully compatible with the majority of new computers.

Either way, it's really weird.

Freakonomics without the economics

Freakonomics was a fun book, and it got a lot of people interested in economics that wouldn't have been interested otherwise.  The podcast based on the approach taken in the book has highlighted a lot of very interesting research and has been a pleasure to listen to.  The writer Stephen Dubner runs the podcast and his co-author Steven Levitt occasionally co-hosts.

A recent episode "Death by Fire? Probably Not" shows what can happen when the economist is left out of the show. It was a podcast about fire safety done by someone completely unfamiliar with the economic concept of opportunity cost.  A new California regulation that went into effect this year requiring sprinkler systems in new homes was applauded as a natural step of fire safety regulation instead of analyzed from a cost benefit perspective. The additional cost of $4000 to $6000 dollars per house was not mentioned. With a national fire death rate of 13.2 deaths per million people, even if these regulations brought the fire deaths of those inside these houses down to zero it would be doing so at a cost of hundreds of million of dollars per life saved.

It's bad when normal journalists misses some basic economic idea that is important to their story, but it is worse when a journalist who has spent the last few years writing books about economics does it. 

For people interested in a fun economics podcast Freakonomics does have a lot of good episodes. However, Dan Ariely's Arming the Donkeys also covers economic topics and does a better job of challenging the ideas of its guests.

Invest in Education!

"Invest in education!" It is the rallying cry of many who think that more education spending is the silver bullet that is going to solve most of America's problems. However, the bullets seem to be hitting student's balance sheets more than America's other problems. In the past year, Mark Kantrowitz of FinAid.org claimed that the total education debt outstanding has passed the total credit card debt outstanding. Unlike credit card loans, these are loans that cannot easily be discharged in bankruptcy. These loans will follow students for the rest of their lives, potentially going so far as to allow lenders to go after tax refunds or a person's social security payments.  This is a debt overhang that cannot be solved by jingle mail*.

The confusing thing about education is that it is partially a private good and partially a public good. An educated person is be more likely to promote positive aspects of society by being a well behaved and productive citizen. Basic education is obviously very valuable, because a literate population is capable of many more things than an illiterate population. It is also necessary for most people to be trained in more advanced skills, since there is a level of educational attainment that it is important for a developed country to reach. In the past, economic growth has been associated with increases in education. Economists have documented this trend and like many forecasters they assume that the relationship that has occurred in the past will continue to hold into the future. However, the nature of education as a private good might be causing this correlation to breakdown. 

Education is also a private good because it serves as signal to other people of the student's qualities, and that signal only benefits the person holding the degree at the expense of those who don't hold a degree or those who hold a lesser degree.  Bryan Caplan suggests that the qualities signaled are work ethic, intelligence and perhaps even the conformity required for an employee to be a productive corporate worker. If education is being used mainly for signaling purposes, the limited number of degrees being given out by top colleges would be a feature and not a bug.  The value of adding on more slots at lower tier colleges is also low because lower tier colleges generally price themselves similarly to top tier colleges so that they don't look like they are lower tier.  

An important question is whether or not education is more of a cause behind students becoming more productive students or if the kids who are selected to become students were already going to be more productive in the first place.  One way to measure this is to look at the demand for university material by non-students. If many non-students thought they could receive most of the benefit of an education by attending lectures given at the top colleges and doing the work on their own time then we would see more security outside of lecture halls keeping non-students out.  ITunes would also be able to sell their lecture series for nontrivial prices that they are now giving away for free at ITunes U.  Another question is how much additional spending directed towards education helps. When inflation in education has increased at four times the rate of inflation since 1986, it should make people stop and think before deciding throwing more money at the problem is the obvious answer. 

If education is more about signaling than teaching, then it means that as spending increases most of that spending is merely being wasted on an inefficient status competition.  Perhaps this is why Peter Thiel has been calling education a bubble.  One way he is drawing attention to this bubble (while at the same time cleverly expanding his network for recruiting talent) is with his "20 under 20" program. This program involves paying 20 students to leave college for at least two years. One of the ingenious things about this program is that the applicants are being promoted as very elite students with articles about the program specifying that many of the applicants are from top colleges. If this program is repeated year after year and becomes fixed in people's minds as one of the most elite programs out there for aspiring entrepreneurs it should give the winners the status they were hoping to get from their college diploma. They will only go back to college if they think that they will really benefit from the education.

Those who agree with this general perspective and think bubbles should have direct tradable implications should review Steve Eisman's May 2010 Ira Sohn conference presentation about the for-profit education companies.  If there is a subprime of the education bubble it is probably the for-profit education companies. For-profit education has the unfortunate distinction of being among the lowest status providers because of their willingness to take almost any type of student in order to make a profit.  If people wake up to the idea that most of the value in colleges comes from increasing a person's relative status** and that these colleges are at the bottom of the totem poll, then the for-profits will be in trouble.

It's nice to think that education is a race to the top that everyone can win, but we shouldn't ignore that once we get near the top the real game starts to look more like king of the hill.

*The situation is more complicated than this for many of the people who took out multiple loans on their houses. It will also be hard for some of these people to walk away.
**Certain government workers who are merely trying to get additional certification in order to quality for higher pay might be the last to care until their compensation system is changed.

Links: The "Some people on the internet are wrong!" Edition

1. Naked Capitalism's april fools prank came 3 days late. Just to consider one of the most glaring errors of the letter, the author seems to have no idea that many academics earn money based on the value of their connections and knowledge from their governmental positions. Thus, their main patrons would be factions that might end up running the government. See Elizabeth Warren's use of weird metrics in order to make politically charged studies which led to her current position as a special advisor to the Consumer Financial Protection Bureau.

2. One of the bloggers at the Economist has no idea how hedge funds work. They are unable to tell the difference between a decrease in assets due to investors pulling money out of a fund and negative returns. To be fair, the blogger is apparently based out of Kansas City.

3. In one of the recent episodes of "To the Best of Our Knowledge", Lone Frank brought up understanding the "Coke/Pepsi" example as one of the main case studies of neuromarketing. The Coke/Pepsi example is derived from the Pepsi challenge, where the less popular Pepsi is preferred in blind taste tests, but Coke is preferred if the drinks are labelled. This leads neuromarketer proponents to believe that there is something special about the branding of Coke or memories associated with the brand that makes it more appealing than Pepsi and that this difference can be seen directly via brain scans.  However, these neuromarketers are missing a key fact that was pointed out in Malcon Gladwell's book Blink. Coke tastes better in the long run because the sweetness of Pepsi gets more annoying over time. Given this data, the experimental subjects used by neuromarketing analysts are probably associating the brands with their memories of consuming the actual product.  Advertising and other fond memories associated with a brand should also have a measurable impact on the brain that would be interesting to study in more depth, but it is kind of sad that the neuromarketing flagship study* is based on false premises.

*This label is based on how it was the one study that was brought up on the podcast and because it is the only study mentioned in detail on the wikipedia page for Neuromarketing.

Who Predicts Well - Does it apply to financial markets?

What makes a good predictor is a very interesting subject to those involved in financial markets.  Tyler Cowen highlights an excerpt from a piece by Dan Gardner and Philip Tetlock:

"…what separated those with modest but significant predictive ability from the utterly hopeless was their style of thinking. Experts who had one big idea they were certain would reveal what was to come were handily beaten by those who used diverse information and analytical models, were comfortable with complexity and uncertainty and kept their confidence in check."

Having one big idea might be a dangerous thing for money managers, but it seemed to work for some people who shorted the housing market.  The way to reconcile these two facts could be that the people who made money shorting the housing market weren't applying their one big idea about the future downfall of America or the idea that easy money is going to destroy the economy, but instead were the type of people who came to their conclusions about the housing market after studying various metrics and utilizing different analytical models.

It should be noted that just because the experts who are supporting their one big idea might be worse predictors in general, they may not be wrong in the long run. Those who favor a long gold strategy tend to have one big idea: You can't trust fiat currency (particularly the fiat currency of the United States), as throughout history all fiat currency has approached zero.  They see every sign of inflation as a sign that gold should go up in value and every sign of deflation as a sign that gold should go up in value.  Even though they have the wrong cognitive approach, over the past 10 years they've been right far more often than they've been wrong.  If someone marginally familiar with Tetlock's research decided to sell gold because they didn't like the arguments of people buying gold they would have lost a lot money.