EDM

In the next ten years, I predict that all cars will be equipped with electronic driving monitors (EDM) that will provide feedback on your driving style and efficiency as well as provide insurers information on how safe you drive. Initially the driving monitors will be optional and insurers will offer discounts to people providing driving data.

So thereafter, tax payers will be convinced that valuable money and police person-power is saved by not needing to enforce traffic laws like speeding, stopping at stop signs, etc. if EDMs are required.

http://blogs.zdnet.com/BTL/index.php?p=1499
http://ezinearticles.com/?Ensure-That-Your-Teen-is-Driving-Safely&id=921264
http://casr.adelaide.edu.au/T95/paper/s24p3.html
http://www.theregister.co.uk/2005/11/15/vehicle_movement_database/

Cost of Living Index

The cost of living is a funny thing - two people may make the same amount of money in two different places, but the cost of living can make the real value of their incomes very different. Take, for instance, the cost of living in Ottawa, Canada versus Austin, Texas. Based off of data I pulled from the web, the cost of living in Ottawa (75.6) is 64% more than in Austin (46.2).

As an arbitrary example, if you made $50,000 in Austin, you would need to make $81,818 in Ottawa to have the same standard of living. ($108,000 in New York City (100)) Or, put another way, if you made $50k in Ottawa, you would have the same standard of living as someone making $30,555 in Austin.

Career Advice

If you are a young 'un in Central Texas and you want to have a high paying job, you could do worse than to visit
http://socrates.cdr.state.tx.us//iSocrates/occprofiles/profile_select.asp

You can sort all the careers in Central Texas by expected growth, salary, etc.
Here, for example, is a list of careers that have over 100 new openings per year and make over $80,000 per year.

Cluster Flock

We are all individuals, but wouldn't it be funny if someone could predict what you are going to do tonight - what you are going to eat, read, watch - and know what you value most in life?

e.g.:
Andy pulls into his row house in his Honda Civic. His wife he helps his wife prepare dinner and take care of their 4-year-old while listening to a classical music radio station. She goes to the gym while Andy puts their daughter to bed and reads a developmental economics book, "The Bottom Billion". Tomorrow will be his turn to go to the gym.

I've been thinking about consumer profiling (clustering) with the ever increasing availability of consumer data... specifically, the "analysis of consumer lifestyles to create a detailed customer profile... [these results] are combined with geographic (place of work or residence) and demographic (age, education, occupation, etc.) characteristics to develop a more 'lifelike' portrait of the targeted consumer segment."

You can (if you fall in to one of the top groups) check out your own profile by (US) zip code at http://www.claritas.com/MyBestSegments/Default.jsp?ID=0&SubID=&pageName=Home

A great many products are moving towards this predictive profiling: Google (Sponsored Links), iTunes' Genius, Amazon, other online booksellers, and TiVo Suggestions ("Users can rate programs from three "thumbs up" to three "thumbs down." TiVo user ratings are combined to create a recommendation, based on what TiVo users with similar viewing habits watch.")

I believe we will soon be seeing much more sophisticated "suggestions" software that pulls from larger data sources and has an almost "Ah, ha!" marketing appeal. A new era of direct marketing is on its way that will have us pegged so accurately that we will be uncertain of our own opinions ("Is the calculator wrong? Maybe Excel didn't add up those numbers right...").

Politics are very closely linked with marketing. While micro-attention to the polls has been around for several election cycles, I believe a much more detailed voter profile is coalesing which will allow for much more effective hooks. Modern and future election campaigns know what people care about, what media they use, and what they had for breakfast...

Political return on investment

Here are some of my views on the economic stimulus package. There are three main parts to this issue: the core issues, inflation, and politics.

The aim of the short term stimulus is to get people to spend more. Consumption is down dramatically, i.e. the market for things that are produced is down. The question is, how do we best get people to spend money? Should the government spend money on contracts which get companies to hire more people and make CEOs richer? Or should the government give everyone a couple of thousand dollars? Which method would cause inflation to increase faster?

Is consumer spending really the core of the problem? Why are people spending less? Do they have less money, less credit, less confidence about the future? Contrast today with the Monetary Policy Report submitted to the Congress on February 11, 2004, "Uncertainty in financial markets had declined, and rising consumer confidence and a wave of mortgage refinancing appeared to be supporting consumer spending." If consumer spending is the problem, then the deeper problems are uncertainty in financial markets and the crash in the value of equities and real estate. These are the issues that need to be addressed

Returning back to the question of inflation, is there inflationary slack in a depressed economy? Normally giving everyone money would simply drive up the price of everything, but if supply exceeds demand, retailers would be happy to get rid of their excess inventory without raising prices much. But after they spend their money, would consumers continue to spend? Would the core problems continue to persist?

What about the politics of government spending? The current bill is widely acknowledged to be loaded with special interest spending. In fact, the political return on investment for this crisis opportunity is far too much to be passed up. Never before in history has a political party had the opportunity to spend so much money to promote its interests. Would it be possible to forego this opportunity - the nation is crying for leadership, something must be done, and the precedent for spending is already established.

In short, my view is that the government spending option is mostly political and will be much less effective than a tax rebate along with serious financial market reform.

Road Kill

Some people say population control is the key to saving the world. US policy makers are putting into effect laws which will kill off nearly 2000 Americans a year. One of these could be you or me, but hey, what can you do? The key to the key is to disguise the murders in something morally upstanding.

Save the world by increasing fuel economy! The "unfortunate" result will be a 4.4% increase in miles driven and/or number of people killed. Since roughly 42,000 Americans die a year in traffic accidents, a 4.4% increase equals nearly 2000 additional deaths.

I think this plan has merit!

- or -

maybe we could use tolling...

http://freakonomics.blogs.nytimes.com/2009/02/03/the-rebound-effect-of-higher-mpg/

http://freakonomics.blogs.nytimes.com/2009/01/06/why-youll-love-paying-for-roads-that-used-to-be-free-a-guest-post/

http://freakonomics.blogs.nytimes.com/2009/01/08/why-youll-love-paying-for-roads-that-used-to-be-free-part-two/

Probabilities

"What are the chances, eh?" asked the cop. "... that I'd pull you over twice in one week?"
Considering the fact that as I waited 10 minutes for him to run up the ticket, I watched well over 100 cars go through the stop sign without stopping, and that I had not gotten a ticket in many years - I'd say the probability is pretty low.

At a party this weekend, someone brought up mbti (personality type) and I reacted with determination to piss all over it. I explained a central fallacy with the idea that traits are normally distributed, so the probability of being a particular "true" type are vanishingly slim. I've written about this before, but I feel like makeing a correction and adding a little explanation. (And what is the probability that someone will actually read both entries?)

In a normal distribution, 68% of the values will be within one standard deviation of the mean. In regards to personality type, 68% of people are normal for a trait and only (100% - 68% =) 32% are "outside of the norm". For two traits, only (32% x 32% =) 10% of people are "outside of the norm" for both traits. For three traits, only 3%; and for four traits, only 1%.
So, only 1% of the population is actually a "true type", (e.g. intp, esfj).

Traits Probability
0 68%
1 32%
2 10%
3 3%
4 1%