Tuesday, January 31, 2012

Sticking it to the man

Is there a natural tendency (some more than others) to stick it to the man, whomever this might be? This week/month, ‘the man’ happens to be Google, and for all its transgressions in the past and all it plans to commit in the future I’ve decided to switch my home page from Google to Bing. Ironically, 10 years ago I probably would have designated Microsoft to be ‘the man’.

There are a few grains of truth to these

In order to make many we have to lose money
In order to win big, we have to be willing to lose big
In order to create jobs, we have to destroy jobs
In order to build you up, we first have to tear you down

I'm just not sure how much.

All triggered by this MR post.

Monday, January 30, 2012

Stolen phones

With the increase in crime and along with it - stolen phones, I am puzzled that even with built-in GPS, the police can’t track the phones - like, er, well, on the TV shows, where the perp is tracked down by cell phone pings. Apparently this only works on TV for the most part and even Snopes themselves are confused over what can and cannot be done with built in GPS. The phone companies seem to hold the key and they are unwilling to make cell phone thefts unprofitable by making the phones traceable or simply locking out the device.

Just  one day after an 11-year old was robbed of her smart phone right outside Alice Deal Middle School in Northwest DC, it has happened again.

On Wednesday, another young student had her phone taken while she walked through Fort Reno, right by her school. In both cases, police say a young man ran past the victims, snatched the smart phone and took off running. Investigators say the two incidents could be related. Police say this is part of a larger trend where crooks of all ages are stealing smart phones and selling them on the black market.

DC Police Chief Cathy Lanier said smart phone thefts are up across the District, and she isn't getting any help from the phone companies in battling the problem.

Lanier spoke to a group of citizens this week where she said, "I spoke to the head of security at Apple and said, can't you disable these phones permanently when they are stolen to take the profit out of the market for the bad guy? They said, well we can but we're about customer service, and we like to give customer service when they call."

Along with this, what I would like Siri to be able to do - or what I would like to be able to say to her if my iPhone were stolen (if I had an iPhone, that is): “Active self destruct sequence.”

Thursday, January 26, 2012

Intermediate outcomes

Andrew Gelman’s post on the “fight” between Martin Lindquist and Michael Sobel on one side and Judea Pearl and Clark Glymour on the other is strictly for those who are familiar with directed graphical models (DGMs) which also include certain types of structural equation models (SEMs). But this is the second post he has made of late on intermediate outcomes. (Here’s his other post.)

I have nothing to add really except that the phrase intermediate outcomes reminded me of the book Because A Little Bug Went Ka-choo which is full of intermediate outcomes.

Teaching and STEM

One of MR’s top post touts the payoff from majoring in STEM fields (science, technology, engineering, mathematics). Yet as this NYT article states, majoring in STEM is just too damn hard. Moreover, Catherine Rampell notes it is less work to just to pick a non-STEM major - by about 4 hours a week. (From someone who has spent about that much time a week in Organic Chem lab for one semester, I can attest to that. Not only that, I came out smelling of chemicals which hot showers didn’t always seem to be able clean out. Needless to say, OChem wasn’t for me.)

The highlight of the NYT article however was the following (emphasis mine):

“We’re losing an alarming proportion of our nation’s science talent once the students get to college,” says Mitchell J. Chang, an education professor at U.C.L.A. who has studied the matter. “It’s not just a K-12 preparation issue.”

Professor Chang says that rather than losing mainly students from disadvantaged backgrounds or with lackluster records, the attrition rate can be higher at the most selective schools, where he believes the competition overwhelms even well-qualified students.

“You’d like to think that since these institutions are getting the best students, the students who go there would have the best chances to succeed,” he says. “But if you take two students who have the same high school grade-point average and SAT scores, and you put one in a highly selective school like Berkeley and the other in a school with lower average scores like Cal State, that Berkeley student is at least 13 percent less likely than the one at Cal State to finish a STEM degree.”

Several thoughts come to mind:
1. If majors are a way to signal ability and I think that this is partially true, do we want to turn major choice into policy? My knee-jerk reaction is no because if it is a signal of ability and we subsidize entry into STEM then the average quality of STEM majors would fall.
2. Peer effects of being in a selective school seems to indicate that the two people with the same GPA and SAT score are not really the same people. One is higher ability than the other if the one in the selective school succeeds. I use ability in a very general sense to encompass resilience, persistence, grit, motivation or whatever quality that allows them to succeed.
3. The article also notes that it is possible that the problem lies in the way selective schools teach STEM (among other reasons). I’d say yes, and not just in terms of STEM but college teaching in general.

This last point is made by Larry Summers (HT:MR) though he does not address STEM specifically. The value of group rather than individual effort could perhaps stem (pun not intended) the outflow from STEM.

…  collaboration is a much greater part of what workers do, what businesses do and what governments do. Yet the great preponderance of work a student does is done alone at every level in the educational system. Indeed, excessive collaboration with others goes by the name of cheating.

One approach that has not been getting much press of late is the concept of peer teaching. Ironically, as the article notes, the leading pioneers of this method of teaching are STEM professors. But the hardest thing to change is:

... getting ... other professors to stop lecturing will be a hard sell. Change is slow in the academy, and professors tend to be rewarded for focusing on their research, often at the expense of their teaching.

Wednesday, January 25, 2012

Social disorder

The recent crime wave in the usually staid and wealthy upper NW neighborhoods of DC (see one of several stories in WaPo here) has a feeling that the stakes have just been upped. What started as a series of thefts from vehicles and breaking and entering have entered into a phase of armed robberies and home invasions. The police say that the criminals who commit these crimes are very different in the sense that someone who does a smash and grab is different from someone who would commit a home invasion. In the parlance of economists criminals specialize based on their comparative advantage.
While there may be an element of truth, it seems that one crime is just a hop, skip and a jump to another - it is easy for breaking and entering to become home invasion, armed robbery and murder. After all it is straightforward to acquire the tools and skills. It is whether they have the nerve or the guts to grab a gun and start pointing it at people and perhaps even using the weapon. It is whether they have the audacity to simply invade people’s homes by getting together a group to just smash windows and barge in shooting and for lack of a better word, pillaging the homes.

After all, social order is very much voluntary. We all pay our taxes - voluntarily. If we suddenly woke up and decided not to do so there is very little that the IRS can do. They simply do not have the manpower to to enforce tax laws on everyone. Sure as they say, you may get a few of us but you won’t get us all.

In a similar vein if a proportion of the population decided to start committing crimes, there is very little that the police can do. Sure, they may get a few of them but in the end as there are more of those who decide that it is profitable to commit crimes enter the market so to speak, they can overwhelm a civil society. In a sense there is point at which everyone starts to rebel and there is little that can be done. Think of how riots can erupt, for instance.

Yet like riots, the madness for lack of a better word dies down - somehow. So hopefully, this crime wave is what it is - just a wave.

Tuesday, January 24, 2012

Reinventing yourself

Over the course of three weeks in July, the Manchester International Festival presented newly commissioned works by 28 artists known more for challenging audiences than for pleasing them... The whole program was intended to be, according to festival director Alex Poots, a celebration of “risk-taking, the pioneering spirit, and valiant attempts.”

Dickson Despommier ’64PH, a Columbia scientist best known as the progenitor of a concept he calls “vertical farming,” fit right in. He had been invited to announce that the nonprofit organization that runs the biennial festival will soon create a towering multilevel greenhouse inspired by his vision: Alpha Farm, to be located in an abandoned eight-story office building in Manchester, will hold several floors of broccoli, lettuce, tomatoes, onions, carrots, and strawberries, all cultivated beneath huge banks of lights.

A microbiologist by training, Dickson Despommier spent the first 30 years of his career holed up in a Columbia laboratory studying a tiny parasitic worm called Trichinella spiralis. His major contribution was to describe how this worm, which thousands of people ingest every year by eating undercooked meat but which rarely causes serious illness, can survive in our bodies for long periods by burrowing into muscle tissue.

In 1999, the National Institutes of Health decided that his research had run its course. “I lost my grant support,” Despommier says. “At age 60, I was faced with reinventing myself.”

But at least he still had a job from which he could reinvent himself. Most would have just gone through the motions of tenure - teach as little as possible and hold out until retirement.

He then began studying how infectious diseases take advantage of certain environmental conditions. ... He also created a new graduate course at Columbia’s Mailman School of Public Health exploring how the use of chemical pesticides and other ecologically disruptive farming practices affects human health in subtle ways, such as by altering the relationships between parasites and their hosts at various levels of the food chain.

Halfway through the semester, the students rebelled against what they considered their teacher’s relentless negativity. “They said, ‘Okay, we get it. Farming is bad for the environment. So why don’t we figure out how to improve it?’” Despommier recalls. “Their idea was to grow crops right in Manhattan, on rooftops.”

Despommier indulged his students, assigning them a group research project in which they were to calculate how much food could be grown if every available rooftop in Manhattan were made into a garden. “It was a purely theoretical hypothesis, really, to imagine that every property owner could be enticed to plant crops on his or her roof,” says Despommier. “But it would test the limits of their idea.”

The students, after poring over maps at the New York Public Library, became discouraged. Even if every square foot of flat, undeveloped rooftop space in Manhattan were devoted to growing rice — an unusually nutritious and calorie-dense grain — the resulting harvest, they estimated, could feed only a tiny portion of the borough’s residents. “This told us that rooftop gardening, for all its benefits in absorbing carbon dioxide and stormwater,” Despommier says, “wasn’t going to be much of a solution.”

The idea of urbanites growing their own food had captivated Despommier’s imagination, ...

My tooth WAS handled!

An amusing post title (“You can’t handle the tooth!”) from MR kindled some memories. The post points to how some tasks can be handled by dental assistants much in the same way that physician assistants (PAs) now take over some care giving from doctors and how this is possibility is being opposed by the American Dental Association (in the same way the AMA fought against PAs.

When I was growing up all my dental needs (and there were many, but more on this below) were taken care of by women (usually from villages) who had come to study to become - well I don’t know what they were studying to become but they were being trained to clean, fill cavities and all other dental care (except extraction) and they were supervised not by dentists but by other graduates of the program. (And yes, they were all women who lived in dormitories near the college. I hesitate to call it a dental college but for all practical purposes that was what it was. I don’t know if this is still the case though but the college still exists.)

Not only later when I was in the United States did I learn that only medically trained dentists were allowed to fill cavities. My two opposing thoughts were and still are:
1. They need doctors to do this!?
2. Was I getting substandard care while growing up?

I had more than my share of cavities I feel and up till today I can’t be sure if it was my bad dental hygiene or the possible screw ups that happened while I went to the college for dental care. (Yes, it was free but at the cost of being a guinea pig.)

Friday, January 20, 2012

Belief falsification

I’ve been skimming Timur Kuran’s “Sparks and Prairie Fires: A Theory of Unanticipated political revolution” and I probably should take more time to digest it. The element of the model that didn’t really appeal to me was the seemingly exogenous shift in the threshold function that was needed to explain revolutions or to set the stage for a revolution. In many ways the model needs to be extended to include feedback effects as more agents reveal their true preferences which in theory should shift the threshold function. In any case, I say seemingly only because I haven’t taken the time to digest the model and was too busy drawing parallels with financial crises.

Here’s the abstract from Kuran’s article:
A feature shared by certain major revolutions is that they were not anticipated. Here is an explanation, which hinges on the observation that people who come to dislike their government are apt to hide their desire for change as long as the opposition seems weak. Because of this preference falsification, a government that appears unshakable might see its support crumble following a slight surge in the opposition's apparent size, caused by events insignificant in and of themselves. Unlikely though the revolution may have appeared in foresight, it will in hindsight appear inevitable because its occurrence exposes a panoply of previously hidden conflicts.

Here’s my thought on its relevance to financial crises and hence the title of the post:
A feature shared by major financial crises is that they were not anticipated. Here is an explanation, which hinges on the observation that people who believe that the economy is experiencing a bubble hide their beliefs either because of reputational or incentive effects or fear of retribution by their employers. Because of this belief falsification, an increase in stock prices that appears to be based on fundamentals might see its foundations crumble following a slight shift in sentiment, caused by events insignificant in and of themselves. Unlikely though the crash may have appeared in foresight, it will in hindsight appear inevitable because its occurrence exposes a panoply of previously hidden conflicts.

Remember Henry Blodgett who privately called a stock a ‘dog’ while rating it a buy publicly?

Thursday, January 19, 2012

Are sociologists more ethical than economists?

I was surprised to read in the recent debate about the economics code of ethics at the AEA meetings that the code of ethics would be modeled along the lines of (gasp!) the American Sociological Association. (See excerpt at the NYT.) Does this mean that sociologists are more ethical than us?

Like others I am skeptical that a code of ethics would do much. It’s sort of like, well, having MBA students and investment bankers take a course in ethics and then expecting them to behave, err, ethically. Unlike Lant Pritchett however, I believe that disclosure is important. (See coverage at Slate.) Pritchett believes that any argument should be able to stand on its own weight regardless of the affiliation of the person who is making them. So if Frederic Mishkin believes that the economy of Iceland is sound because of its fundamentals even if Iceland has paid him a consulting fee we should be able to evaluate Mishkin’s arguments fairly without any knowledge of this side payment.
Unfortunately in economics there are very few facts but lots of opinions and arguments and what actually constitutes facts is always debated. There is nothing more factual than a clinical trial - it is all about collecting data and running experiments. Yet would we be swayed if we knew the results of a trial was funded by the company trying to get a drug approval rather than the NIH?

Do we eat corn flakes because it is good for us or because Michael Phelps says its good for us? If eating corn flakes is good for us then we should be able to eat it without any regard for who is telling us to eat it. Likewise, an economist making an argument is just that - an economist making an argument. But if an economist is paid to make an argument or if the economist has ties to a company that may benefit from the arguments that the economist is making then the economist is doing more than making an argument. The economist is making an endorsement.

So we arrive at the question - are sociologists more ethical than economists? While both the American Economic Review and the American Sociological Review requires data sharing only the former seems to be more transparent in making the data accompanying published articles more easily available for download.

Vado a bordo, cazzo

Vado a bordo, cazzo: I could not resist typing this into Google Translate: “I go on board, fucking”. The alternatives (click on the words and Google will pop up some alternatives) may be a little better, e.g. “I go on board, dick”, etc. but none seem to be able to drop the pronoun.

Wednesday, January 18, 2012

Can social media become a weapon of mass hysteria?

With the increase in crime in our neighborhood there was an inevitable increase in chatter on the listserv as to what the community could do. There was talk about setting up a Twitter account so that we could tweet about suspicious people or cars in the neighborhood. Never mind that people were already posting to the listserv on a somewhat of a real time basis. For example, someone walking home posted about police activity at a nearby intersection. There was also an increase in postings about what someone may have heard while talking to someone else, etc.

But … perhaps with a lack of events or incidents, the talk died down and we didn’t become massively hysterical. Which leads me to believe that perhaps social media cannot become weapons of mass hysteria. It would be interesting to see an analytical model of information with behavioral feedback applied to social media and its role in creating social action such as protests and revolutions.

Wednesday, January 11, 2012

How much stuff do we have

In a previous post on the stagnation of median incomes I had suggested that another alternative measure would be “how much more stuff” the median household or median person have had over time. The data source that might contain this data is the Consumer Expenditure Survey. Unfortunately, this data is not publicly available for free download but is available for sale. My knee jerk reaction is: Isn’t there something wrong about selling data that has been collected using public funds? The only other data is from the NBER but this only goes up to 2003.

Assuming we had the data, the measure might be whether the proportion of after tax income that is spent on food, housing, expenditure and health has decreased over time by the median person or household.

Tuesday, January 10, 2012

There goes the neighborhood

A couple of days before Thanksgiving a woman was mugged a block north of us. She was knocked unconscious. The day after Christmas a couple was robbed about half a block down from us and in the past 2 weeks there have been 3 robberies within a couple of blocks from us. Some of these have been armed robberies. These have usually happened after dark and in EST that means after 6 pm. Meanwhile, the listserv is burning up (well, not quite but there has been some discussion) over the benefits of motion detector lights versus having lights that are always on as a neighborly service to those might have to walk in dark areas of the road.

To DST detractors, I say it’s time to scrap EST and go to DST permanently. It is tempting to also make a link between the recession and crime but this article in the fedgazette suggests that it is not so clear cut. And I am not surprised. Highly aggregated numbers tend to hide (if any) underlying patterns there might be. Meanwhile, highly disaggregated reports such as those that might be provided by Crime Reports may also obscure spatial patterns that may exist.

The neighbors are accusing ‘boys from SE’ (presumably DC) and I am sympathetic to the idea that while there may not have been an increase in the overall crime rate (property or otherwise) there has been a definite shift in the geography. Whether these are coming from SE DC or not remains to be seen. Whereas crime may have occurred in and where the criminals lived in the past, the recession has caused them to migrate beyond their usual territory since the pickings have become slimmer in their neighborhoods. But because only a small portion of them would move to commit crimes elsewhere there may not be any noticeable drop in say one section of the metro area and a noticeable increase in another. (I use noticeable as in ‘eyeball comparison’ rather than in a statistical sense.)

Moreover, with the increased use of credit cards and the falling prices of electronics, more robberies have to be committed just for the criminals to maintain their previous standard of living. So in the past whereas one criminal might survive on a robbery or burglary a day, he now has to commit multiple crimes.

I am also skeptical that a short term recession would increase the proportion of people who become criminals but a longer recession (of which this might be considered one) could have an impact on the crime rate. So aggregate measures of crime rate over NBER recession dates would not yield much correlation if the recessions are short. This is suggested in the following paper which finds that the ‘lost decade’ of Japan has increased crime as a result of the effects of the recession on the low skilled. The data used in my mind is less than perfect and is suggestive rather than conclusive.