Monday, November 24, 2008

The Ethics of Financial Engineering

As I write this, the Dow-Jones Industrial Average is somewhere south of 8100, down 35% or more from its 2007 high and showing few signs of fatigue in its downward trek. General Motors hourly threatens to go bankrupt, credit markets are doing an imitation of the last Ice Age, and newspapers are running old pictures of soup lines during the Great Depression of the 1930s, I guess to get us used to what's coming. For such a young recession, it's already gotten plenty of publicity. But in all the finger-pointing about whose fault it was that we got into this mess, I have not read anyone who has addressed the question of what might be called economic and financial engineering, and the ethics associated with it.

I remember being surprised the first time I heard that a couple of our better electrical engineering graduates got high-paying jobs with a credit-card company, of all places. But I'm surprised no longer when I hear that present or former engineers often get hired by banks, brokerage firms, and other outfits that deal in highly technical and complex financial machinery. The attention to detail and problem-solving skills that engineers learn can be applied fruitfully to finance and securities trading as well as electronics, and the pay can be better, too. I call it "machinery" although in reality it's mostly software and rules devised by lawyers and technical types such as former engineers and physicists. But the complexity is there, and there is a good argument that such complexity played a significant role in the current recession.

From what little I do understand about the situation, when all sorts of home loans (the good, the bad, and the ugly) were bundled together by means of software-mediated deals, they sold like two-dollar Miley Cyrus concert tickets at a middle school full of teenage girls. To make things more complicated still, financial institutions started selling things called "credit default swaps," which were some sort of unregulated insurance against the eventuality of loans turning bad. My point is not to explain these things in all their gory details (which I couldn't even if I had to), but to show that computers and technical people who can keep track of these things, and figure out the rules by which they operated, played essential roles in this situation.

Before electronic computers became generally available, the most complex math a banker had to deal with was figuring out compound interest, and there were tables for that sort of thing. The complexity of a given financial deal was limited by, among other things, the labor it would take to figure it out. If somebody came up with some kind of security that came with a formula that would take three women punching calculators for three days to figure out, nobody would have bought it.

Not so today. If you took all the computers away from today's traders, the whole system would come to an instant halt, not only because computers are the medium of communication (so-called "electronic trading" is involved in virtually all transactions), but because a lot of trades are initiated by automatic triggers that write buy and sell orders based on electronically reported prices.

This is not to say that speculative booms and busts are possible only when you have engineer types and horribly complicated automated trading involved. The classic textbook example of a boom-and-bust phenomenon was the tulip-bulb mania of the early 1600s. Substitute tulip bulbs for bundled home mortgages, and you can see the same psychology at work: rising prices, a spreading perception that investing in tulip bulbs is a great way to make money fast, a few people made richer but only if they cash out early, and then alternate reality sets in: hey, we're only talking about tulip bulbs here! What's the big deal? And the crash follows, wiping out thousands of tulip-bulb plutocrats.

Engineers or technical people are not to blame for the mass psychology of crashes. But as they are often endowed with perhaps an above-average grasp of logic and what used to be called common sense, I would hope that they could serve as a kind of reality check or brake on things when matters really get out of hand. Of course, engineers working for a firm whose whole existence is based on complex derivatives or credit-default swaps or tulip-bulb futures, are not going to have long stable careers in such firms if they start questioning the fundamental assumptions on which the operation is based. On the other hand, it's looking like they won't have long stable careers anyway, now that many of the outfits are going broke.

I have no illusions that many of my readers are working in the financial industry. But if you ever happen to end up either working in it or dealing with it, remember that when a deal gets so complicated and computerized that even the people who are buying and selling it don't really understand it—then maybe it's too complicated. Complexity in the service of necessity is one thing, but complexity simply to confuse the buyer is wrong. And it looks like there were a lot of confused buyers out there who have lost faith in their vendors, to the detriment of the economy as a whole.

Monday, November 17, 2008

Gateways to Engineering: "Our Mr. Sun" 52 Years On

In my perhaps overly generous definition of engineering ethics, I consider the question of why people become engineers a legitimate topic of enquiry within the field. If for some mysterious reason young people all of a sudden lost interest in becoming engineers, we'd have real problems getting engineering projects done, and all the good things that result from engineering wouldn't happen. Also, even though it's been two years since I did a blog on the movie "The Prestige," I'm still getting comments about it. So these two factors lead me to draw your attention to a curious film called "Our Mr. Sun." But first, some context.

The nineteen-fifties were unique in many ways, some good, some not so good. Having helped to knock flat nearly every other industrialized country's manufacturing capabilities during World War II, the United States enjoyed an unprecedented decade of prosperity as we produced a lot of neat new stuff that the rest of the world wanted. One beneficiary of this abundance was the Bell System, which back then legally monopolized the U. S. telecommunications market. With a small portion of their government-regulated profits, Ma Bell devoted itself to what it perceived as good works, the nature of which some historian of technology ought to explore one of these days. One of these good works was a well-funded series of entertaining and (they hoped) educational films, which became known as the Bell Laboratories Science Series.

It sounds like what we call "infotainment" today, but compared to today's thirty-minute ads for weight-loss nostrums, the Bell films are almost the exact opposite. Today's infotainment is produced as cheaply as the advertisers can get by with it; Bell went out and hired top-notch directors such as Frank Capra, and gave them pretty nearly a free hand and generous funding. The whole point of today's infotainment is to sell you something; the first film in the Science Series, an hour-long production called "Our Mr. Sun," has about three minutes devoted to the Bell System's early research in solar cells, and one gets the feeling that if Mr. Capra had decided to cut it for a good artistic reason, the Bell people would have swallowed their pride and gone along with the cut.

When one asks why a telecommunications monopoly would spend their shareholders' money on such an apparently profitless enterprise, the non-historian is reduced to guesswork (the real answer may be buried in the AT&T Archives in New Jersey, but my historical-research travel budget for this blog is busted). One reason might have been that Bell, and a lot of other people besides, were worried that U. S. citizens didn't know enough about science, and needed to know more. Although the film's format combined animation and live action, it was clearly intended for a wider audience than the kiddies, since it reportedly aired on television at 10 PM, which was not a good time if you wanted a lot of children in your audience. On the other hand, over 600 prints were later distributed to schools and civic organizations, so reaching the younger generation must have been at least a part of their intentions.

One way to get a grip on why they did it is to watch the film, which I did last night. (It is now in the public domain, and a URL for a streaming source is given below.) Technically, the story is in the form of an allegory a la Pilgrim's Progress, with characters such as Mr. Research (played by professor of literature Frank Baxter), Mr. Writer (Eddie Albert), and the voices of Mr. Sun (Marvin Miller) and Father Time (allegedly played by Lionel Barrymore, in his last role). These latter two worthies appear via animated segments, which makes it easier to conduct interviews with beings such as stars and personifications of non-material ideas.

Capra knew his way around lively stories, and the film holds up surprisingly well in both the dramatic and the technical senses. Dramatically, it does not induce that cringing sensation that the hyper-corniness of so many didactic films of the 1950s produce in us today. Capra managed to get across three thousand years' worth of history about how the sun was originally regarded as a god, to today's present view of it as a flaming ball of gas, and pinpoints the turning point with the name of a specific Greek philosopher. Then the viewer is treated to such things as the physics of nuclear fusion as explained by a magician, the problem of future sources of energy as symbolized by an "energy bank" measured in horsepower-hours, and speculation that future energy shortages will be solved in the short run by nuclear energy, of course, but eventually we might run out of uranium and then we'd have to develop better solar cells.

It's easy to throw rocks at such clouded crystal balls (to mix a metaphor), but the science that was state-of-the-art then was explained well. We get to see a clip of Hans Bethe, who originated the explanation for the carbon cycle of nuclear fusion in the sun, and get treated to what scientists knew about how chlorophyll (or rather, Chloro Phyll, a diminutive cook in a plant's metaphorical chemical kitchen) turns sunlight and water and carbon dioxide into sugar. In one of the weirder sequences, a cartoon scientist dressed in a chef's outfit starts with beach sand, purifies it in a blender, adds a "dash of arsenic" (!), cooks it in a boron oven, and voila! out pops a pan of solar cells from the oven, like cookies. (You're tempted to say, "Kids, don't try this at home," but the Bell System, in a separate but related public-education program, made available to public schools a do-it-yourself solar-cell kit, complete with a set of little fire bricks to build your boron-diffusion oven with.)

Your correspondent was too young at the time the film was released in 1956 to see it in its initial release on TV, but seven others came out over the next eight years, and chances are that anybody going to school in the 1960s saw at least one of these films. In an essay on the series, David Templeton notes that many young people who saw the films later became scientists and engineers, and some cite the series as at least one reason why they chose their technical fields. So in that sense, at least, it looks like Ma Bell got her money back.

It's hard to imagine anything like this taking place today, for a number of reasons. Telecomm monopolies have passed from the scene, corporate altruism is not popular with shareholders these days (what is?), and there is no chance in perdition that a modern film director could get by with the framing motif that Capra, a committed Catholic, chose to begin and end "Our Mr. Sun" with. The first words you see are "The Heavens declare the glory of God" (the first line of the Old Testament Psalm 19), and nearly the last words you hear are those of St. Francis of Assisi, who viewed Nature not as our mother, to be enslaved to, nor as our mistress or subject, to be exploited and dominated, but as our sister, to be loved, cared for, and regarded as a fellow creature of one's Creator. Of course, this was back when prayer in schools was not only permitted but often required by law. Whether getting rid of that kind of religious intrusion in public education has contributed to our current parlous state in which the future of engineering in the U. S. is at least somewhat in doubt, I will leave as a puzzle for the reader.

Sources: "Our Mr. Sun" can be viewed at the AVGeeks archive at http://www.archive.org/details/our_mr_sun. David Templeton's essay can be found at http://www.metroactive.com/papers/sonoma/09.23.99/bellscience-9938.html.

Monday, November 10, 2008

Watching Teenage Drivers with Webcams

Over two hundred teenagers in southern Maryland are now driving around with a webcam on the rear-view mirror of their cars. Whenever they turn or brake sharply, the resulting g-force triggers the camera to record a 20-second sequence of what went on inside and outside the car before and after the incident. These dynamic snippets go via the web to a company in San Diego that reviews them, attaches little helpful comments about how such dangerous driving incidents can be avoided, and notifies the teenager's parents that the video is now available for viewing.

Although deaths and injuries in automobile accidents have been declining slowly for years, over 40,000 people died in highway-related accidents in 2007. Anything that makes that number go down without severely compromising some other desirable outcome of automobile use is worth considering. And at first glance, the specter of Mom or Pop looking over the teen driver's shoulder, so to speak, seems like a good idea. A similar study done in 2006 showed that drivers who started out triggering the webcam a lot with their jerky, high-risk driving, eventually learned to reduce their triggering rate (and thus drive more safely) by four-fifths. It's too early to tell whether a similar improvement will result from the Maryland experiment. But one thing is already clear: the teens don't like the idea, even though some grudgingly admit that the system has improved their driving.

Do the teens have a point? Is the webcam an intrusion on their privacy? Obviously it is, but then you have to ask whether the chance of saving someone's life is worth a little less privacy. And it's not like the thing was on all the time. Teens do other things with, and in, cars that I'm sure they wouldn't want their parents to see. But when the company that operates the system says it won't forward anything that's "embarrassing to the teen" in their words, that seems to be enough to satisfy most young drivers. Of course, if the company were ever to betray that trust, the entire system might suffer a black eye that it might never recover from.

This system is just one example of how technology is making it possible to monitor more and more aspects of our daily lives, in ways that were unthinkable back in the days when George Orwell wrote 1984. One of the creepier images of that novel was the spy cameras everywhere, monitored by secret police whose presence the citizens were reminded of through the slogan "Big Brother is watching you." A sure-fire argument against that kind of thing ever happening in reality used to be that you'd never be able to man every camera everywhere, because eventually you'd end up hiring one half of the populace to watch the other half. But notice that the in-car webcam uses smart technology—namely, accelerometers—to select only those incidents worthy of study, thereby reducing the work of human editors to manageable proportions.

So as time goes on, it will be more and more practical to acquire webcam data on all sorts of activities, and still be able to handle the massive amounts of raw input intelligently. Is this a fundamental threat to privacy, liberty, and all that? Or is it a tempest in a teapot?

The answer hinges on those who are doing the spying, or monitoring, or whatever you want to call it. In the case we're discussing today, a private company is involved with consenting families, and if the company does anything out of line, they are liable to lose business fast. That's one of the best constraints against misbehavior. Governments do not have such a negative incentive, which is why government-sponsored monitoring of behavior can be more problematic. A case in point is the increasingly obtrusive safety inspections for airline passengers. In certain airports, systems are now in place that use millimeter-wave sensors to see through a person's clothes. The people who inspect these images are not co-located at the inspection point, but still, you wonder if and when this kind of thing will be abused.

It seems the best thing to do in these cases is to ask whether the system is doing any good. In the case of the in-car webcams, it looks like they may well improve driver safety, which is good for everybody. In other situations, such as in-flight security, it's harder to evaluate effectiveness except with tests in which people try to sneak by the inspection stations on purpose. And the news regularly carries reports that inspectors often fail these tests. On the other hand, we haven't had any U. S. planes get bombed or turned into flying missiles since Sept. 9, 2001, so something is working, at any rate.

The other factor to consider is the continuing decline in monitoring technology cost. The current webcam system costs $900 plus a $30 monthly fee, but if it proves popular, these costs could go down to where it would be offered as an option when you buy a new car. If insurance companies like it, you could get a discount on your teenager's insurance rate if you agreed to install the device. And once it's in there, it will work for everybody: the teens, Mom, Pop, and Grandpa. So one day we may all be driving around with spycams in our rear-view mirrors, who knows? Let's just hope the people operating the cameras then are as trustworthy as they are today.

Sources: An early report on the in-car webcam can be found at the Washington Post website for Oct. 24, 2008 at http://www.washingtonpost.com/wp-dyn/content/article/2008/10/23/AR2008102303821.html. The 2007 automotive fatality statistics can be found at the U. S. National Highway Traffic Safety Administration website http://www-nrd.nhtsa.dot.gov/Pubs/811017.PDF.

Monday, November 03, 2008

The Ethics of Externalities

You may have never heard of an externality, but engineers (as well as nearly everybody else) deal with them all the time without realizing it. The term comes from economics, and means an effect of an economic transaction that happens to somebody who was not directly involved in the transaction. That's pretty dry, so let me give a juicy example.

In the early nineteenth century, the chemical called sodium carbonate (washing soda) was obtained by burning a type of seaweed found off the coast of Spain. But when Napoleon ticked off England so much that the British blockaded French ports, that cut off France's supply of soda from Spain. The French government thereupon offered a prize for the best process of making soda without seaweed. A chemist named Le Blanc found that if he heated ordinary table salt with sulfuric acid, he got an intermediate chemical (sodium sulfate) that was easily transformed into washing soda. Le Blanc won the prize, the French were able to wash clothes again, and eventually the Le Blanc process took over as the main commercial way of making soda.

The trouble was that a by-product of the process was hydrochloric acid. At first the manufacturers just let it go up the chimney, but nearby farmers began to complain that it was killing their crops. This effect was an externality to the economy of making, buying and selling soda. Eventually chemical engineers found a way to capture the acid and sell it too, but not all air pollution problems are so easily solved.

One of the major externality issues these days is the problem of carbon dioxide emissions. Every time anyone burns a carbon-bearing fuel (coal, especially, but to some extent gasoline and oil as well), the resulting carbon dioxide goes into the air and plays some role (exactly how much isn't totally clear) in global warming. The prophecies about what consequences global warming will have if we don't do something about it range from the negligible to the apocalyptic. This fuzziness about how much carbon dioxide does what amount of damage is one of the classic problems with externalities. In a straighforward economic transaction between informed parties, the price paid says a lot about the relative value of the commodity. If the price goes up or down, that represents information that buyers, sellers, and even economists can use about the thing being traded. But the person harmed (or occasionally helped) by an externality doesn't spend any money, and therefore the economic equivalent of the externality is much harder to determine in many cases.

What have externalities got to do with engineering ethics? A lot, as it turns out. Many externalities are hidden, often from the traders and sometimes even from the third parties being affected. Returning to environmental externalities, such infamous incidents as the terrible ground pollution in the Love Canal area of Niagara Falls, New York came about because standard practices at the time allowed chemical companies to dump toxic waste into the ground with only minimal precautions, and nobody gave much thought to the possibility that someone in the far future might want to come along and build a school on the former toxic waste dump site.

Once you start looking for externalities, you'll see them everywhere. Anyone who buys a piece of new electronic gear is creating a future externality that arises when the thing is no longer useful: where does it go then? Into a landfill? A landfill next to whose property? Or maybe it goes off to some third-world reprocessing facility—then what? As the saying goes, "you can't just throw things away anymore because there is no 'away' anymore."

Even the cyberworld has externalities. Say some online game clogs up a server so much that other people just trying to get their work done experience slowdowns. That's an externality, and one that's hard to evaluate as well.

Just knowing about an externality doesn't mean that it's easy to deal with. Economists say you can pass laws or tax externalities to right the potential wrongs they represent. But that assumes you can put a value, either economic or moral, on the externality. Obviously, if a third party is injured or inconvenienced by some transaction that he or she has no control over, there is at least the potential for injustice, and in a just world such things wouldn't happen. Then you have to ask whether in the grand scheme of things, this particular injustice due to the externality is worth worrying about or fixing compared to everything else that's going on. This is the problem known as life, and we don't do life guiding here.

All I wanted to do in today's blog was to let you know about a concept that I have found useful in thinking about a wide variety of engineering ethics problems. Perhaps the most important thing about externalities is to recognize them when they occur. Depending on how serious they are, the ethical engineer may or may not want to address the issue, but if you don't see it, you'll never be able to do anything about it.

Sources: Wikipedia's article on "externality" is helpful, although it concentrates mainly on the economics of the concept. The story of Napoleon's washing comes from a modern reproduction copy of Asher & Adams' Pictorial Album of American Industry, 1876, p. 19.