The President is clearly a fit man with great reflexes. During the election campaign he sank a three pointer on demand for the camera which earned him my everlasting admiration. This time he swatted a fly that was bothering him during an interview in the White House. Nailing a fly is never easy, however, I am somewhat skeptical about the bona fides of the White House fly. Was it a house fly? If so, is it possible that the North American house fly is an entirely different species from the flies that I grew up with in India? They do look somewhat fat and happy over here, compared to the lean, mean third world variety. I don’t believe - and I say this from considerable experience - that a human being can swat one of those Indian flies with their hands. With a fly swatter, maybe, but not your bare hands. I mean no disrespect to the Prez, but that fly was not the real thing.

The science behind fly swatting was revealed to me in an article in Cosmos, a science magazine.

Long before the fly leaps, its tiny ‘brain’ calculates the location of the impending threat, comes up with an escape plan, and places its legs in an optimal position to hop out of the way in the opposite direction. All of this action takes place within about 100 milliseconds after the fly first spots the swatter.

In other words, the fly flies away from the threat and is quite quick about it. Now, flying away from a threat is not exactly great insight (although watching it on high-speed video must have been fun). For example, if I encounter a family of dangerous Canadian Geese on my running trail, I come up with an escape plan and place my legs in an optimal position to hop out of the way in the opposite direction, exactly the way Cosmos describes it. So flies aren’t unique in this respect. I guess what makes flies special is the speed with which they make the decision and act upon it.

The physics of the fly swatter are not explained in the article, but if I were to take a stab at it, it must be about two things:

1. The whiplash effect of the swatter creates very high speeds at the business end of the swatter.
2. The holes in the swatter don’t create the envelope of rushing air that a hand or a flicked tail would create. This points to the possibility that it is not just the sight of the impending threat, but also the envelope of rushing air that tips the fly off.

We had lots of flies to contend with in the small town in India that I grew up in. With no TV (we didn’t get a strong enough signal until I left for college), no video games and very hot summers, it became necessary to kill the ennui with something, anything, more interesting than breathing. During one of those hot summers I discovered that while I could not swat flies with a rolled up newspaper or my bare hands there was a better way.

The technique - Scoop and Stun - which was even superior to a fly swatter. The way it works is that you anticipate the fly flying away from your hand and catch it in mid air. You then throw the fly down hard on the floor. The fly dies from the impact without creating a mess and you can just sweep it away. Another advantage of this technique is that you can target flies sitting on stuff that you won’t be able to swat because you don’t want to break it.

While we celebrate the lightning reflexes of the American president, Indians can rest assured that their politicians are unlikely to follow in his footsteps. Swatting flies in Hindi - “makkhiyan marna” - is a colloquial phrase that means sitting around doing nothing. Being shown swatting flies on TV can make a politician look bad. It’s not as bad as having shoes thrown at you in press conferences, but is still avoidable.

Related posts:

1. Bizarre Bullet

No related posts.

The data plotted in the chart is the margin of victory for the home team at the end of regulation (negative if they lose). There is a very pronounced spike at zero, implying that the % of games that are tied and go to OT is much higher than if it was closer to a normal distribution around a mean of zero.

There is a great video if you click through to the post, which shows how the distribution of the lead of the home team changes in the last 40 seconds of the game. Things look pretty normal till about 20 seconds are left. Something happens in the last 20 seconds that makes things converge to a tie.

The comments from readers try to explain this. The explanations that ring true to me are:

1. Shot selection - If the trailing team is within 3, it will attempt a 3 point shot. If it is within 2, it will typically go for a higher percentage shot for 2 points.
2. Fouling to stop the clock - The trailing team will keep fouling to stop the clock in the last 20s which will typically expand the lead. If they luck out and pull even, they stop fouling. If they pull ahead, the other team starts fouling.

Said differently, if the two teams are separated by 3 or fewer points in the final 20s and the trailing team has the ball, they will try to run out the clock and make a final shot attempt to tie the game (or win it if they are trailing by 1). If the leading team has the ball, the other team will foul them in the hope that they will miss a free shot and they get possession of the ball.

Ely offers his explanation here. Another thread of discussion on this is on Yglesias.

No related posts.

The reason Gawande’s seminal piece has caught the attention of the blogosphere and the administration is because it uses data to point the finger where so far in the healthcare debate, it hasn’t - at doctors themselves. It’s easy to blame the insurance companies or the pharmaceutical companies. They are big, powerful corporations and it is natural to assume that must be twisting the system to their advantage thus raising costs for everyone. Doctors, whom we trust with our care, are much more difficult to confront. But confront them we must. It appears that the root of the problem in American healthcare is over-care.

Doctors are paid by insurance companies (or patients) by the amount of work they do (number of visits) not by the results. Insurance companies pay a standard rate per visit. They make no distinction between the rates they pay to a doctor that provides great care and one that provides poor care. The doctor therefore has all the incentive to increase the number visits and none to increase the quality of care (above a certain minimum, obviously, otherwise they won’t be left with any patients).

There are other perverse incentives in the system. The doctor might decide that they can do some minor procedures in the doctor’s office itself thus generating more profit for themselves. Which is all nice entrepreneurial stuff, but it leaves a huge conflict of interest on the table. The patient probably can see it when they are sent for a lab test that they don’t need, but when all they have to pay is the co-pay, they don’t much care.

The conflicts of interest can be deeper and uglier as Gawande reports - kickbacks from hospitals, favours from pharma companies. The bottomline is that much as we would like to put doctors as care givers on a pedestal, they are economic actors and respond to the incentives they are given. And those incentives lead to expensive, over care which leads to higher health insurance costs on businesses and workers, a large number of uninsured and a rapidly rising state and federal health budget. Unseen by us, this great transfer of wealth from the ordinary tax payer to health providers, has crept up on us.

Another industry that went through its own trials because of conflicts of interest was the broker research industry of the late 90s. CEOs of public companies, investment bankers and research analysts had this nudge-nudge-wink-wink thing going on - analysts would give their top ratings to every company that was a client of their banking colleagues. The bankers would then kick across a portion of their fat fees to the analysts. Everyone was happy but the research that investors were getting was not worth the paper. It was like Lake Woebegon - all the companies were above average.

Then the bubble burst and it was left to Eliot Spitzer to clean up the mess. He prosecuted a few analysts and extracted the Global Settlement from the big broker dealers which regulated the conflicts of interest. There were minor requirements like better disclosure of existing client relationships on research reports but also some big ones which threw up these Chinese walls between Research and Banking. No sharing of fees was the big one. Some of these things have worked, some haven’t, but Research has never been the same since.

Now Research, even in those days was a tiny industry (less than 5B) compared to the US Healthcare industry (over $2 trillion). To tackle conflicts of interest in the Provider industry will take enormous willpower, money and time. The lobbying might of the American Medical Association, the Pharmaceutical lobby and the Insurance industry may just be too much even for a popular President.

For instance, Gawande cites Mayo Clinic as a provider that gives some of the best care in the country and a very low cost. The doctors at Mayo Clinic are all on a salary. They basically have no profit motive and the only thing they care about is patient care.

Can all American doctors move to a salary only remuneration model? There is absolutely no chance of that happening. The only way that could happen if everyone worked for an NIH like setup that the UK has. Everyone agrees that that will not come to pass in the US.

But the option of doing nothing doesn’t exist anymore. The problem with a democracy is that things have to get very bad before corrective action can be taken. A significant percentage of voters need to feel the pain before the right thing can be done. Has that time come yet? I don’t know. I just don’t know.

No related posts.

Kavya Shivashankar won the 2009 Scripps Spelling Bee Championship, as has been widely reported. I watched the last few rounds of the contest. The Bee doesn’t sound like an exciting sport to watch, but once you get into it, it can be quite riveting.

The only word that I knew how to spell in those last few rounds was ‘menhir’. Amazingly, it tripped up Aishwarya Pastapur, who tied for second place. Why did I know how to spell menhir? Because of Asterix - that amazing comic book series that is unfortunately not widely available in the US.

Three years ago, I blogged about the unusual success that Indian kids were having in Spelling Bee. The post was titled Indian-Americans and the Spelling Bee Conundrum. Only 1.8% of the college educated US population is Indian-American. But as many as 20% of the top ten finalists from 2001 to 2005 were Indian-American. Obviously, this is way outside the bounds of random variation.

Well, that percentage this year was more than 50%. 6 out the 11 finalists (Scripps list) were Indian-American. There is an interesting interview on NPR of Balu Natarajan, the first Indian-American winner in 1985. He credits this success to the North-South Foundation a non-profit that promotes education in India. NSF organizes a Spelling Bee every year in the US which effectively is the breeding ground for the National Spelling Bee winners. Kavya won the junior NSF Bee in 2004. Incidentally, Balu Natarajan, in 1985 won on ‘milieu’, which would be a softball for today’s contestants.

It is ironic that Indians, whose mother tongues all spell words the way they are pronounced, should excel at Spelling Bee which exists because of the illogical, wayward spellings in the English language.

Related posts:

1. Indian-Americans and the Spelling Bee conundrum

Google Voice is a pretty unique service. It is like having your own personal PBX system for free. You get a phone number when you sign up for the service. Calls to this number can be routed to a different number (home, mobile, work etc.) or sent straight to voice mail based upon time of day, caller ID or rules that you set up.

Management of call routing, greetings, voice mail administration is where Google Voice shines. Instead of doing a full review, which I am sure you will find in other places, (like this one) let me just mention the couple of things it does very well for me.

I do a lot of calling out using Google Voice. If I’m in front of my computer, it is super easy for me to search for a name and hit Call on the Google Voice website. Google Voice calls my number first and then it calls the other party and establishes the connection. It might sound a little convoluted compared to just calling from your cellphone directly, but it isn’t - for two reason. One, I find searching for a number on Google Voice easier with the full keyboard of my MacBook (my cellphone is a Blackberry Pearl - yes, I could get a full keyboard Bold, but I like my cellphones small). Two, from GV, I will generally initiate the call from my land line giving me a better connection and the use of my landline handset.

Google Voice address book just uses my Gmail address book so there is no syncing involved. This integration was one of the features added when Google relaunched the service. Another new feature was international calling. US domestic calls are all free as is the rest of the service. International calls are cheap. India is 7 cents a minute. Czech Republic, where the wife is currently holidaying, is 3 cents a minute. The rates are all transparent. At this time, if you asked me how much I would pay for a call from my home land line to the Czech Republic, I wouldn’t know. I know the AT&T India rates are awful, because of which we have the Reliance Calling Card. But after Google Voice, I’ve stopped using Reliance too.

Another neat feature is voice mail transcription. It’s a little spotty though and more often than not I have to dial in to hear the voice mail and confirm what I read. I think the problem is that it transcribes American accents well but it can’t handle an Indian accent or Indian names. Voice to text is a tough ask anyway so I won’t take away any points on that.

Surprisingly, Google Voice does not have a mobile app. So I can’t use it from my cellphone when I am not near my MacBook. Well, I can, but I would have to call my GV number and then key in the number I want to call. Nobody’s going to want to do that. Which is quite a problem. I am going to great lengths to get folks to change my phone number in their address books from my cellphone to my GV number, but if they keep getting calls from me from the old number that creates a problem. Plus without the mobile app, it’s easier to just use Reliance for international calls when you are out and about than GV. The Reliance dialer for Blackberry just takes every call made to India on your Blackberry and routes it through their own service. (Note: Reliance mobile dialer is available only to business customers for some reason).

Overall, I find Google Voice to be a very useful service. And I don’t even use half the features. I hope they keep improving it. Meanwhile, my GV number is the only number I give out from now on. Umm…except when I’m in India when I need to have an Indian mobile. Unless…GV could offer the same service in India and allow me to map both numbers to the same account. Now that would be something.

Related posts:

1. Tech Blowback
2. Rants From Delhi Airport

In Mexico I encountered a familiar problem with bottled water that I face in India all the time - when you try to open a new bottle, you invariably spill some water. [Update: My experience in India is with Bisleri bottles primarily. A reader points out in the comments that there are other brands that don't have this problem.]

There are two reasons for why this happens. One, the plastic wall of the bottle is thin. To unscrew the top, you need to grip the bottle hard, which squishes the bottle. When the top unscrews, the bottle is still a squished and the water gushes over. The second reason is that water in the bottle is filled right up to the top which gives no leeway at all when the bottle is squished.

In India and other emerging countries, keeping the costs is not so much about making a higher margin. It is about keeping the prices down so that you can capture a larger market. The food grade plastic used in the bottles is probably the largest variable cost for the bottler. It is no wonder he tries to minimize the amount of plastic per bottle.

That explains the thin plastic walls. And my guess is that it also explains why the water is filled all the way up to the top. The volume of water must be what it says on the bottle - 500 ml. or 1l. By filling the water all the way to the top the bottler can make the bottle just a wee bit smaller and save on the plastic.

The irony is that on account of these cost cutting measures, assuming that everyone spills a little water when opening the bottle for the first time, the consumer net, net, gets less water per rupee and the nuisance of cleaning up the spillage. On the other hand, it spares the earth - the plastics are petroleum products and aren’t biodegradable. Things are never black and white.

Related posts:

1. Water, No Ice
2. Water, No Ice Featured in the Mercury News

Have you ever wondered about visitor registers? They have become less common in the US, where in places like New York building security is the real stuff. In New York you won’t find god-knows-what-purpose-it-serves vistor registers. On my India trip, there were visitor registers everywhere. For some meetings I had to fill out two entries in quick succession, one in the building foyer and one before entering the office of the company I was coming to meet.

I have never understood what is achieved by visitor registers. (If anyone knows any better please leave a comment.) To me it seems like a perfect waste of time. I walk in to a building, mumble something to the guard at the reception and he asks me to fill out the register - my name, company, my host’s name and company, my cell phone, mother’s maiden name…- how this improves security is beyond me. My photo id is not checked. No call is made upstairs to ascertain if I indeed have a meeting as I claim I do. So what is it for? What do they need my cellphone for? To call me if the building is on fire?

And what of the registers themselves? Where do the registers go when they get filled up? Is there some visitor register heaven in a warehouse on the outskirts of every metro? Or maybe telemarketers buy them up to make cold calling lists. Telemarketing calls on your cell phone - that’s a rant I’ll leave for another day.

I spent most of my trip to India in Delhi this time. Of the eight days I spent in Delhi, we didn’t have broadband for six days. Unbelievable but true. Apparently, MTNL changed their back-end infrastructure and the old modems didn’t work anymore. Frantic calls to MTNL support yielded nothing. They said something like - these things happen from time to time; it should be restored by tomorrow. Our neighbours were in the same boat - no broadband, and no clue why.

The next day, we learnt from the internet cafe downstairs that everyone had to change their modems. Apparently, MTNL had put an ad in the papers a while back which informed their hapless subscribers about this matter. And that was that. The subscriber had to make an appointment for the modem to be replaced. If you didn’t read the ad, tough luck. My father had to call the area manager at MTNL and throw his weight around a bit (he’s retired but he can still carry it off) and we got our new modem yesterday. Today I had all my meals brought to me in front of the computer.

I am now in the airport lounge with a huge backlog in my feed reader. So I fire up the old Macbook in the lounge and check if there are any comments on the blog. There are quite a few, but of them, fifteen comments are from a guy calling himself Rohit Khanna. Many identical comments were posted multiple times. None of the comments had much to do with the post. All the comments had a link in them to a website - www.toostep.com. Here is RohitK’s comment stream on IntenseDebate. All the comments have links back to toostep.com. This doesn’t include his commenting on blogs that don’t have IntenseDebate.

What’s going on here? This is not your garden variety comment spam. Akismet is pretty good about catching comment spam left by bots. This is human generated comment spam. The comments make some kind of twisted sense and they do relate to the post’s content. But they don’t pass the smell test. The comments are posted with the sole objective of driving traffic to and raising the google rank of toostep.com. I don’t mind genuine comments that link back to something related - say the commentor’s own post on a similar theme. But I certainly don’t want my blog to become a repository for comment crap. If you are a blogger, you too shouldn’t stand for this. Ruthlessly delete the comments and expose the company that is sponsoring it. toostep.com - shame on you.

There - got it off my chest. Venting always helps. But venting on your own blog is way better.

Related posts:

1. Annoyed Reader
2. Salon.com coverage
3. Notes from TiEcon Delhi

The amount of money spent is, of course, a key determinant of electoral victory. We will cover that in a later article on Corruption and Campaign Finance. In this article we will discuss the state of the art in actually getting the bribes into the hands of the voters.

The most important method of delivering voter bribes is cash. Its advantage lies in the fact that it is uncomplicated. Voters understand exactly how much they are getting and can readily compare it with the going rate of a vote. Most will give immediate feedback to the politician on whether it is adequate or not, and how it compares with what the other candidate is giving out. This allows the politician to do another round of bribes, if the need be.

Operationally also it is simple. Party funds are typically warehoused in paper currency itself. In the event that funds have to be retrieved from offshore accounts, havala transactions can readily convert them into rupees. Gunny sacks are then used to transport them to the point of distribution.

There are some disadvantages with using cash as well. The big problem, in one word, is corruption. Leakages - into the pockets of party workers - can be as high as 30 to 40%. Politicians have tried their best to get party workers to run a clean bribe distribution system, but to no avail.

To overcome this problem, many politicians distribute gifts instead of cash. Saris, liquor and when desperate, mixie-grinders are commonly used bribes. While this increases the operational complexity in procurement and managing the supply chain, leakage is considerably reduced. When very large sums of money need to be distributed - for a statewide bribery program or where the bribe per voter needs to be very high, typically for a completely hopeless candidate - using gifts instead of cash is the method of choice.

Several innovations in the business of buying votes have been made in recent times. One such innovation is to use utility bills to pay the bribes. When funds available are limited, politicians prefer to target neighbourhoods where they are weak instead of spreading the money thin across everyone. But the practical problem with this is that when bribes are being distributed, news spreads quickly over cellphones and then in the resulting crowds there is no way to ascertain where the voters come from. Also, many unscrupulous voters will sometimes take advantage of the crowds and present themselves to be bribed multiple times.

Using utility bills is a brilliant solution because they all carry a residential address. And by requiring voters to present original bills for the last month only, multiple bribe collectors are almost eliminated. The bills need to be retained by the bribe distributors for audit purposes, which makes this method unpopular amongst party workers. Politicians’ search for honest party workers to run this process has not yielded much success. Consequently, in spite of its advantages, this method has not spread widely.

Another innovation that has been used very successfully, is to promise a very large bribe to be paid if elected. This works very well for the political party making the promise because a) it is conditional to being elected and b) the bribe is paid out of public funds. This allows them to ratchet up the size of the bribe greatly. In the last state elections in Tamil Nadu, for example, the DMK promised every family a color TV and then made good on their promise after they were elected.

The only downside to this is that voters in the past have tended to be cynical about election promises. However, surveys indicate that a high percentage of voters are able to distinguish an election promise of good governance, which is unlikely to be kept, from the promise of a future, conditional bribe of a tangible product. This is understandable since good governance is ‘in the eye of the beholder’ while a color TV either is or isn’t.

There is a great tradition of voter bribery in India. There are legends about it dating back to the 1950s. Once such legend concerning voter bribery apparently occurred in the 1950s in Orissa. Biju Patnaik, who was the front-runner to be Chief Minister, as expected by voters, had distributed cash to ensure his party’s victory. His opponent Hare Krishna Mahtab, who did not have the resources that Patnaik had, needed to do something to counter. According to legend, Mahtab’s party workers spread the rumour that Biju Patnaik had distributed fake currency notes. Then they generously offered to exchange the fake notes with real notes! Needless to say the Mahtab camp greatly underestimated the sophistication of Oriya voters, and lost the election.

Today, bribing voters has become so common that all serious candidates must spend a good chunk of their election funds on it. Since other issues matter little (fomenting communal hatred being an important exception), candidates must compete with each other on the quantum of bribes per voter. This is good for the economy as it redistributes wealth from industrial houses and foreign corporations to the aam aadmi. Also, almost all the money distributed as voter bribes is spent immediately which gives a boost to local businesses like country liquor distilleries. In total, the voter bribes industry now accounts for an estimated 2 percent of GDP and about 50 basis points of annual GDP growth. At the G-20 summit later this week, as world leaders discuss what each country is doing to restart the global economy, Prime Minister Manmohan Singh will argue that voter bribes in the runup to the general election should be treated as India’s ’stimulus package’.

In order to encourage more spending on voter bribery, the election commission recently took the important step of releasing two promotional videos on prime time TV. In these videos, Mulayam Singh Yadav and Jaswant Singh, senior leaders of two important political parties were shown distributing cash to voters. It is hoped that this will accelerate election spending by candidates in the last few weeks before the elections. In the current weak economic environment, the election commission must be commended for this bold, creative step.

I would now like to invite comments from readers on how we can take voter bribery to new heights.

[Disclaimer: From the first couple of comments it appears that my attempt at satire did not quite work. So, for the record, the facts and figures in this post could be grossly exaggerated or just plain wrong. Like Bollywood movies, to enjoy this post, you must suspend your disbelief.]

Photo Credit: Wen-Yan King

No related posts.

The Rational Information Technology Policy

Be totally blind, deaf and dumb on whether to use or not use IT tools.

Details:

* The government has no recommendations on who should use IT, in what manner IT will be used by people, households, and firms.
* The government will not directly fund or subsidize the adoption and use of any IT tools
* The government will neither support nor oppose the use of any IT tools in any legitimate activity. The government will be agnostic towards the adoption and use of all IT tools
* Tools are tools, not ends. Use of tools helps achieve ends. The government is interested in ends, not in means. Depending on the context, the appropriate tools will be selected.

There you have it. I have solved the problem of an IT policy.

My rebuttal argued that for a variety of reasons, an active hand of the government in promoting the use of IT is desirable.

He responded thus to my post

The misunderstanding of the role of government is distressingly common. A blog post on 6 AM Pacific declares my post on the rational IT policy “to be wrong-headed.” Then adds the non-sequitur, “I think it is important for any government that comes to power to nurture and encourage the use of IT in government, business, education and at home.”

I am hard pressed to see where it is that I have advocated that the use of IT should not be nurtured or encouraged. The government has a role in enabling the use of IT where it is appropriate. What I am against is the government mandating of specific tools and technologies. It is not the government’s job to pick winners. It has to get out of the way of people and businesses and let them figure out what is the best use of their resources. The ones who use the tools are best able to judge what is appropriate, not some bureaucrat in some government office. The bureaucrat has the right to choose what is used in that government department, not elsewhere.

Reading his post and the few others that he wrote on the BJP IT Policy Document, I came away with the distinct impression that it was not just that he didn’t want government to pick the winners. In fact, he saw all IT as a tool, or a means to an end, which did not deserve government funding, policy or advocacy. The users of the tools would determine its demand and that was sufficient to decide its fate. I disagree with the notion that tools, especially ones with a lot of leverage like IT has, don’t need a helping hand from the government.

I would even take it further. The government should, in many cases, actively favour one technology or one option over another. I can point to any number of policy measures - mobile phone standards (GSM/CDMA), local/mobile number portability, net neutrality (in the US), promoting open source software in government (in China) - where there are multiple valid paths but the government must choose. Making the right choice (or in some cases making any choice) can lower costs, increase competition and increase penetration.

Not everything can be just left to the market. This variety of libertarianism, if it ever had a hope, has been squelched by the global crisis. Markets don’t self-regulate. People want well regulated markets. They want a government that is a force for good, not a passive-stay-out-of-my-way government. I understand that it isn’t always easy to determine the boundaries for a force-for-good government. And corrupt actors within the government will cause leakages. But that is no reason to not have an active government making policies such as an IT Policy.

Related posts:

1. IT and the Government