SK – Page 140 – Pertinent Observations

Mike Denness and WTC Bombers

Professors who are insecure with respect to their ability and competence demand, rather than command respect. They institute complicated procedures which ensure that students need to suck up to them. Professors who know they are good don’t care. For example, the better professors who taught me at IIT never took attendance (everybody would be marked present), and would yet lecture to a full house most of the time. Lesser professors would get finicky about attendance. And other such trivial things. By forcing students to do things in a certain way, by “being strict”, they assumed, that students would respect them. It is a wonder that none of them thought this might be counterproductive.

In our third semester at IIT, we had this course called “Digital logic and VLSI Design Lab”. It was a decent and useful course. You would build digital circuits and test them out. No rocket science to it, but something that was useful in the long run. And because there was no rocket science to it, the faculty (one of the more insecure professors) had instituted a complicated process so that he gets some respect (or attention at least). Actually it wasn’t that complex. Before an experiment, we had to write up about the circuit and how we go about the experiment and get his signature on our write up. The lab assistant had been instructed that we should be issued components only after our report had been countersigned by the professor. Nothing too complicated, but a small step to ensure we suck up to him.

Things were mostly smooth, but one day the professor was late to arrive. Or maybe he was there and we didn’t see him – I don’t remember correctly. I don’t know how it happened but we managed to get the components from the lab assistant without our report having been countersigned by the professor. In a jiffy (after all we were three bright IIT boys) we had finished the experiment. And we called the professor to show him the results.

The experiment didn’t matter to him. He didn’t care one bit about the elegant circuit we had constructed. He only looked at our write up. His signature was missing. And he went wild. I won’t get into the details here but he went absolutely ballistic and threatened to annul our experiment, and possibly even fail the three of us in that course. “Such indiscipline is not to be tolerated”, he said.

“Sir, but this is not fair”, a teammate interjected. It only ensured that the professor went even more ballistic. “You guys must be reading the newspapers”, he thundered. “You see what is happening in South Africa? Is that fair? There is absolutely no fairness in this world, so you won’t get any brownie points by arguing that something is not fair” (the professor was a big cricket fan. The events in South Africa pertained to the one match suspension of Virender Sehwag and a suspended sentence to six other Indians, handed out by match referee Mike Denness).

“I don’t want my students to be this indisciplined”, he went on. “You never know where this will take you, if it is not nipped in the bud. One day you will do your experiment without taking my signature. When that is tolerated, you get encouraged to more indiscipline. And so it grows. And one day you will be bombing the WTC”. None of the three of us was able to react to this (this was in October 2001).

I don’t exactly remember how it ended. If I remember right, we had to dismantle our set up, take the professor’s signature on our write up, re-issue the components and re-do the experiment – but I’m not sure – maybe we were let off. But it was an important lesson for us – if indiscipline is not checked right up front, you could go on to be a terrorist it seems!

Rail length growth in India, or why you should not trust visualizations at face value

My colleague Nitin Pai extracted some data from the latest issue of EPW that shows the growth in total route length of Indian railways in the last 20 years. To get a better understanding of how the rail length has grown, I draw a simple graph. This is what I found:

Data source: Economic and Political Weekly May 18, 2013 vol xlviII no 20

From this graph, it looks like the growth in Indian Railways route length has been pretty impressive. You will also notice that the graph is not monotonically increasing – there are years where the route length is lower than that of the previous years. I would suspect that is due to conversion of metre gauge to broad gauge tracks.

But then if you take a closer look at the graph, you might notice that the y axis doesn’t start at zero. So you might want to see what the growth looks like if you were to start the y axis at zero. Here is what you get:

Now that the axis has been plotted starting from zero, you notice that the growth in rail length by the Indian railways is not all that impressive.

Moral of the story: If you are a user of a visualization, make sure you check things like axes, scales, etc. before jumping to conclusions. You never know what tricks the person who made the visualization might have been up to. If you are making a visualization, however, keep in mind that a lot of your consumers are not going to look at the visualization too carefully, so make use of axes, scales, etc. in a way that embellishes your story.

Betting against your co-investors’ misery

I don’t remember the name of the driver who drove me home that rainy night. I remember asking him, but it was almost three years back and I’ve forgotten. As was my usual practice then, I was sitting in the front seat of the cab, and chatted up the driver as we navigated the heavy Koramangala traffic. Our conversation would get interrupted by him getting calls asking him for money.

He had lost a considerable amount of money in chit funds, he explained. There was a chit fund in which he and his family members had invested. The fund gave them what he called “good returns” in the first cycle. The fund opened again for a second cycle, and once again he and his cousins all invested. This time, however, the fund manager had disappeared, taking with him his investors’ money.

While running a fund well for one cycle and disappearing with investors’ money the next is a classic fraud scheme in undocumented financial services, what intrigued me was that the driver and all of his cousins had invested in the same chit fund. The reason I was surprised is that in chit funds you bet against the misery of your co-investors. Let me explain how this works.

It is like a game played out by N investors over a period of N months (notice that the number of investors and the period of the cycle are equal). Each month, each investor contributes a fixed sum to the pot. And then the pot gets auctioned to the same set of investors, and goes to the investor who is willing to take the biggest “haircut”. Investors who are more desperately in need of money are likely to bid to take a larger haircut than those who need the money less. Once an investor has taken the discounted pot, he loses the right to bid for the pot in subsequent rounds (though he continues to contribute the fixed sum). Each round, the money left over after paying the “winning” investor is distributed as dividend among other investors (with the fund manager taking a portion as management fees).

Notice that this single instrument serves as both savings and loan instrument, with the catch being that it is all in-house. The “haircut” that investors are willing to take while they bid for the pot can be considered equivalent to the interest rate they are willing to pay on that loan (it is a loan – since they need to continue to pay the “premiums” for the full period of the fund). The assumption is that when you assemble a chit fund with a set of investors, each of them has different preferences in terms of when they need the money. Some see it as a savings instrument, and are liable to bid for the pot in later rounds. Others see it as a loan instrument, and bid for it in an early round. All the fund manager is doing is to facilitate this intra-group lending through a formal mechanism.

Also notice that the larger the haircut your co-investors are willing to take, the bigger is the profit that you stand to make. If nobody in the group is desperate for money at any point in time, the bids for the haircuts are going to be rather low. So if you invest in a chit fund and hope to make money off it, you must have reason to believe that there are other investors who are much more desperate for the money than others. So if I invest in a chit fund along with you, I’m betting that at some point in time you’ll need money so badly that you are willing to take a large haircut, so that I make a good dividend. So in effect, I’m betting against your misery!

So if a chit fund investor is actually betting against the misery of other investors in the same fund, there is no point in two people from the same family to invest in the same chit fund – they stand nothing to gain by betting against each other. It was a long drive home that day, and I explained it to the driver, and told him that it doesn’t make sense for him to invest along with his relatives in the same fund. He understood, he said, and added that he didn’t want to invest in the fund that he had lost money in (where the manager decamped), but his relatives accused him of being a “traitor” for not investing along with them, and so he complied! And they all went down together.

Surveying Priorities

Earlier today the Lowy Institute put out the results of a survey it conducted on “India’s views of the world ahead”. While the report contains some excellent insights (including Indians’ perception of various countries), the problem is that it doesn’t establish what people’s priorities are.

For example, there is a question that asks people how important it is that “India has the largest navy in the Indian Ocean”. Some 94% of respondents think it is important, but neither the question nor the answer acknowledges the cost of being the largest navy in the Indian Ocean. Of course, having the largest navy in the Indian Ocean is a great thing to have, but what about the cost?

This is the problem with “uni-directional surveys” – where questions are independent of each other and no relation between factors is established. For example, everyone wants low taxes, high level of government-sponsored welfare, full employment, good wages and a strong military. The reason differences between political parties occur is because it is impossible to have all of it at the same time, and different parties have different positions on the trade-offs.

Table 24 of the Lowy survey illustrates this. The question is about domestic policy goals, and respondents are asked about the importance of each. Is it of any surprise that over 90% of respondents think each and every one of these goals is important?

Extracted from the Lowy Institute report on Indian Views of the World Ahead (http://www.lowyinstitute.org/files/india_poll_2013_0.pdf)

In order to capture trade-offs, I propose a different kind of survey. One where the respondent is told “The government suddenly gets an extra Rs. 100 which it has to spend on either strengthening our military or providing food security. What do you choose?”. The survey I propose will have a series of such “binary” questions, where respondents have to allocate the government budget between various programs. That way, the true preferences of the respondents can be captured.

One last point on the presentation of the above table. The survey uses a “4 point Likert scale” (“not at all important”, “not very important”, “fairly important”,”very important”) to record responses. First off, marketing research theory recommends that such scales have an odd number of choices (3 and 5 are the recommended numbers). Secondly, the report has chosen to group the first two choices under “total not important” and the latter two under “Total important”. As you can see from the table, these “total” columns are presented in boldface, thus drawing attention. Consequently, given the amount of information in each table, no one really looks at the columns not in bold face. In other words, the Likert scale could have had only two points (important – not important)!

Water Subsidy in Bangalore

Pavan Srinath yesterday wrote about the water subsidy in Bangalore, arguing in favour of “crisis pricing” of water in order to tide over the current water shortage. To support that he has produced the chart produced below which shows the total subsidy a household gets as a function of consumption.

The interesting thing to note is that there is “indefinite subsidy”. Ideally you would expect to get subsidy only up to a certain level of consumption. However, the data here shows that irrespective of how much you consume, you still get a significant subsidy for the marginal liter of water that you consume.

Pavan’s own comments on this chart can be found on his post at The Transition State

Revising the Food Security Bill Numbers

Mohit Satyanand replied to my earlier post on Food Security Bill with a couple of comments. He mentioned that only about 40% of the beneficiaries are going to get rice while the other 60% are going to get wheat. He also pointed me to the site of the Food Corporation of India where they give the official “all in” costs of rice and wheat (Rs. 27 and Rs. 19 respectively). I still believe that the wholesale market price is a better measure of the all-in price, but it would be useful to see what the subsidy number works out to given the official government numbers on prices.

Notice that the total subsidy has now come to about 6% of the budget, which is still massive. There are of course other problems with the bill – such as distortion of markets, but those are outside the scope of this blog so I’ll stop here.

Time for bragging

So the Karnataka polls are done and dusted. The Congress will form the next government here and hopefully they won’t mess up. This post, however, is not about that. This is to stake claim on some personal bragging rights.

1. Back in March, after the results of the Urban Local Body polls came out, I had predicted a victory for the Congress in the assembly elections.

2. Then, a couple of weeks back, I used the logic that people like to vote for the winner, and this winner-chasing will result in a self-fulfilling prophecy that will lead to a comfortable Congress victory.

These two predictions were on the “Resident Quant” blog that I run for the Takshashila Institution. It was a classic prediction strategy – put out your predictions in a slightly obscure place, so that you can quickly bury it in case it doesn’t turn out to be right, but showcase it in case you are indeed correct! After that, however, things went slightly wrong (or right?). Looking at my election coverage Mint asked me to start writing for them.

As it happened I didn’t venture to make further predictions till the elections, apart from building a DIY model where people could input swings in favour of or against parties, and get a seat projection. Watching the exit polls on Sunday, however, compelled me to plug in the exit poll numbers into my DIY model, and come up with my own prediction. I quickly wrote up a short piece.

3. As it happened, Mint decided to publish my predictions on its front page, and now I had nowhere to hide. I had taken a more extreme position compared to most other pollsters. While they had taken care to include some numbers that didn’t mean an absolute majority in the range the predicted for the Congress (so as to shield themselves in that eventuality), I found my model compelling enough to predict an outright victory for the Congress. “A comfortable majority of at least 125 seats”, I wrote.

I had a fairly stressful day today, as the counting took place. Initial times were good, as the early leads went according to my predictions. Even when the BJP had more leads than the Congress, I knew those were in seats that I had anyway tipped them to win, so I felt smug. Things started going bad, however, when the wins of the independents started coming out. The model I had used was unable to take care of them, so I had completely left them out of my analysis. And now I was staring at the possibility that the Congress may not even hit the magic figure of 113 (for an absolute majority), let alone reach my prediction of 125. I prepared myself to eat the humble pie.

Things started turning then, however. It turned out that counting had begun late in the hyderabad karnataka seats – a region that the Congress virtually swept. As I left my seat to get myself some lunch, the Congress number tipped past 113. And soon it was at 119. And then five minutes again back at 113. And so it continued to see-saw for a while, as I sat at the edge of my office chair which I had transplanted to in front of my television.

And then it ticked up again, and stayed at 119 for a while. And soon it was ticking past 120. All results have now been declared, with the Congress clocking up 121 seats. It falls short of the majority I had predicted, but it is a comfortable majority nevertheless. I know I got the BJP number horribly wrong, but so did most other pollsters, for nobody expected them to get only 20% of the popular vote. I also admit to have missed the surge in Independents and “Others”.

Nevertheless, I think I’ve consistently got the results of the elections broadly right, and so I can stake claim to some bragging rights. Do you think I’m being unreasonable?

Food Security: Making sense of the numbers

I’m not convinced by the official figures for the subsidy that is required for the food security bill. According to my calculations (shown below), it is likely to be a whopping 11% of the budget, and this is excluding administrative cost.

I bought rice this morning from the kirana store close to my house. I paid Rs. 48 per kg, and it was not the most premium quality. Assuming that the food security act will provide rice that is of slightly inferior quality, and taking out the retailer’s margin, I think assuming Rs. 40 per kg is a fair estimate for market price of rice.

I’m pasting a screenshot of my spreadsheet here:

As you can see, the proposed bill intends a subsidy of 11% of India’s budget this year (to put that in context, the Fiscal Deficit is 5%). Also note that the calculation above doesn’t take into account the administrative cost of implementing this scheme.

Errata: The sixth line in the spreadsheet should read “subsidy per person per year'”. The numbers, though, remain the same

The Upside of ADHD

Attention Deficit Hyperactivity Disorder (ADHD), a condition that I have, isn’t all bad. In fact, it was a recognition of the qualities of this “disorder” that led me to stop my medication for it. I figured that I prefer the with-ADHD me to the without-ADHD me. I found the latter too boring, not creative enough, and unable to connect seemingly unrelated things – something that I’ve always taken a lot of pride in.

Yet another positive of ADHD, I realized yesterday, is that it allows you to lead a “markovian” life. It allows you to easily get rid of historical baggage while taking decisions, and makes you look forward by taking a decision based on the present. Yes, it can sometimes be a bit troublesome, as it prevents you from following long-drawn-out plans, but mostly it’s a good thing.

It makes you disregard that you’d taken a decision for some reason in the past. It makes it easy for you to disconnect from your earlier decisions, and look forward. It doesn’t allow you to be swayed by emotions – on account of some decision you’d taken in the past, and instead makes you rely on rational reasons.

There’s this documentary called “ADD and loving it”. Maybe I should watch that. And maybe I’ll have something to add to that.

Banking activity and economic activity

Out on Capitalmind, Deepak Shenoy has an excellent post on the penetration of banking services in India, where he points out that 30% of all bank deposits in India are in Mumbai and Delhi. I encourage you to read that post in full.

Having read that, I was interested to see the per capita figures and compare them across states. On a whim, I decided to compare that to per capita state GDP and this is what I got:

Data source: RBI website Note: Maharashtra, Delhi and Goa have been left out because they are outliers. Some other states (Chandigarh, Gujarat and Mizoram) have been left out since their latest GSDP figures are not available — Data source: RBI website
Note: Maharashtra, Delhi and Goa have been left out because they are outliers. Some other states (Chandigarh, Gujarat and Mizoram) have been left out since their latest GSDP figures are not available

While the direction of causality cannot be clearly established, this clearly shows that banking penetration is highly correlated with economic activity.