by Pranab Bardhan
All of the articles in this series can be found here.
At ISI we were assigned statistical assistants who’d take our large data analysis jobs to the IBM computer at the Planning Commission, but for relatively small jobs they’d do the calculations themselves by furiously rotating the handles of the small Facit mechanical calculator they each had, you could literally hear the noise of ‘data crunching’. This was before electronic desk calculators came to Indian institutions. I remember buying a small Texas Instruments calculator in a short trip abroad and was quite impressed by its capacity; and I told TN that I did not need to learn the operation of Facit machines, which I saw him cranking all the time. (This reminds me of a British economist, Ivor Pearce, who told me that just before the War he used to work for an accounting firm where they had not yet heard of log tables; he said he finished the whole day’s work in just an hour by using the log table and read books in his office the rest of the time). Of course, I am told today our tiny laptops/smartphones contain computing capacity million times larger than the biggest IBM machines in India at that time.
The statistical assistants at ISI were literally called ‘computers’ (I was a bit taken aback when on the first day a man came to see me and said “I am your computer, sir”). One day when I was chatting with this human ‘computer’, he said some years back he had worked with a foreign professor who was rather short-tempered and used to scream at him for the slightest delay or lapse. (It so happened that I knew this professor). I said he should have protested if the professor was unnecessarily rude. He gave me a sneaky smile and said that he and other ‘computers’ had taken their ‘revenge’ on that guy. When I asked how, he said they used to mess up his calculations without the professor knowing about it. I was aghast (and also made a mental note not to trust his data analysis of that period). This was an example of what Jim Scott, the political scientist, has called ‘weapons of the weak’ in his eponymous book; many decades earlier the famous Czech novel ‘Good Soldier Švejk’ had satirical accounts of passive-subversive resistance of military authorities by the soldier.
I have myself faced such subversive behavior on the part of lower officials in Delhi in a somewhat different context. Once I went to the office of a particular Ministry in search of some data which I knew they had, but the officer in charge openly said that he could not show me the data, though he did not give any good justification. I told him that these data were collected with tax payers’ money, and since no national-security issue was involved, he was duty-bound to release the data. He just smiled dismissively.
Back to office, TN told me that in India when I wanted something in a Ministry I should not go to the lower officials; I should instead work my way from the top down (this was called ‘proper channel’ in official parlance). TN gave me the contact of the higher-up officer in the same Ministry, who when I told him what I needed immediately called the lower official and asked him to share the data. Next day I went to the same official who had refused me before. He was now full of oily politeness and said that it was the great fortune of the Ministry that a professor like me was going to make good use of the data. But for the next few months on one excuse or another he made it very difficult for me to lay my hands on the data. After a lot of running around I finally got the data, but I tried to fathom the reason behind his delaying tactics. Was it his resentment that I went to his boss instead of buttering him up? Was it his way of asserting his passive-obstructive power (the weapon of the weak)? Or did he expect some bribe from me? (In general, on bribery in Indian offices, apart from the ethical problem, practical problems abound: how to know whom, when and how much? Sometimes touts are there to help in this matter.)
Occasionally there are even more unpredictable barriers to research with data in India. A few years later, when I was at the Delhi School of Economics (DSE), for big data analysis I used to send my research assistant to the IBM 360 mainframe computer located at a different part of the campus, where after you submit your job on the long queue, you’d get back the results after 3 or 4 days. For small jobs there was an antique IBM 1620 machine in the basement of the DSE building. One time I had a relatively small job and I needed the results quickly. So I asked my assistant to take it to the basement and get it done at the smaller machine. She soon came back and said the computer was ‘down’. After a few hours I sent her again, but it was still ‘down’. I decided to check myself what was going on and went downstairs. Usually professors themselves did not go there, so when I was there the officer in charge rushed to greet me. When I asked what was wrong with the computer that it was down for so long, he said, nothing was wrong with the computer, but they were not to run it when the air-conditioner was not running, and he did not know why that was.
I then went and found the man who was in charge of the air-conditioner; he said there was nothing wrong with the air-conditioner but he’d not run it as the water tank at the back of the air-conditioner was empty. I presumed he was referring to the water-cooled condenser in a large commercial air-conditioner which was connected with a water tank at the back; in a hot climate that could be important. Of course, he did not know why there was no water in the tank. By that time I was desperate. I went to the garden at the back of the building where the tank was supposed to be and indeed found the tank empty. As soon as the gardener, who lived in a small cottage nearby, saw me he ran to come and greet me. When I asked him why there was no water in the tank there, he sheepishly mumbled something in Hindi that I could not catch. By that time a small crowd of peons and others have gathered around me. They told me that in connection with some religious festival the gardener had village relatives visiting him, and given the water shortage in his cottage they had come and used up the tank water to take their bath. I was dumbfounded, contemplating the chain of events. The gardener’s rural family by taking the simple action of bathing, had through a whole chain made a university department’s computer system dysfunctional, and disabled my analysis of data (which incidentally were about the problems of rural households in India). Is this what they call the ‘butterfly effect’ in chaos theory?