Playing with Analytic Rigor & Terminology: My Unfolding Vogage 068

On an invitation from the Entrepreneur Development Cell of the Indian Institute of Kharagpur, I along with my wife, travel led about 3 hours in a car to reach the IIT campus a few years ago. The organizing students extended a very warm reception and looked after us with care. I was scheduled to address them on the subject of 'Rise of the Knowledge Entrepreneurs' on the following day after lunch. The students took us around the sprawling old campus. We also visited a students' hostel and also the Management School building. Then, it struck me and I requested the students to find out if there was a professor of economics named Partha Basu in the economics department of the Institute and in case he is in station whether I could meet him after my address was over. Partho came to meet us at the IIT Guest House where we were put up, took us to his residence and we had nice time together. He was my colleague at United Bank of India and we were meeting after a gap of 30 years. Parto was two years junior to me in the University and joined the bank probably three years after I did. He was a serious researcher and had lot of problems satisfying himself with the adequacy and representativeness of stratified sampling for a survey he was conducting for an office note. He would accept and write only on the conclusions that come out after satisfying all statistical tests: he could discuss and even have a conviction on a particular hypotheses being true, but would not justify the hypothesis being true unless it has passed the r. rigorous statistical tests of inference. The problem of office communication is that the readers and users of these communication cannot tolerate pleading ignorance. One cannot just say that I did not get support for any of the alternative hypotheses I had tried to test with statistical rig our. One needs to make a personal judgment based on whatever statistical tests one has conducted and whatever theoretical arguments one could marshal for and against all alternative hypothesis, and say this is the most likely Truth and the cost of not accepting this Truth in case if it is really True while formulating policy / making a decision is such and such. Office communications cannot say that one is unable to arrive at a conclusion. The art of drafting an office communication lies in taking the reader through the notes in such a logical sequence that the reader gets increasingly convinced at every stage without any question occurring in the reader's mind so that by the time the person finishes reading the note , the person starts believing that he/ she himself has discovered the Truth contained in the Note and takes responsibility for agreeing with the office Note. Unfortunately, very few researchers would combine their expertise in statistical inference testing and expertise in building a case for a particular hypothesis being true on the basis of knowledge and analysis other than inconclusive statistical tests of inference. Businesses decisions and policies cannot wait for conclusive Truth to arrive,: they are continuously made and revised based on new knowledge and past mistakes identified. Academically, this may look like playing with scientific rigor. I learned that the hard way in the United Bank and later: but serious researchers would interpret that as compromising with scientific rigor. Partho, a very nice gentleman,strongly against male domination over women in families, and with lot of flair for argumentation, would soon leave United Bank and join the Indian Institute of Technology as a economic teacher and would have enjoyed his life thoroughly thereafter.

There was however more things to learn about playing with scientific rigor and terminology. My boss and his boss drew my attention to a two line paragraph from the minutes of a meeting at the Ministry: I do not exactly recall those few sentences but they had recorded that the Secretary or Additional Secretary had advised the banks to formulate models of business plans covering matrix of parameters for rural branches so that these branches turn viable in a shortest possible time. The bosses wanted to know what did the word 'matrix' signify in this context and whether I can build such models. I told them although these sentences conveyed very little extra meaning than they would have if the words 'matrix' and 'parameters' were not used. But I could draw up such a model where I could safely use the terms 'matrix' and 'parameters' and yet these terms would really be insignificant in the essence of building of the model. I also, told them probably the Secretary / Additional Secretary had picked up these terms while visiting the World Bank on an assignment or he must have faintly recalled some terms he had learned while studying mathematics at the undergraduate level. But I could use the terms matrix in the general sense of an array of numbers and parameters as various elements of costs and income rates that along with business activity variables determine bank branches profitability. The bosses relied on my knowledge of undergraduate linear algebra and theory of equations and crossed check if my definitions of matrix and parameter made sense by looking up at the dictionaries. Then, they told me to go ahead and quickly prepare a Note as the Chairman of the Bank had expressed the desire the our Bank should be first to respond on having taken action on this part of the minutes of meeting at the ministry attended by all nationalized bank chairmen and representative of the National Institute of Bank Management. I started my work and also sought their permission to visit a few rural branches so that I could make my model empirically realistic.

I chose the Kolaghat Branch for a five-day visit. It was then a rural / semi-urban location with rural command area. I used to take a daily trip first by bus and by a local train from Howrah station to reach Kolaghat in about 75 minutes. The daily trips cost me in terms of getting an attack of 'conjunctivitis' eye sore that would also pass on my fiance resulting in one of her eyes appearing smaller in size than the other for a weeks before we got married. However, all the five days of my association with the Kolghat branch meant a two hour visit to some parts of the branch's business hinterland for finding prospects of business growth over the next few years, examining the branches past record of performance in these areas and discussing a complete five year business plan for the branch covering deposits of different categories, advances under different categories, staff productivity and manpower deployment, variable and fixed costs and profits. Back to the office, I would access information on all these variables for a sample of relatively new rural branches to mind out the mean and variation in the business activity level and composition, manpower deployment and cost elements for the first three/ four of their operation. Once this information was analyzed, it was easy to develop .internally consistent, feasible and yet challenging time paths of the growth in different business activity levels and the costs, the net interest margins and the profits. The idea was to generate three four alternative time paths of these variables for the first 3 to 5 years of their operations that would be consistent with their reaching the break-even point within that time frame. Thus, I would now have only to present these time profiles, of variable with underlying parameters relating to staff deployment, staff productivity and interest rate margins and etc in the form of an array of numbers that I could call as Matrix 1, Matrix 2 and Matrix 3 as business development plan models to guide the new branch managers to pursue the objectives of meeting priority sector targets as well as reaching break-even point at the earliest. I wrote out the Note and after the bosses were satisfied they sent copies of the Note to the Ministry bureaucrats and to the representative of the National Institute of Bank Management. The latter responded quickly appreciating the work but pointing out that the concept of matrix had not been really applied: he was right, he knew his undergraduate linear algebra well. I was given the task of replying to him, which I did in a language that would close the issue and yet bosses would be satisfied. The bureaucrats would not respond. And, the funny episode of playing with scientific terminology would end with my bosses satisfied.