 Hello and welcome everyone. This is the second talk in the third chapter of faculty seminar series that we have been doing in computer science department. This talk is going to be given by Millen Sohani, the speaker need no introduction. Millen faculty member of computer science department, the head of Sitara. And this talk I would like to mention is a CSE Sitara talk. But I can see lot of faces interested from all areas of engineering science here. So, the talk is about the elite university and RV2 selective. So, without much ado let me pass the mic over to Millen. Okay. So, thank you very much for giving me this the stage and this opportunity to talk to all of you. So, my name is Millen Sohani and I am going to talk I am the current head of Sitara and also faculty member in CSE. So, I am going to talk about the elite university and RV2 choosey or 2 selective. So, by the way this talk is joint work is ongoing work with Vinish from school of management and an M. Tech student Monique who is here as well. Monique. Monique is here. So, and I must also thank placement office for a lot of data which they did which they gave me and they shared and a lot of help in other slightly confidential data. So, the motivation for this talk is actually the gaping inequality that we see in our society you know poor development outcomes. But there are you know very old institutions and the government structure is there but we see very poor governance and also there seems to be an allocation problem you know where there is where we want engineers to be there are no engineers and where we want engineers where we do not really need them they seem to be there. So, I think that this the basic question was that how is knowledge being formed in our society and you know how is it being transmitted. So, that was a major research area for me and I have been looking at it for the last 4-5 years. So, this is the third sort of paper in this series that we are investigating. So, also by the way this knowledge and development how knowledge is formed and transmitted in 2011 Joseph Stiglitz also pointed to the same issue that knowledge transmission and development are intimately related. So, here is the talk outline I am going to talk first about the elite Indian university that what are its features and how elite are we really and how is it and what is the role of society in the university. The second is I am going to look at the placement data. So, after all for any educational institution what do the graduates do after they graduate is an important measure and we are going to look at that and we will set up the methodology analysis and come up with some preliminary conclusions. Then for a deeper analysis we will have to look at how an economy actually measures abilities and you know gives wages. So, what are the productive processes and by which wages are determined and then we will apply this sort of theory to the data and come up with some conclusions. So, we will actually compute wage curves for IIT Bombay. And then the final issue is about you know the political economy or you know as they say what is the meaning of meritocracy and are we really satisfying those axioms. So, I will propose some axioms based on Stiglitz again and just check whether we are satisfying those axioms are we really a meritocracy and how do we fail it or do we pass it and so on. And finally you know I will spend some time on how do we come out of these some of these problems that I will point out. So, first of all I am you know it is clear that the society and university form a very important virtuous cycle. You know the university is the repository of all knowledge and practices and in turn and the university receives support from society and in turn the university actually produces agents or employees which would who find place in society you know working as workers in factories as singers you know as civil society agents whatever. So, it is you know so this virtuous cycle is really how the university and society get connected. So, the elite university of course is much more important it is supposed to provide thought leadership you know the arts and you know long term research and is in a way connected to the destiny of this society right. So, it is also a symbolic of what this society values. So, it is in that sense the elite university is quite important. So, in this talk I am going to look at you know there is an input side of how you know how students are selected and what are they trained and you know how they enter the university we are not going to look at that we are going to look at the output side. So, as you know how do you graduate where what career options do they take and what are they you know what are their compulsions and how does it benefit our society. So, this talk is only about output side of the picture. So, by the way stop me there are some quick questions otherwise you know we can you know I will take questions at the end. So, first of all you know let us talk about the Indian elite university. So, it has a long history I mean the university is of course started in 1850s the modern university, but the true elite university started after independence the first is of course the IITs which started in the 1950s ISC, ISRIS, IIMS, TIFR you know JNU there are many such now many universities and they actually one key feature is that they cover most of the areas of you know knowledge. So, for example, economics, science, technology, social science, mathematics and so on. So, there are of course, new ones coming up like the new ISRS and some new IITs right and key features of these elite institutions are that they are centrally funded typically fully they are autonomous they have a research orientation internationally trained faculty and fairly transparent and highly selective admissions of students and a focus on excellence and global standing. So, these are some of the key features of our elite you know I mean these are well understood and I think that I think that is pretty clear right. So, now let us look at how elite are the IITs. So, you may ask why focus on the IITs well the first reason is that we are here. So, I am here at IIT you are all students of IIT or faculty members or some way connected with IIT and the second is that you know engineering and technology is of course very important for development outcomes you know product business money are all engineering services and really you know it is us engineers who are supposed to deliver these outcomes. So, it is you know so that is why we will focus on the IITs just to give you a picture you know roughly 200,000. So, these numbers are in crores. So, in the education I mean you know Indian budget is roughly 200,000 crores of which 60,000 crores comes from MHRD right and from. So, from that about the centrally funded institutions command about 3,000 crores of which the IITs consume 2,000 crores. So, this is how you know this is how the budgets look like. So, for example IIT Bombay support from MHRD is roughly 200 crores give or take 50 crores I think this is reasonable. So, I may be making errors of plus minus 20, 25 percent, but this is the ball path. So, as such if you look you know you may say that 2,000 crores is a very small fraction of 200,000 crores it is still 1 percent and per capita investment per student or per student is roughly 10 to 15 lakhs per student. So, that is the number that we are talking about. So, besides this the IITs also get roughly equivalent support from various research agencies like DST, DBT, Ministry of IIT and so on. So, our you know the investments roughly double. So, just to give you you know just to put things in perspective if you look at the Mangalyan project it was roughly 400 crores. Then if you look at ISRO its budget annual budget is 5000 crores then you know Maharashtra water supply and sanitation department which services 28,000 gram panchayats in Maharashtra has a budget of about 1000 crores and Mumbai University with 600 affiliated colleges has a budget of 400 crores. So, that tells you what the ball park you know what is the kind of support that we get and you know per capita and what are our you know what is the sort of relative picture. So, it is more than the money it is the intellectual space that the IITs occupy right that is far more important. So, if you look at the exams you know the JEE and gate JEE define you know what training and student needs to start doing engineering and gate defines what finishes you know what is the training which you expect from an engineer. So, these both these are administered by the IITs right and if you look at TechWip 2 World Bank project to World Bank project to improve the education technical education quality in India you know it says on chapter 1 page 1 that the gap between other colleges and IITs needs to be bridged and that IITs should pay a catalyst the role of a catalyst right. So, it is very important that you know if we are the role models to all engineering colleges in India then what exactly are we up to right. So, that is very important that we understand that. So, of course we dominate the research agenda, research funding allocation and moreover nowadays many state government you know school curriculum are now orienting themselves to these entrance exams right. So, if you look at NCRT or even Andhra Pradesh state board they are their definition of what is science actually is based on what is being taught or what is needed to get into IIT or what is being taught in IIT and so on. So, the definition of science in higher you know in school is itself dominated by what IITs you know are doing or teaching. So, I think that besides of course the social the sociology that we IITs are we have the bragging rights in this in this country we go out we are an IIT and you know lots of doors open. So, I think that I need not go through the process you know that there is JEE and there is advanced JEE and so on. Just I suffice it to say that roughly one in 200 applicants gets in. So, the odds of getting into IIT is one in 200 which is fantastic I mean nowhere in the world you know even in China you know the odds are not that hard ok. So, not that difficult. So, of course so I will call this as my measure of selectivity one in 200 that is the selectivity by which students get it ok. This of course if you go into computer science you are calling one in 2000 right. So, the selectivity changes by discipline right. At the PG level we have the disciplinary gate exam again roughly the selectivity is one in 200 of course that but the you know they inherit the selectivity from the department. So, a mechanical engineering gate versus the computer science gate has you know they have inherited that the you know the selectivity of the parent department ok. So, that we know. So, now let us look at the output side more carefully. So, we have you know so the university actually is producing you know is feeding employees or agents into three sectors the companies, the state and of course civil society. So, company needs employees, states need department engineers and civil society needs artists you know authors, sportsmen and so on. So, an output of of a university can be you know classified into these three these three core sectors and these in turn serve our. So, these are serve our society. So, what I am going to look at is the placements or the output of our of the IITs and see what sectors do they go into the economy, how they serve the economy, what companies do they join and who owns these companies and what do these who serve you know these companies whom do they serve. So, that is our basic analysis ok. So, here is the so the basic question is who joins where where are people going. So, this we have looked at the 2013 placement data. So, you know around March between 80 and 90 percent of the placement was done in 2013 by March or the middle of March. So, I am looking at the you know the seven key engineering department. So, aero, chemical, civil, CAC, electrical, mechanical and material science and the three key programs of BTECH, dual degree and MTECH ok. So, I am not looking at the two year or five year MSc's and the smaller programs of energy sitara or you know others or MDS and so on or PhD. So, I am looking at the key you know the core the you know the biggest chunk of of this data and that is about 833 of you know. So, about 81 percent of the total people who actually got jobs through the placement office in 2013 ok. So, the sample is a large sample or you know fairly you know a majority sample of the thousand people who were placed. So, it is about you know 833. So, of this 324 were BTECH students, 180 were dual degree students and 329 were MTECH students ok. Is this framework clear or any any quick questions? So, it is clear. So, 2013 placement office placement data core engineering BTECH dual degree MTECH and. So, this is the data that we will look at right. So, here is the first point. So, here are. So, I will just use the pointer. So, here are the. So, I have just the the rows are the programs and the columns are are the department ok. And each entry for example, 32 comma 11 that 32 people were placed from chemical dual degree and the average salary annual salary in rupees lakhs from 11 lakhs ok. So, let us look at this table carefully. So, you will see first of all that the CAC BTECH commanded an average 33.4 lakhs per annum ok. And the MTECH was 14.8 and you you will see you know you know just stare at this you will see what the what it looks like. Is it clear? Yeah. Any is the data clear? So, this is the first table that I would like to I would like to show you. So, of course, you know the Indian economy has yeah correct it is including the foreign. So, I will I will come I will classify. So, I will classify the foreign placement domestic placement and so on. So, we will come to that point. So, here you will you will see first of all that these numbers do not add up to 100 because there are you know there are some others which I have not put. So, the others column I have deleted these tend to do a lot of other things compared to the dual degrees of MTECHs ok. So, here in general you will see that. So, you will see first of all that the MTECHs do far you know the you know the MTECHs command a far lower salary than the BTECHs ok. So, is the training different? No, but the selectivity of the students the point of selection was different right. So, JEE exam is much more highly placed than the in the gate exam right. So, we see that that is seen here right ok. So, by the way you will see here that. So, the dual degree you know roughly you know I mean the electrical dual degree 16.4 this is the second high or the highest after the CSE BTECH and then there is a rough parity between the BTECH and the dual degree and there is a sharp drop in the MTECH column everywhere there is a big drop ok. So, the next category yeah the next yeah for gate it is difficult. So, I will come to it for gate it is because they in you know I am they they try to probably get into CSE then get it so did mechanical and then from mechanical to here it is roughly 1 in 200 again you know it changes for from discipline to discipline, but it is roughly you know top 1 percentile get it or half percentile get it. Yeah, but it they have inherited the you know the discipline time selection they have inherited ok. So, MTECH mechanical admission and MTECH computer science though they may have you know in the own stream it may 1 in 200, but at the BTECH time ok. So, here is the other sector wise data. So, this sector wise this classification is done by placement office right and we also did it using the NSSO NIC classification they roughly match, but I am using the placement office classification here and you will see that the first surprise. So, these are the class sectors engineering into ET finance IT fast moving consumer goods then consulting R and D education and others ok. So, this is the rough classification and we see here that in the BTECH right only 22 percent of the BTECHs actually went in for engineering right. So, 24 percent went in for finance, 21 percent for consulting and 24 for IT. So, roughly one fourth spent across the three service sectors and one fourth in the core engineering sector right. Now, in for the dual degree is roughly the same picture except that you know the consulting is a bit higher and the IT is a bit lower right and for the MTECH 50 percent you know MTECH actually went in for engineering job ET job. But you see here that in if you look at all the major sectors the engineering salaries are the lowest. So, you see that ET is 10.2, 13.0, 13.3 then 10.0, 13.2, 11.6, 12.9 then 58.6 the major other sectors you know 15.0 and well non IT right. So, engineers are actually paid the least amongst all possible sectors right largely speaking right. So, the second class so this is just by you know just by sectors and programs and degrees. Now, the second classification we did was by company ownership right. So, so company you know so there are company ownership and location of the job. So, if a you know the company may be multinational or an Indian owned company serving a global society or an Indian society. So, this can be you know we this classification was done by you know by looking at their sometimes at the web pages or their you know annual reports or some and we did this classification. So, I will just explain this. So, here are the five classification that we make the super GG is you know globally owned global revenues and the locations abroad ok students place you know getting in global companies you know serving a global client till and you know place. So, an example of this in 2013 was Sony Japan ok Sony hired people and they placed them in in Japan somewhere ok. Then GG is globally owned global revenues placed in India. So, all the other four are Indian ok and an example is Goldman Sachs right. Indian owned global revenues imposes which in you know in in hire anyone but this is a typical example of an Indian owned company which is serving a global client ok and globally owned Indian revenues. So, another example Proctor Gamble or for example Samsung electronics right and finally the Indian owned Indian revenues you know for example Tata Motor is that clear. So, this classification is clear right. So, just classifying companies by who you know who which you know where what is the ownership and where are their revenues coming from. You know after all we expect that a company serves a particular society and you know and you know and this service actually benefits the society right. So, we will we have we are going to use this. And then we then we classify by program across these five sectors right. Then we see here that the super GG you know this 15 percent. So, now these are in percentage. So, 15 percent of BTECs actually went abroad for jobs in the super GG and their average salary was 46.8 lakhs ok. Then the GG companies you know 41 percent of BTECs actually went for global companies serving a global client like LinkedIn or Google or Mercedes you know or Airbus right. And then you know so this is IG. So, Indian companies with only seven and GI and well Indian companies 21 percent with an average salary of 7.3 lakhs ok. And if you look at the dual degree you will get the you know similar you see a similar pattern and for the M tech the major the major sector is this the global companies you know 56 percent of the people M tech students actually went in for global companies with a global client exactly like generally electric or you know companies like this. So, if you see here the you know the II sector is really 21, 19 and 15 and is the lowest paid ok. So, here is you know the whole table in more detail right. So, we see here that it is essentially all the average numbers are here. So, for example, if you look at ET engineering technology well the total number of students 830 or 830 is 281 right in finance 134 and so on IT 198. So, these are the total ok. But of these you will see that the GG's are 116 right. So, out of these 281 a full 116 have gone to global companies serving a global client right. So, those serving an Indian client you know the Indian society are number 24 and 64. So, a total of 88 students or 88 graduates from roughly 800 actually ended up being engineers for Indian society right. So, that is what the data says yeah. Why do not you classify IT with the AD? Well LinkedIn is an IT company. So, it is not clear what skills are. So, one thing that I did was what are the skills that we train them for you know. So, we do not really particularly train them do any training for IT. Even in computer science not clear what is an IT training. So, it is a service it is a general generic you know service sector job is that clear. So, this we see that I mean the table or if you look at GG or if you look at finance then is clear that you know I think the data speak for it yeah. Well there are many computer engineering or you know product you know same you know for example, Microsoft your Autodesk AutoCAD so many I mean proprietary products right. So, for example, you know there are so many GIS companies you know ArcInfo or whatever ArcGIS LinkedIn is it is an IT company LinkedIn is also well I okay yeah correct correct. So, we have you know so all basically the excel sheet you know will be put up you know you can look at the data and you know examine. So, for example, Atkins so is it an engineering design company or a consulting company or for example, Pricewater or scooper. So, we need to you know we need to you know we need to do that classification carefully. So, we have done as far as we could do. So, we looked at their web pages we looked at what was the profile which the you know the jaff you know what was being posted by the company yeah. So, this is the this yeah correct correct. So, in fact, I in the future analysis I am going to ignore the super gg because they they are not you know if I if I am I am going to look at your locations in India because I am going to plot what are the you know what are the wage productivity for India okay good yeah. So, by the way the second thing that we did was how much the CPI matter right after all we are training them you know in so many courses and you know right. So, does CPI matter at all correct we I mean you know a typical student would have would do about 15 you know chemical engineers how to do 15 courses in chemical engineering or maybe more right. So, we expect that if you are spending so much effort in it that ultimately a better CPI should matter right. So, here is what I have done is that we have just regressed for each category we just regressed on on you know computed a slope you know how does by their CPI and then whether that slope if we if it is statistically significant right. So, only then we reported correct and then we see here that the only statistically significance you know slopes are these 0 0 0 0 say at 5 percent is and these. So, 1 2 3 4 5 6. So, these are these are the sectors where CPI matters right and you see here that the sectors are I T super G G I I consulting I I FM C G G G finance G G I T and a final one where E T matters that is Indian companies working for global clients. So, global clients like CPI right Indian clients do not care right that is what the data could be that the global clients have exhausted the exhausted the top CPI. You have saved up the top CPI in. Yeah may be but may be I mean. So, then may be. Indian clients go like. Well, but then but then that still tells us something to do right I mean we should probably not I mean you know we should maybe not we have failed people at 7 or God not God I mean something you know it does tell us something to do right. So, the point here is one important point yeah yeah I believe you know both you know I have just kept those which seem very small because there are some you know which are completely insignificant ok. So, I have not listed all the you know the major categories and all engineering categories are listed ok. All the engineering categories are here. So, you see here that. So, you will see what the slopes are by the way on this side are the guinea coefficient. So, the guinea you know a large guinea coefficient means that the salaries are you know varying in that sector. So, there must be some other knowledge other than CPI like for example, whether you whether you are M I organizer or not that seems to matter right. So, a large you know guinea coefficient tells you that there is some other metric which is going to which is mattering. So, here you see that for example, you know GG finance has a very large you know you know it has a very large coefficient. Here you will see other you know 209, 198, 213 and so on right. So, the so by the way. So, why do the service sector companies the GGs care about your CPI? Well they implemented by a cutoff right all students know that they just a 7.5 cutoff or 8 cutoff. So, they do not really particularly care about your CPI they are implemented by a cutoff. They do not really look at your transcript or talk to your guides or look at specific courses in which you have done just 6 or 6.5, 7.5, 8 cutoff and implemented. So, any questions? So, here are you know. So, there are more detailed data sets you know on the in the paper and also in you know I will put it up on the web page eventually. So, here is what you know one can conclude that from this very small sampler that global companies serving global consumers is the biggest winner of our IIT placement right and the in addition to that the super GGs. And we also see that engineering is the least paying of all major sector right. So, they right and Indian engineering from there is even much smaller right Indian is even smaller and about 10 percent of our output actually goes to Indian companies doing engineering for Indian society right. We see that the m tech program you would think that we like to think that the m tech program is our you know is the real service to the country. And you see that you know m tech program largely serve engineering global engineering say the global service engineering service sector that is what it is serving right and UG program really serve finance and consulting. And most of these profiles do not need you know any of your engineering training or they at least are not responding to the training that we are giving. So, these are the conclusions that we are that is what the data show. So, obviously this is a very serious I think that this is a very serious issue and we must look at it very carefully ok. So, now so there are let us look at the two problems there is a misallocation away from engineering and away from the Indian economy right. So, remember that IIT is not IIT IIT is a state funded institution correct yeah. So, actually I have tried to understand you know for example, a lot of we are sure I have tried to by the way I have tried to keep track, but it is not easy. So, in fact I have my recommendation my recommendation at the end of this is to you know is for MHRE to you know for at least state funded institutions have a very clear idea of where what we are doing right. So, I think that is a very important point. So, anyway so there are two accusations that or rather two you know conclusions that there are inescapable is that there is misallocation for a state funded institution away from engineering away from the Indian economy right and the second is that you know our training seems to be relevant right. So, the average salary for a service sector may be you know may 10 lakhs and the average salary for an engineer is 6 or 6.5 lakhs. So, are we are we capable of delivering an extra 3.5 lakhs value that is the question that we need to ask right is our training good enough to raise the engineering salaries from 6.5 to 10 or whatever 8 or 10 12. So, I think that it is you know it should it should matter in better salaries it should help Indian engineer company yeah yeah I will come to that I will come to that I will come to that I will come to that. So, this is a this is another you know where are the engineering jobs right you are going to. So, then well there are many responses to this why are we running big departments with you know there are no engineering jobs right in says or whatever why are we running the running such big department we do not need to correct. So, that is one level of response, but I will you know I will elaborate on that further. So, it is really required that we understand what is the cause for this. So, you know. So, in an economy we must understand how does production of value actually happen. So, I have just showed you some two pictures of two different blacksmiths ok, but in general there are you know say we if you look at biscuits there are various machines you know M 1 M 2 M 3 or facilities or factory and so for example, here you will see that machine M 1 or facility manufacture 10 tons per day and the operator ability say normalized between 0 and 1 is 0.3 right and a more advanced machine M 2 produces 50 tons per day and needs a needs a more clever operator with a scale of 0.4 and M 3 200 with a scale of 0.6 right. So, is this clear that there are different machines. So, in an economy for the same you know output there are different machines which are producing different goods and they need a different levels of ability or skill at at manipulating those machines right. So, it is like I mean we could have done it for other cycles or you know banking or other you know other sectors as well. So, the point is but of course that. So, this productivity per worker translates into wages or the salary after taxes rents training costs capital costs and so on is that clear. So, so if for example, if you just look at biscuits then this would be the sort of the wage you know productivity curve which is that. So, there is a minimum ability which is needed to produce biscuits there is a maximum ability which is needed to produce biscuits beyond which there is not going to be any salary rise is that clear. So, for example, at T 1 if your if your ability was less than 0.3 then you are not in the biscuit market. If your ability is between 0.1 0.3 to 0.2 you would be allocated to some machine. If your ability is beyond 0.6 well your salary is not going to change you are too smart for the job but you are going to get the same money. Yeah. So, higher and advanced high tech machine actually it does not need to be complicated. No it may I mean. So, actually so good. So, I will I will right right. More than. More than. So, different skills. So, actually. So, these these are all stylized you know what are abilities right. So, actually you have come to a point where you know abilities are they J level abilities are they English speaking abilities are they to be press buttons on a screen correctly what sort of but let me just use a straight line you know that there is one straight line only I am plotting ability. So, this is a stylized version I am just trying to explain I mean for for me to do that. So, this is just a stylization. So, actually I mean that is paper that study 4 which is coming and it is about how do rents actually materialize in this society. So, we will take it up you know in in fact you are right. So, a lot of our elitization is coming about because you know machines are too powerful and now you have to select one person from 100 applicants. So, you use your J rank it does not really matter. It does not you know the operator who operates say Goldman Sachs right is is no clever or you just need some person who will do the job provided he is in the top 10 percent right. So, then you can you know so I have let us not going you know so a lot of these elitization actually happening because of over production. So, let us we can come talk to that later. So, this is a simple sort of ability to wage curve right. So, if you have so this is clear that if you have between this then you are going to be in the biscuit you are employable in the biscuit market, but if you are beyond that well sorry tough luck you are going to get you know you are going to be paid the same as anybody in 0.6 right. So, of course that in a typical economy there are many products right. So, here you see that there are so the bank banking job cycle manufacture biscuit production and so on. So, all these curves will overlap you know wage curves will overlap and you see that how jobs how people with different abilities get allocated to jobs right. So, here you will see that people under A naught are unemployable between A naught and A 1 are going to go to biscuit market between A 1 and I think A 3 are going to be employed in the cycle shop right and beyond A 3 you are going to get a banking job right. So, just by the machines which are producing these items or the facilities and the abilities that they need we can come up with how exactly is the allocation happening right. So, it is very clear that for a person you need to identify your ability and convey it to the employer and by that you know the job market actually happen is this clear. So, essentially the IITs or whatever university performs an important role of labeling ability right. Otherwise you know if you employ you in the biscuit here and you mess it all up then that is not good right. So, now I am going to actually compute these curves for the 5 or 6 major sectors in the IIT placement game and these are I have picked up and I am going to ignore the super GG. I am going to look at these 6 or 7, 6 sectors which have employed more than 50 students. So, what I am going to do is on the Y axis I am going to plot their salaries on the X axis their abilities. Now, how do I you know plot the ability. So, what I am going to do is since CPI does not matter I am going to plot it using J rank or you know the department ranking right. So, I am going to rank the department you know and then rank the M tech and the B tech department and merge them and they just plot it you know. So, here is the you know here is the X axis right. So, the so this is 6 is the top and this is ordering for the UG. So, by the way so B tech others does not matter if you are not in CSE or doubly does not matter your club together. So, CSE by the way that is how the data shows right. So, it does not matter whether if you are not in CSE or doubly does not matter. So, these are this is how I have plotted it and you know then I select you know the top entry and low and behold here are the curves right. So, in this you see that for example, this curve is the finance global finance curve right. This curve there is the IT GG then this is the consulting right and these three are the engineering curve. So, if you see here engineering starts dominating in the M tech others right B tech others some you know you may get a consulting job, but otherwise engineering, but for others it is just going to be finance consulting IT and so on. So, if you look at the you know if you look at the sector you know the ability to wages curve now it is very clear how the allocations are happening right. It is because I mean and note that it does not really depend on what your department was just depends on your basically J rank or gate rank or gate department right. So, I mean this is the picture that essentially so excessive selectivity is actually you know we are select selectivity is just too much and that you know and then you have this the domestic engineering is this curve and global service is this curve right and if you are on this side of this intersection point well you are going to get allocating to global service right. If you are in the middle somewhere like NITs well then it is something else then there are those BPO job waiting for you if you do not want to do Indian engineering right. So, the sector curves actually reveal what is going on inside the you know in the allocation process. So, I am short of time so I will not you know I mean one question is that can we not can our training does our can our training be up yeah. So, I am just saying that because if you are 1 in 200 right if you are 1 in 200 you have I mean you know there is there is if you say if you I mean yeah I mean sorry sorry yeah. So, if your selectivity of the skill that you exhibit is really is over here then you know for your skill level you are going to skip domestic engineering and go to global service right. So, the point really is that if you want to place well in Indian engineering there is no need for any skill identification more than this ok. Now, the point is that if you are labelled oh he is a IT whatever you know AIR 200 right just the processor labelling has identified you right. So, if you say he is in the top 10,000 now Deutsche Bank will be confused right is this clear. So, what I am saying is that if you just label that he this is the top 10,000 this is the next 10,000 or whatever top you know 20 50,000 actually. So, at here this point is really you know out of yeah top 10 percent. You are doing a very branding you are making unbranded goods no label. 10 percent top 10 percent. But in any market there is a label that branding. Yeah. So, we can talk about this. So, branding by the yeah we can you can. By the way yeah I think you know just the view I think you need to write it down and you know submit it. You know you cannot just say this in my view and so on. No I mean in the sense that unbranded no no branding and unbranded these are very complex notions right. So, it is not that oh it is just you know you consume generic Vaseline it is not like that. You see by in other you know for example Berkeley California the best department in civil engineering in the U.S. it admits one in six okay. So, the admission rate at Berkeley California you know Berkeley civil engineering is one in six. Now do you would you call it just oh it just you know just throw away the branding no. So, the whole point is that after you get in there is some training which makes it you know which is being done. So, they are taking people one in six and they are converting it to something special right. So, the whole actually that is the slide I missed sorry it is this that why is our training not making them not fetching 13 lakhs in the Indian engineering market. That is the question yeah. Can we go back to the graph? I have written a problem on the statement in the sense it says that if you have higher ability you go to global because you get identified. But you made a converse statement saying that the entering industry does not need ability that does not what no no entering if you entering industry you know domestic engineering you do not need all this you know competence or the ability this does not come out of any of this the converse is not converse. If you have the ability the probability that you go to global that does not mean no it does not mean no no it does not mean that is the statement to make no it does not mean that but if you assume that there are machines which are used to manufacture stuff and what skills do you need to operate those machines or what sort of technology are Indian industries using investments in R and D are they making so if you look at all of that then as it stands they do not probably have room for you know somebody really more skilled so I mean they so I am stretching it but there is some pay money salary in the manufacturing sector which is which is domination which is pay because it is a global economy so but that is outside I mean in the sense currently as it stands as it stands they cannot they do not have the ability to reward higher skills so all these girls are saying so they do not have the ability to reward higher skills correct well the purple purple curve would have to hold admission test doge bank would have to go out to because all those top 10 percent the top 1 percent that they want are now they could be Surat Kal they could be you know you do not know so they will probably have to hold independent entrance test and then identify the top 1 percent and then you know give them job they would bring it down the cost curve would shift down because the finding cost you know how do you find that the sorting cost become substantial is it clear okay so that this labeling the J actually enables the labeling you know enables the identification of where are these talented guys so in principle we are you know we are labeling people we do not need right so now if you look at but still there are a lot of people who you know who feels you know it's like you know it's gold buried should we not discover it how can it be left lying you know lying there so much talent which is wasted right so let them you know get global jobs and let's see you know what happens after all do they not remit the money home does it not you know create more demand in the economy and so on so that argument also one should really right and see whether it you know holds true or not right so here is where I you know so I will explain how do really this globalization and production really work so let's look at you know a simple company by the way this stylization is by Mastin and Kramer okay and Mastin got the Nobel Prize in economics from Princeton and Kramer is an economist from Harvard so it is not my you know it's a branded concoction okay so you look at this so it's so it's a let's look at a small company which has an assistant and a manager okay so the assistant has skill B manager has skill A fine and the production is defined as A square B so A is you know the skill of the manager square times B okay by the way this economist will know this is the famous cop Douglas production function right so variations of cop Douglas so it's very standard right so you see that if there are there is a person of skill 2 and skill 3 then who would you make the manager well the higher skilled person you will make the manager so for example 3 square into 2 is 18 but if you made 2 the guy with 2 as the manager 2 square is 4 for the 12 right so it's clear that if these 2 people form a company then the 3 person is going to be the manager and the 2 person is going to be the assistant right okay so now let's look at a society okay so you can see that I am not looking the wages how do the wages get divided I will I will skip that but if you look at a society suppose this society has you know 2 people of you know it's composed of 2, 2, 3, 3 okay 2 people of ability 2 2 people of ability 3 now there are 2 ways of forming a company so one can form a company which is 2, 2 and 3, 3 or another way is 3, 2 and 3, 2 correct so now you see here that if you form the company 2, 2 and 3, 3 your net output is 35 right but if you form it 3, 2 and 3, 2 your net output is 36 correct so it's clear that if the social output has to increase then the companies are mixed right so there is a lower you know so you don't say that all you choose you stay apart and you form your own company we are going to form our own company no in fact it makes sense to have these mixed companies okay but on the other hand so on the other hand if you look at 2, 2, 4, 4 right so now the higher set the elite state is slightly further ahead not 2, 2, 3, 3 but 2, 2, 4, 4 then you will see that in fact then you will see that in fact the 4, 2, 4, 2 pair has a production of 64 and the 2, 2, 4, 4 pair has a production of 72 right so you see that mixed companies are formed when abilities are comparable close by and when abilities are far away you know it is the separate companies which come into which come into place right so different this further abilities further away companies are separate abilities are close yeah huh it's a cop Douglas so it is I mean branded okay okay huh yeah yeah like all branded things no no no but it is no no it I think no by the way by the way this maskin is a you know he has he has a lot of data also to show that the lot of yeah lot of productions I mean so the cop Douglas by the way only recently in the last 10 years is the cop Douglas seem to be breaking down in the US okay so in the last 10 years something really fishy is happening but otherwise the cop Douglas production functions held true for 50 years okay since the data they started gathering data yeah no so actually there are no no so I agree so these are so by the way so there are many papers which connect ability in one country with ability in another for example now if you look at the PISA you know the rating ASA or PISA rating the top 1% yeah that's not going to mention at all they give a breakdown correct correct so I think that but there are a lot of research is you know when people migrate from here to some other you know what if they have these skills how does it you know translate skills in that other a lot of work which has happened so I mean these are all stylized models but style have some validity so basically just you know the the point to note is that separate companies when abilities are far apart mixed companies when abilities are close by right yeah so now let's look at globalization right so supposing that you have your own society 2233 and you have a you know a society which is 6 6 nearby near you okay right there is another society and you have your society 2233 and that society 6 6 so before globalization you know you would have done this this would have started as 6 6 right now supposing that you those 3 3 you make them for an exam and then you discover that they are actually 4 4 that they pass the entrance exam and they are now labeled 4 4 now what happens is the first thing what happens is that they will you know form this company right the 23 will break down and 2244 will start and if you globalize then this is the new society and in fact you will see that 22 and the 6 4 and 6 4 dominate right so actually what happens is that those 4 4 join companies with the 6 guys and form companies together correct so if you see this what happens is that you know let's just understand what's happening well after globalize even more inequality and one key important result is that there are far fewer managers in your economy now in the 2233 there are 2 managers and 2 assistants in your new economy there are you know there are 1 manager of skill level 2 and 3 assistants in other societies there are 2 managers 6 6 right so you have lost you know management skills right and you have you know increased inequality and you have you know there is obviously less research and local problems because the manager is not not familiar society correct so just by identity you know identifying a talent which you don't need you have you know change all of these pictures right so all of this allocation and how you know so essentially what I am trying to say is that so we are we seem to be intent on a trajectory you know if I just say Bharat in the elite you know we are so you know you are making companies like this and now you see suddenly this identification you are selecting a small bracket and showing them to be of a different skill and essentially IITs are ending up enabling these companies right just by identifying that look here are these talented people right so they are essentially so you know I mean the conclusion of this is that excessive identification of talent can actually be harmful in a globalized world right and that merit systems have to be very carefully designed right you have to see what can happen right you know so I think that these 2 are the takeaways from from whatever analysis that we did so so I think just I have 5 I have some 5, 8 minutes 8, 10 minutes so I think that one must really I mean now step back a bit yeah step back a bit and understand you know what is the what is the political economy as they say you know how is the how are we organizing our society so frequently say that no we have there are still 7, 8 minutes well there is by the way data so we have no no so I think it will be much different so for example how many people become assistant is going to be very very different so if you look at what are our you know what are yeah we still look at data but I think that what do these jobs when they graduate what are they really doing in Deutsche Bank if you join Deutsche Bank what are you going to do right so it so I do not think that MIT graduates are going to join you know they may join finance companies you know add a very different level I mean well we have to see the data yeah yeah yeah yeah yeah yeah yeah yeah no no no no no of course not you see but you know for example we consume 65 kilos of steel per capita right and the you know US consumer 250 to 300 right so we need to you know if we are we have to have more buses and bridges and so on we need to make at least 100 or 150 kilos of steel per capita right so for them to move into a service sector economy is very different from for us to you know so if you look at our you know all these industrializing measures we are nowhere there so we need so here is the the final so you know when I say something is a meritocracy you know if it is a state funded so it is essentially a closed loop so there is a university which labels right and there is a production system which uses this you know labeling and generates more output and there is a political system which distributes this output so that everyone is better off right so after all in the labeling there are winners and losers right because you know if you know a company does not know by label then it is you know it is going to make some average decision right so so in the sense then there has to be a political system which actually redistributes income and then the people in their wisdom say yeah I like the university right and that I support it because you know the sorting actually benefiting me so if it is a democracy and 50 percent of people must vote for the university then all of these things must happen right so in that sense and then the you know the people vote in vote for the university so this I would call a meritocracy you know state funded meritocracy so if you look at you know you know how what we pass here you know what does our elite systems pass yeah so I think that if you look at these four four conditions I think we fail the first three okay and maybe popular support the IT do have right so we do not we do not redistribute we do not our production system cannot use the labeling and well our you know our labeling is of course not correct but of course the ITs do have public support so why is that so really if you look at our education system it is really an exit you know hatches exit door so you know you are in English you know you are in good one Wadi you write English you learn English and then you are in error then you pass engineering exam you are in you know Bombay and then you you know write J and then you become global so at all points you are leaving the society that you know the education is actually helping you go to the next society it is not really training you to be a better performance not closing the loop it is not helping you know become a better black you know bridge manufacturer or whatever it is a society which is help you know oh you are good you should go ahead to the next society so it is actually a train in which we are you know helping people move to the next next bogey right so we will see that so you can you know I mean there are so this assumption this train assumption is a very mysterious assumption right and so we are essentially you know we are created this fair and proper people can migrate to the next society right so each of these are very fair exam based on merit correct so you learn English you know and you go to the next society then you pass some other competitive exam the next and so on so it is a mysterious I mean I do not know why we have built that system so the outcome of course is really terrible you know if you look at 50 percent of people their practices have not changed in the last 50 years you know they are still using the same way of you know the sanitation water system you know this jaggery this is a jaggery plant by the way look at his legs this is a jaggery plant in Daun so all the photos in this talk are from nearby area they are not from you know Odisha or you know really Bundogal places they are right nearby 100 kilometers as the crow flies this is from Daun Daun we do not have to go too far right so now of course what to do right the big question is what to do so I think that so one is to realize that the road I mean we keep talking about MIT so the road to MIT actually goes through Gavanwadi and goes through the kitchen of Hirabai in Gavanwadi so that I think we should realize that you know it is going to be the source of new problems and new engineering jobs of the highest caliber you know by the way so this Chula is though you know not less complicated than the IC engine right so I think that the focus that this realization is really very important that this is where the engineering jobs are and we need to really understand it much better right so I mean what I am saying is that there has to be a slew of new companies which are you know saving drinking water system for better sanitation system public transport you know just enabling more public transport or more logistics of it you know influencing policy and essentially delivering efficiency and making it pay 10 lakh rupees per annum I think it is possible right the gg jobs are paying 10 or 12 lakh rupees per annum I think if we do all of this and you know it is possible to pay a fresh graduate 10 lakh rupees per annum if we do all of this right I think it is possible so then how to do it I think we need to you know redefine engineering it has to be sorry it has to be more it has to be more interdisciplinary you know focus on seductively planning engineering services engage with society look at the informal sector you know it is yeah 80 90 percent of our manufacturing informal sector right so they have not really big opportunity so and of course broad based engineering so whatever you do so you know the Amravati college was doing the same thing then you know the VNIT is doing the same thing Kolhapur college is doing the same thing so and I think that is a great situation so if we have a broad community of engineering then that will confuse the gg okay so I think that the allocations would be more proper by the way that is not the point that is not that is not the point so I am just saying that just for fun that it will confuse the gg for sure so I think that you know there are if you look at the history of you know even water supply when did water supply come into Europe and did it need the Navier-Stokes equation to produce water supply system the answer is no okay did so by the way they still are battling with sanitation right so I think there are many problems that have been solved even for the developed world so I think that so essentially an engineering which is full of case study that which you can share I fixed this I did that I redesigned this and I implemented that and so on so I think the theory and practice and fieldwork should go together so that is the main point right so of course MHRD our parent department and DSD should also behave differently so for example the first thing they should do is to maintain data on placement in prescribed format each MHRD institution must do that that is the only way we will know what is happening right and then they should rethink on you know all these fancy tech wave and gate and accreditation and so on so we may be we will be accrediting of you know a model of engineering which we do not need right so we have to be careful then for DSD reserve a substantial part of research funding for regional outcome recognize case studies and take as valid research output right then you have various things right so most important is actually a shared vision of knowledge and practice of history of science and understanding of the history of science and technology what it really means what was the past what is the future right and the elite I like IIT and I think that it should be a pillar of civil society and not you know a hand made in for global companies or global engineering it should be a pillar of our civil society so if you do not do that the road to Gavanwadi is going to go through MIT okay so that is what you know so in other words right it is what is going to happen is that it is the global people who are actually going to serve Bharat you know through the World Bank and through you know all these other you know all these other western elite universities all of them have now a South Asia department right and they are coming here and they are solving our problem and they are solving it in the way they want it okay so I think that it is important that we have the shared vision of knowledge practice so there are some references this talk will be on my web page thank you yeah question how that interface with the society or industry happens and it is meant to be relevant right yeah yeah yeah yeah so I think that I have we have explained so of course are you all right that is certainly I mean I think the coursework and I will we can detail it out later any other yeah there are two things you mentioned one was making the data companies big and large the other is that having sorted people to a large extent with free selectivity you are now letting the global companies the second problem can be solved by essentially privately funding the day no no I so I think that I still think that elite institution is a necessity no no our sorting which is why I think you have become the relevant sorting for those yeah but I think I think I will take it offline because this is a more involved question yeah no no correct no certainly I mean I but I think that you know I think MHRD should really design these formats and maintain them I think that MHRD is about time that you know it because these are this is very useful I think we definitely need more data yeah yeah you tell our student what we expect them to do after four years this is like you said open-loop system so the students learn whatever they want to learn and do after graduation whatever they want to do they are no need such expectations that we test them they will be an optimization this is one I think the problem and the second thing is that is quite between what we do in IIT and what is needed in the country reporting of disconnect. I think that disconnect is not really good. So, we need to introduce that connection between the Indian problem and what IIT does, bring a connection then the country will be lose. No, that I am saying that. So, I think this is just to show that the data also motivate. So, I think sitar has been arguing for this for many years. So, this is just saying that the data supports the argument. So, data is supporting that we need to look at engineering in a very different way and something which can be shared by a lot of people not just IIT, but by regional engineering colleges. I think broad basing the practice of engineering is really very important you know and that can only be done when we look at local problems. So, it is per necessary that you look at you know. So, engineering as a social skill you know looking at how problems are they come about manipulating you know talking to stakeholders trying to deliver solutions. I think companies or society or you know collectors or whatever or even Hirabai from Kaunwadi. Yeah. So, you go to an enriched domain to solve problems and you know and to a chain of directions you go to a more enriched domain and you solve problems there and then finally, come back. So, you show the strain where you train to get to the next stage and to the next stage and so on. Yeah. So, I think there are the you know the you know I have written the paper in current science about it that loop should close and at least it says that you know what steps should be I think that is an easy problem you know what should be the curriculum what should be the course work. I think for example, that you know the two just going out and solving real problems I think that is the simplest thing that we should start doing and I think that I am not against theory at all I love it, but just that it should be parallel to you know field observation that is the main yeah. Thank you very much for the analysis. Well thanks for coming. So, at the end I think you said we have to redefine engineering and you showed some slides and pictures that you said you think that this kind of salary is about that can be done. Definitely. Definitely. Well a firm belief well I can prove it. Is it a gut feeling? Is it a gut feeling? No, no, no, no, I mean Sitar is running an M tech program based on that I mean we have not reached it, but I know it is possible I mean I need we need support we need to sit down together, but it is possible. Well I mean there has you see I will tell you in Shahapur Taluka you know spends 3 crores per year you know and installs these schemes drinking order supply schemes each costing 30 lakh rupees roughly you know easily 2 or 3 fail in the first year itself. So, there is 30 lakhs waiting for you to save if you can save one drinking order scheme you know made you know whatever 10, 12 lakhs. So, there it is easy I think the inherent inefficiency in our system is so much that 10, 12 lakh rupees a year is easy money I think it is beyond that that the fruits would get higher. Yes, but nobody has pushed it nobody has pushed it. So, I think that exact no, no exactly but that is where I think that IIT is an elite institution should push that Shahapur is failing we will not allow it to fail. So, I mean in western countries I think more or less it is organically grown society there are society needs and there are university which works on those needs and there are institutions. Correct correct. So, it is and but we continue that is the trouble. So, I am just wondering is it possible to is it possible to do what you are suggesting because it is a very big problem. No. There is a civil society there is a political government there are institutes like this and all of these have to work in tandem to make this happen. Actually the problem is big, but I think that IIT has is any it is I am not saying that Rajaram Bapu Institute of Technology should do this. I am saying IIT should do it because IIT has the local standard you know it has the you know it has the validity to legitimize this process exactly. So, I think IITs can do it. Yes. Verify. Yes. I will tell you one thing I want to participate you in the verification process, but I have just one question. Yes. The jula organic architecture of the slug or such thing which has really implicate engineering, implicate architectural excellence in that which I never thought. If I have to really go to MIT to go over for it, do I have a claim to capture those engineering do I have really a claim to capture that techniques or do I look at them also from my knowledge which actually MIT thought. So, these are I think that you see I think these are complicated questions, but the answer is that it does the road does go like the way I said. The history is the you one can look at many historical I mean I think that there is a wide body of you know how is how to represent science and what are scientific discoveries, what what happened in India are the slums really organic enough are they enjoying themselves there you know. So, there are many questions we should not just say because traditionally it is good or just because it is bad I think that we need to make a I think that you know that I just recently I was reading that Charavaka and you know that Devi Prasad Chattopad had written about Charavak and the Upanishad and the you know the whatever and the dialogue between the Charavaka making you know Ayurveda and you know completely empirical body while Upanishad completely you know well whatever a materialistic argument and how they. So, I think it is a very rich tradition. So, we need to look at of materialism as a materialism. Yeah. There is also not even as far as the places that you mentioned why we are having a training and education system. Yes, definitely. Yes, yes. Yeah. Yeah. Yeah, yeah. No, but I think I think. Yeah, but let me now I mean no yes sure. No, so I think that culture is a very different cup of tea or music. I mean there are culture for example cricket. So, there are many examples. Now, does cricket need training? So, I think these are I think I would like to restrict you know maybe if we should look at case studies. I think it needs examination that is all. No, I have no knowledge of FBI. I mean I think we should look at it. Yeah. There is room for any theory or abstraction. Oh, plenty of room. I mean I think that so the essentially as a micro level I can understand how this works but as a micro level if a person has to contribute in the form of research. Oh, yes. So, I think there is plenty of. There is also moderate implementation problem. Oh, there is plenty of research in it. Plenty of research. So, do you think that at the outset it is a now there should be study on people who are entering the IIT system to JEE. Yes. Are they really made for engineering? Yes. Will they have integration for engineering? Yes. Maybe there are mathematics. Yes, you know JEE makes a profit every two three several crores and they can just fund you know a few PhD just studying this whole problem and they have not done it for so many years. Well, when we say that they are going for IT concept and they cannot. They are not really made for engineering. Correct. So, I think that exactly what I think is worth studying. I mean and there is the money JEE and gate are making huge profit. The entry into mathematics physics chemistry. So, by the way Gao in China does not have it has language, mathematics and it has three papers. There is language, mathematics and I forget it does not have I think science it does not have. That study is important. Language. No, no, no. Are they. I have studied how the Chinese you know admission work. Yeah. So, like. I might be simplifying, but do you want to like people to work with governments to solve local problems? No, but many or work with you know blacksmith or work with you know public transport system or work with you know Karzat Taluka bus depot. It seems like you have to work with governments. And one of the reasons why I think I mean I am trying to get in the right direction. But one of the reasons that will that put people off may not even be the money really. It is the fear that in a government setting merit will not be. No. Yeah. So, by the corruption in the matter of design. Corruption is not about it is a matter of designing protocols which you know which sidestep corruption. Everywhere there is corrupt. I mean if you. So, by the that remind Ashok Gadgil professor from Berkeley was here. He said that he when he was a student his guide from Rosenfeld or some really famous physician or physicist used to travel to meet congressman. And they were as corrupt you know in those Chicago club or whatever. So, but just meeting congressman and that puts a pressure. So, just meeting politicians MPs MLA just you know changes the discourse that you have. So, we have to make that connection. You know that just that IIT is there and you are saying show me the show me the data side. You know I think changes the discourse. I think that is a big offline. Yeah. Offline. Offline. Offline. Yeah. Offline. Offline. This is yeah. I agree. But it is I we have to and you are from Sitara. So, we can discuss. Offline. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Yes. Then he has got without a model like this. You are looking anyway. Yeah. Yeah. Yeah. Yeah. I think maybe if he has something yeah definitely I have not studied you know, school or whatever not really school or lower I need to yeah. Yeah. Yeah may be I think you should there this last question maybe others I will just be outside. Yeah. No, I think that this talk is you know trying to build to that because I think that I just presented in the senate not going to get passed, but if we give these talks talk about it and you know maybe there is chance that something like this gets passed, so it is towards that definitely. So it is to you know just increase the discourse on this issue, yeah, so last question. Yeah, this is a good excellent point, so I am I am saying I am saying I am saying I am not for do gooding okay, I am for creating value and making money okay, so I am saying that you will you know there is knowledge or there is value to be generated okay, so I am not saying you know I like do gooding, but I am not relying on that, so I think that developing our so development is a perfectly valid economic job okay, so I make so do gooding is so you know is that clear, so when I am saying that yeah, so that is wrong, sitara or whatever I am not for do I mean do gooding is in is a private matter is that okay, so thank you.