How To Transition From Economics Academia To A Career In Data Science - w/ Bhoomika

What exactly does a masters in econometrics entail? When we actually do econometrics, we combine real world data with a theory. Seen with econometrics many times we are trying to estimate causal effect and not correlation. I'm Nam Pandandy and in this episode featured not expert is Bumika Jagi. Bumika is a data scientist and has two master's degrees in economics and applied econometrics and data science and analytics respectively. People who are studying economics should be more aware. They should come out of their economics bubble. And the second thing is invest the time in learning a programming language. Is there any mathematical proof for one of capitalism or socialism/communism to be better than the other? I wanted to understand the inequalities in the world and why they exist and so why some nations might be

exist and so why some nations might be poorer. It sounded like there was like a gap that existed between academia and industry now that you are in the industry. So what are some ways in which those gaps can be filled before it's too late? We go with the fascinating intersection of data science and advanced economics as Vumika shares some crucial insights to set any economics student out there for success. At least if the goal is to transition to industry to complete your PhD you need five to six years and you'll be able to only publish one or two papers by then if you're lucky. How much overlap is there between just pure data science and economics style data science? There is a whole new world which you should be aware of. In line with our theme of learning from somebody just two steps ahead of us instead of an

just two steps ahead of us instead of an expert, my goal with this episode is to highlight this fascinating career track and how to forge your career in this domain. This is the Ready Set Do podcast and please subscribe on YouTube and any of your favorite podcast apps for weekly episodes featuring not experts. And now without any further ado, here's Humika. Welcome to the Ready Set Do podcast where we learn from journeys of not experts who are just two steps ahead of us. Bumika, welcome. Hey, happy to be here. So nice to be having you and finally talking to you about econometrics, economics, and all of the other interests that we're going to cover here today. So, let's jump off the deep end here. And if you had the platform to share anything related to economics to current economics students, what would some of those things be?

what would some of those things be? Yeah, it's a very broad question to answer to be honest. Um it's very difficult to get down to two or three points but I'll give it a shot. I think people who are studying economics should uh be more aware. They should come out of their economics bubble especially the grad school students and the second thing I would say because there are a lot of opportunities out there and should be more aware and the second thing is invest the time in learning a programming language and I'll talk more about those later on but I think I think yeah that that covers what I saying. So yeah, that's really, you know, almost like a controversial take, if you will.

like a controversial take, if you will. So just to backtrack from that a little bit, can you help break down for our listeners what economics is? Because when you think of that, often the first thing that comes to mind is, you know, finance bros wearing your half uh you know, West Patagonia vests running around in Wall Street. But I have a feeling it's not just that, but it's so much more and also like so much less if you think about it. So from your POV, can you help break down what is economics exactly as it pertains to the academia at least? Yes, definitely.

academia at least? Yes, definitely. Actually, you're right. Many people who I talk to would be like, oh, you're studying economics, so you'd be very good at investing or you do stock markets or something like that. Not true. Um there can be so economics is a very vast subject and under the broad umbrella of economics sure you can have a specialization where which can take you to be very good at stocks or financial markets or something like that. But unless you actually do that specialization it's not really u it's not really that it's more about understanding the choices or the decisions we make as individuals and as society. And it's uh it's driven by a few uh I would like to say not driven but like assump we make a lot of assumptions. Economists make a lot of

assumptions. Economists make a lot of assumptions and we build models and it's very technical and it's so it's like we are studying how do I decide if I want to buy this or if I do not want to buy this or go here and that is at an individual level but we can also do it at an econ do the same thing at a country level or a state level or you know at an organization level so it's basically understanding all the decisions that we take and how do we take those decisions and can we reach an equilibrium which never exists. It's a utopian situation. Yeah. But at least we want to understand that and and more broadly I think a lot of times when people or a lot of times I was studying

people or a lot of times I was studying econ was was due to a specific reason. And I wanted to understand the inequalities in the world and why they exist. And so why some nations might be poorer, why some people are poorer and why there is a gap that we cannot fill or at least we're trying to uh get to close the gap but it's not as easy and why is it that why why that happens. So it's more of that rather than being a finance guy and maybe you plan, you know, if you do that specialization, but Exactly. It's just one of the multiple strands that you could potentially look at. But I'm very intrigued by what you just shared here. I think that opens the

just shared here. I think that opens the door to, you know, really just so many questions I have. The I I think the first one that comes to the top of my mind is you know say somebody that's listening to this maybe has a passing interest in economics but never had the chance to you know formally study it. What would you say are some really fundamental concepts that you know are just you know for like for instance for math we have say algebra like if you don't probably know algebra or statistics or maybe calculus you probably you know don't really understand math so can you help break down for us what would be some of the most fundamental concept just in terms of terms for economics that people could then go out and you know start looking

then go out and you know start looking for or just as a beginning of a rabbit hole if you will for lack of a better term. Is it are you trying to ask if someone's get trying to get into economics what they should know exactly like where should I start if I want to get a really nice solid but fundamental level of understanding of economics so I know supply demand is a term that's thrown around equilibrium as you said maybe it pertains to that or not but really just looking for some umbrella terms that are the core of economics just from you know like a study p yeah so there are a lot of online resources but what where we always start is

but what where we always start is microeconomics. If you talk to talk to an undergrad or even grad school, the depth at which we study those different concepts changes as you go from undergrad to grad or PhD levels. But it's always starting with microeconomics, macroeconomics and the and that's the building block of everything. Interesting. That's that's where you start. And um there are also definitely courses in calculus because we use uh linear algebra and calculus in understanding or developing a lot of concepts in micro and macro. Um but I think yeah that is where you start. So by when I say microeconomics you'll be covering consumer theory you'll be covering producer theory you'll be covering um equilibrium or uh where both of those match just like demand and supply right and then there's also a concept called game theory. uh

concept called game theory. uh some people might be aware and so all of those things I mean I think that is where we start at least at the micro level and then same thing happens with macro too and you can go into the depths of all of the things that we study there but that's that's that's the building block of economics that's yeah that perfectly answers my question that's exactly what I was looking for in terms of where should one start because it's from the outset it's just so overwhelming that you know it's just very easy to get just completely burdened I not even knowing where to start. So going back to what you said about the fundamental building blocks.

about the fundamental building blocks. Um I know you have a bachelor's in economics. So maybe we can start there. Can you help uh walk through what that experience was like? And that will kind of be in the backdrop of you have multiple M's degrees as well. So I'm just kind of trying to eventually start here and then close the loop on the type of things that you're that you're taught in bachelor's versus masters when it comes to economics and you know really just generally what your experience was like what were some of the things that stood out the most for you that were most uh impressionable when it comes to leaving a mark on you in the long run.

leaving a mark on you in the long run. Yeah definitely so I think I I I I have an undergrad in economics that is where I started and that's from Delhi University. Um it's a pretty popular course in Delhi University and you start with these building blocks of micro and macro. So I think undergrad is sort of generalist. We don't how many years is that? Sorry, how many years is that?

that? Sorry, how many years is that? Yeah. Yeah, it is it is for me it was three years but it keeps changing. Uh I think there was time when it was four years and it's back to three years. So for me it was three years but it keep back to between three and four. That's interesting. Okay. Yeah. Yeah. Yeah. Sorry but go ahead. Yes. So uh yeah. So undergrad was in econ. A lot of uh u you'll get to see a lot of different things that people study if they study economics but they don't go into into a lot of depth I would say. Um so you'll you'll have a breadth of topics. Um and for somebody

breadth of topics. Um and for somebody who wants to actually go ahead in economics I think that's a good good thing to have but even after that to be honest when I came to the US I uh studied more different kinds of economics and some of which I did not even know existed when I was doing my undergrad. So although it's breadth of topics that you cover in uh undergrad, it still leaves out some things you know but it's still a very good starting point and and from early on I think it's a it's very math heavy uh interesting early on you'll have a lot of courses in math and I think it is important for uh at least the kind of economics that I

at least the kind of economics that I was doing. Uh I think it is important to say that it's the kind of economics that I was doing. A lot of math was involved from the beginning. And what kind was that? Just so just so we're clear. I think I think I think uh many schools and I'll go into the depths of it later on. Many schools can make it a little less technical depending on how they teach things. Sometimes it'll be more focused on um understanding the theory not exactly maybe how it was built not going into the model etc but that's not how I studied it moving forward I became more and more technical uh but so yeah the undergrad was that and then I did a

the undergrad was that and then I did a master's in economics too in India again in Delhi and um uh and that became technical quickly very technical And that was also the first time I was actually doing STA uh using STA. It's a software programming language which most economics people use. I think some other social sciences students also use it. Uh and programming language you said. Yes. Okay. Interesting. It is a statistical software basically. And can you maybe for somebody that isn't super familiar with that, what would be like the closest equivalent common programming language R or and stuff? Yes. R I would say R. Um but we're in economics is sort of a loop or a bubble where kind of

of a loop or a bubble where kind of everyone is using STA. I would say at least most people are. Uh that's not necessarily a very good thing, but that's how it was for me at least. And I also got a flavor of using Python. So my professor was trying to teach us concepts in economics in Python with nobody in the class having any programming background. Um that's it was rough. So I although it was it was technical it was nice and everything but and it gave me a flavor of something that I did not know before but um it it was too quickly we didn't know the fundamentals I say I would say and I think that is necessary and we were doing something some economics things in Python so okay so a bunch of

things in Python so okay so a bunch of data manipulation stuff like that feels fairly advanced to me for you know especially for a Oh really okay wow more than that and so that was I think um at least from my point of view u was not very useful at the time at least but it was good to know that something like that exists so how did you cope with that then at the time what was that like it was not a major part it was just one course you know very little part of that class I would say got makes so yeah that was easy to go about but I think it is

was easy to go about but I think it is still important to have some kind of fundamentals and then and I'll go into more detail of what that entails later on but then I I was trying to think you know at the time what I should be doing next and I was sort of back engineering I started looking at the profiles of people who I was very interested in and what they did and they and I realized that I should probably be doing PhD because a lot of the people that I was looking at had PhD in economics and so my natural next step was to apply for a PhD in economics and I decided I'll be doing that in the US and so that's another different process and I applied here and I got in

here and I got in um at the time I was very interested in development economics. Um so PhD in economics works very differently than other PhDs. Um it's not like can you exactly can you briefly walk through the what that process was like specifically for you for economics uh in the US um like the application process? Yeah. Yeah. like just roughly I mean you don't need to get into the details but I'm just curious to understand what that's like. Yeah, it's pretty much very similar to all other courses in terms of applying but they focus more and or they should at least focus more on the maths course on your GRE. Okay, that's helpful. Yeah, at the time it was like that. I think most schools now wave off GRE. Yeah, that's true. Yeah. Yeah. uh

GRE. Yeah, that's true. Yeah. Yeah. uh but the having a research experience and are having um a lot of math courses in your undergrad is very very helpful. Uh if you don't have research that's okay if you're directly coming from an undergrad but you should have a lot of math background. Um that is what at least most schools look for. Um and so I got into a PhD program. I so I was saying PhD is very different in economics is very different than other engineering programs because in other engineering programs you tend to do lab work or experiments and that's the most important part starting your second semester or something but that's not true for economics at all. Um so for the

true for economics at all. Um so for the first two years around 2 years you need to complete a lot of core courses that are very heavy very math heavy very technical and they are needed to go to the next stage of your PhD journey. So before you do any kind of research or contribute to any kind of research or experiments, you need to go through all of those um courses, all of those exams and clear that and it's very different than a typical other typical PhD programs because I had friends in other programs and they did not care about study for them. The mutin was experimenting or labs or things like that. I see what you mean. I study was not an important component but yeah so uh even within the PhD actually I pivoted quite a bit I went into uh thinking that I'll doing I'll be doing I

thinking that I'll doing I'll be doing I will be doing a lot of development economics but uh within the PhD I realized oh you know house of finance is something that I did not know existed before because I was not exposed to it before and now I'm exposed to this new thing and I really like that um you know subject subject and I wanted to specialize in that subject. But a lot of those things happen within your PhD journey and I realized pretty soon that the PhD is very concentrated on publishing papers and not exactly on what I wanted from it. And so this coupled with other different reasons I decided to not go into PhD at all. Um and I pivoted. I did a career transition and I moved to data science. So I I decided to do a master's in data

I decided to do a master's in data science. Uh but I got a degree in econometrics and so uh yeah that was my brief uh intro to what my jet is. So yeah I mean that that is already so fascinating. One quick question on that. When you say you were working on development economics can you help break down what that entails exactly or like at least briefly? Yeah. So, uh, development economist is basically trying to understand a lot of at least why I went to in into that was because I wanted to understand inequality as I mentioned before like why that happens or how can we make an impact and um reduce inequality but it's it's so it's a lot of concepts around that and why that happens and we study a lot of um

that happens and we study a lot of um you know a lot the structure of the economy why that happens mathematically and so there's there's something called and it keeps coming in different courses. There's something called causal inference which is not exactly correlation if ever you heard of those terms. Yeah. Yeah. Yeah. That's I'm familiar with those terms. Yeah. And so why should we are we trying to we're trying to see let's say and this is a very typical example but um our safety nets let's say in Africa there were so many studies uh if safety nets would actually help people or not with mosquito bites and should they get that or not and things like that. So there are so many things you can do but it's mostly not for actually I think the main

mostly not for actually I think the main purpose is actually having a lot of publications and doing that sort of work. It's called an RCT or randomized control trials. Um, right that uh that people who are up in their career and have been doing this for a while, you need so much funding for all of this. Grad students don't usually get a chance to do that. It's very rare. But I see but um so yeah that is how actually economics differs from other sciences in the sense that you don't really get to do the research or the experiment early on but uh as but I want to say that those core courses are very important if you actually don't do those you cannot do the research moving forward which is why actually you need to complete those before but it's just a different format

before but it's just a different format for econom so help me understand this part you mentioned you have to do a bunch of, you know, what sounded like pretty hardcore research, but you also shared that you would also have to publish a bunch of papers. So, how would one publish papers if they're not able to do the research that you said that you can only do that RCT stuff at a much advanced position. So, that almost feels like a contradiction to me and like not not a contradiction but a paradox. So, how does one work through that? like what what does the happy path here look like that we don't do research early on that's what I'm trying to say research only happens and you need to publish papers but after completing the core courses you'll be doing your uh research

courses you'll be doing your uh research but that's uh and if you're lucky you'll be able to find data and everything goes well and you'll be able to publish a paper but all of that publishing just one paper will take you six years in total around that's the average right as five to six years is to to need a to complete your PhD you need five to six years and you'll be able to only publish one or two papers by then if you're lucky uh and only after so many years have passed and you have so much experience you'll maybe you get a chance to do real stuff like RCTs or something like that but um huh yeah I mean I guess the thing that threw me was I I just thought you would through the course of your PhD you would

through the course of your PhD you would be publishing multip multiple papers. So that that's really helpful context where it's not that way. You even to just publish one it takes like as you said five six years which is yeah for the most part I think that is true unless you are in an exceptionally you know well-ranked school or top 20 or something where maybe the rate is higher. Uh but um yeah for most places that's not true.

um yeah for most places that's not true. That's really helpful context for sure. I mean I for one had you know no idea. So um continuing forward here what exactly does a masters in econometrics entail. I maybe I can lead with uh what I think that would be about and then maybe you can correct me if that work. So what I'm sensing here and also obviously because you said data science.

obviously because you said data science. So what I'm sensing here is you would still be working on some of the more math heavy econom economics applications but now ones that are grounded in maybe big data concepts or you're you're actually handling and parsing big amounts of data. Is that kind of the gist? Feel free to correct me if I messed that up. Yeah. So I would put it like this. I think it's uh when we do the core courses it's a lot of theory and math but when we actually do econometrics we combine real world data with a theory. So we're trying to test hypothesis hypothesis okay trying to see if that is actually true or we're trying to tell people that this is true. So it's basically empirical. empirical in the sense you

empirical. empirical in the sense you observe it, you have data on it and you're trying to test it with the tools that econometrics gives you which is uh which is built on math or stats. Interesting. I see. Can you you mentioned real world data can you just give us an example for us to illustrate or understand what type of data or you know general ballpark we're talking about with this stuff? Yeah, sure. It depends on the I don't think I can answer it very generically but it depends on the context or the question you're trying to answer. I think with econometrics many times we are trying to estimate causal effect and not correlation. So correlations are very easy or at least relatively easy to measure. So, one thing goes up and the other thing goes up and you have an intuition maybe they're correlated and

intuition maybe they're correlated and you check the data for for let's say a year or something and you say maybe they are correlated you know or we're not sure maybe it's something sure we're not sure maybe it's something else is driving it right exactly yeah some hidden variable that we aren't even measuring yes so econometrics gives us the tools to actually measure causality interesting so that is that in the realm of B theorem because that's the only thing I can think of in terms of how I would proceed with that. Yeah, B is something we used to do. Yes, obviously B is used in all of those tools, but there can be instrumental variables.

there can be instrumental variables. There are so many concepts within econometrics that will help you measure uh the the intensity and and uh so basically the coefficient and also the direction of causality. Yeah. M and it'll tell you or at least try to tell you what why if this x thing increases this is the reason um and this is what you should be doing. It's basically used in policy making. So let's say if you want to um if you want to want if you want to um if you want to let's say reduce poverty it's a very general or like how do you reduce poverty there are so many things you could do let's say right? So you're trying to understand what affects poverty you uh Does race affect poverty? Does your education le level affect poverty? So,

education le level affect poverty? So, can you say that if you get maybe 10 years of education, you'll uh probably cross the poverty line? Is that something that is true? We don't know. But that those are the kind of questions um we try to answer in general in economics and econometrics will give us the tools to actually do that. I'm not saying it's always it is it has the tools but it can be useful it cannot be useful because we make a lot of assumptions here right and assumptions on human beings are not always true right the basic assumptions in economics is that uh every person is a rational individual right and that is not true that also brings us to kind of game theory right that isn't that just the whole foundation of game theory where in

whole foundation of game theory where in an ideal world this would happen and both parties would defect or whatever But that's not what happens in the real world. Cuz in the real world, people are motivated by random factors which are close to impossible to measure. Right? That is true. Which is why many times economics fails to actually predict the real world um you know recessions that happen and people are like oh you have all these business models. By the way, we did actually study real world business models which the central bank uses in the US and everywhere around the world. uh but I was just getting started. But they actually use those models to say hey if we increase or decrease the interest rate how is it going to affect everything else right there has to be some mechanism to understand it and economics gives us the

understand it and economics gives us the tools to understand it but with a caveat we're using a lot of assumptions and those assumptions might not always be true actually most of the time can I go and say most of the time they're not true wow well that's concerning it is conserv you know cuz I have heard of I I forget who it's credited to but all models are wrong some are useful I'm sure you you've heard that one as well it's like so it's just such a great line it just encompasses so much but yeah I mean going back to what you were saying that is concerning but I think that is I mean not to go off on too big of a

mean not to go off on too big of a tangent here but personally I've always just been somebody that has been really drawn to economics I remember reading and you will probably laugh um but I remember reading freconomics when I was in like high school. Do you know it by chance? Have you read the book? I have heard about it. A lot of it but I haven't read it to be honest. It's just one of those where that people from the industry or people that are economists absolutely love to just rip into it. They're just like this is absolutely absurd. This is just not what stuff. So I'll give you a quick flavor. I'm trying to think. So essentially they pick random things in the world. So I remember one of the chapters being so they also have a sequel. So it's called super freconomic.

sequel. So it's called super freconomic. So I forget which of the two this was on but they were trying to make the case that global warming is a solved problem. They were trying to make the argument that if humanity wanted we would solve global warming tomorrow. And their big, you know, like big flashy way of doing that was just to inject a bunch of sulfur in the atmosphere, which apparently doesn't cause any harm, and that just kind of reflects back all of the sun rays back to where they come from, thereby solving global warming. So the reason why I'm sharing all of this is what you're talking about kind of feels at least congruent to this type of stuff, but yours feels much more grounded in actual data more so than just quote unquote vibes economics. Does that kind of make sense? Did you ever

that kind of make sense? Did you ever have to deal with stuff like this that is just so much more nebulous than um you know when you actually have data to work with or is that not something that happens? So um as I said like I didn't actually really I was about to start doing research and when you actually start thinking about what you want to answer uh that's the time you actually start looking at data sets and seeing the feasibility of the project because before starting you need to have a full life cycle for it beforehand right and at the time you decide and then results can not make sense but that's okay at least you tried and your approach was right but I didn't space anything like that in academia because everything is pretty fixed at least at least throughout my undergrad my masters even

throughout my undergrad my masters even for the PhD core courses everything is pretty much fixed there is no flexibility flexibility uh flexibility only occurs when you actually start doing research and at that time it's up to you how you frame your question and you have your adviserss and they guide you um they're very helpful u they they'll probably help you understand if or not this sense if the question you're trying to answer doesn't make sense. Do you have the data to answer that question? But yeah, I mean that the book that you the thing that you referenced. Yes. I mean a lot of times the models and the model predictions won't make a lot of sense.

predictions won't make a lot of sense. It's just because we the model is built on assumption. Exactly. Assumpions are not true. So yeah. Yeah. But I still want to say that you need to have a framework to understand how the whole system works. Right. And so you can't start from not knowing anything. So basically then otherwise you'll be going in circles. You need to have some sort of framework, some sort of understanding. And I think that is where economics is very helpful. Amazing. I love that. That actually just puts such a nice ring around the whole thing. Huh.

a nice ring around the whole thing. Huh. And then from there, you know, not to be a purist or anything, but how much overlap is there between just pure data science and economic style data science? Sorry if that question doesn't make sense but no yeah that's perfectly fine. There is a lot of overlap between economics and data science. We study a lot of math and a lot of statistics which is very helpful and very useful uh for data science too. I actually um a couple times in my data science course I repeated classes. Uh the time series was a complete repetition of what I had already done. Um and then I and also statistics. So they had statistics and time series were complete repetition for me. Uh and then machine learning was another elective that I took uh in my uh

another elective that I took uh in my uh data science um masters that also had a lot of concepts which were very similar to you know statistics. Let's that's how we started logistic regression, linear regression and then KN&N all of that trees. Yeah. Yeah. Base but KNN and all things like that uh is something that I did not know before right but it's based it has a mathematical foundation is pretty logical. So you can understand that with the with the tools that economics gave you. It's not too different is what I will say. Uh uh yes it is mathematical but also all the other things are. So it's very similar.

other things are. So it's very similar. I see. It's very similar. Um um yeah. And then I think what I what I did find new though I think a lot of things I got to learn in data science was that that a lot of skills were transferable. Um we both of the courses of math, both of the courses are stats. Uh coding is one thing that we do. Um I used to use data already and I had an elective so I did some Python also already but in data science we coding was um uh it was an important element let me put it like that and uh I I did a lot of um uh

that and uh I I did a lot of um uh programming u that involved Python R uh MATLAB um also visualization tools SQL is something I never did before. Yeah I mean yeah how would that ever come up in in economic circles? Yeah, that part makes sense. Yeah. So, uh that so some things were new but I wouldn't say it's difficult to pick up to it is they have a lot of overlaps. Although like things like SQL are completely new, you can still pick them up if you know one language it is transferable. So that's actually so to take a step back something you said earlier really stuck with me apparently. But you said or you shared how you were trying to learn about some fundamental type questions that you had about the world like why is

that you had about the world like why is there inequality? How do you solve for poverty and stuff like that? Before we proceed, like obviously we have a bunch of ground to cover, but I just want to ask you, were you ever able to get these answers or are these just things that you kind of marinate in the back of your head over the years and kind of just work independently towards or whatever happened to those questions? Yeah. So, yeah, a very good question. I think yes, I'm still interested in understanding those questions and it's not a simple answer at all. A lot of people a lot of answer at all. A lot of people have been working on things like these and I don't think there is one answer to it.

answer to it. Um in my PhD journey actually I pivoted uh more towards I started like that but I pivoted to concepts like household finance and property markets and gentrification and things like that which is aligned but not exactly um answering poverty. So basically what I got interested in was let's say um there's a system of credit scores in the US uh also in India I think all companies have it and so it is a very important factor in determining if you will be given a loan let's say any kind of loan and people need loans you know you need to buy a car you need to buy a house and things like that education loan so how do you actually come up with a number and can that number be so important

important that loans are very important by the way. I feel like if a loan is used for the right reason, it can actually amplify your success a lot. So for me it became critically important to understand why uh or how do people just rely on one number. It is part of being financially inclusive in a society. So that is one aspect of understanding poverty. I was moving towards that aspect because there are so many aspects of poverty. It's not 100% uni dimensional thing, right? So um so yeah for me like I think I um kind of was trying to be more narrower in what where I was going. So that is where my interests lie. I have a curious mind. I keep learning but I don't have an answer

keep learning but I don't have an answer for you right now as sorry I mean just to clarify I I didn't think you would just bring out of your hat the answer to that I I was I was more so you know trying to understand if it feels like such a multi-pronged such a multifaceted problem I personally wouldn't even know where to begin so I guess what I was getting at actually was is this like a you know like a perpetually permanent pet project for you that you kind of go keep going back to when you have some free time as like a I don't know you know passive way to occupy your mind and

know passive way to occupy your mind and such and just work on cuz I mean yeah I don't even know how one would have an answer to that or if there is even an you know so yeah I mean we're aligned there yeah I didn't think you would have I know I know it's just it's it's a it's a thing most a lot of people are working on actually if you try to Google and understand and I'm actually trying to remember but There was one person I was following a lot um not one but multiple people at the time uh who were working on this uh thing of reducing poverty from different angles right and uh when you are actually doing your PhD you tend to study a lot of these things and so

to study a lot of these things and so your mind chooses to uh stick with one or two uh wherever your interests lie and for me it was more on the lines of financial inclusion and why because I really think Getting access to a loan is really important and it can your future right. Yeah. It's it can change the whole generational standing of you know the family cuz once you break out of that cycle you're good then and you can leverage loans to break out of that cycle of poverty. So completely aligned.

cycle of poverty. So completely aligned. Yeah. It's one of the most important tools that we have of breaking that machine for short. So I can see how you were yeah gravitating towards that. Yeah. So financial inclusion is something that I was gravitating towards but it's just one again component of understanding poverty but I think um so Abijit Banerjee was a Nobel Prize winner. I don't know if you've heard of the name and I have heard of the name actually. Yeah. Yeah. Yes. And he works on this uh he works a lot on development economics. Uh and his wife his wife is the reason I actually chose development economics. Oh wow. I am forgetting her name completely right now but uh she she's also a Nobel Prize winner and they

she's also a Nobel Prize winner and they won it together and they have done phenomenal work in this area. They started the concept of using RCTs uh that was basically actually used in science laboratories and not and not in social social sciences. So they popularized it. Um so she she she was a reason I started thinking that I should be doing this. Um one of the very important reasons and one of my professors were actually working with her. Oh wow. That crazy.

working with her. Oh wow. That crazy. Yeah. You were in two degrees of separation from her which is which is pretty ridiculous. Yeah. If you think about it for somebody that's your hero. That's so awesome. Yeah. Yeah. So at the time I was I was like oh wow. glad I had the same reaction and I was like I should definitely you know uh I'm going the right direction and this is all good. So that she was my inspiration. Um she did a lot of work in this area and she continues to do a lot of work in this area. Uh but again you know poverty can be understood from a lot of different angles and I pivoted in my PhD journey. I think a lot of people people do pivot and that is a natural phenomena. Uh because you you you get

phenomena. Uh because you you you get exposed to a lot of new things and a lot of new um um theories that you did not know existed before. I never studied um behavioral economics. Uh I never studied house of finance before. I got a chance to learn about that, study that uh from a from a different angle than that I knew before. Uh I had professors who were teaching that I'd warten before. So it even though I did not directly study behaviorally called my professor was previously teaching that. So it's basically a spillover effect in whatever he was teaching which was household finance. It still had um teachings from behavioral economics too. I see. Very cool. And I promise this is the last tangent before I bring us back on track

tangent before I bring us back on track here. But I'm actually itching to know in all of the study that you've done around you know um empirical study I mean around data and such and uh various tools that were at your disposal around uh just really this field. Is there any mathematical proof for one of capitalism or or socialism/communism to be better than the other? because all the answers I hear around this subject are purely informed of we'll just say not data you know so is there anything that you can share on this particular subject I realize this is way out there but I just had to ask yes so I wouldn't pick sides so people there are different groups they have their own um causes there is proof for both it depends on what kind of uh thinking you

depends on what kind of uh thinking you follow and who you follow there There's no one answer and everybody has their own um pick on it. Interesting. It's at least illuminative to know that there it is you could make the case even using data for both sides. So I feel like that's a pretty good takeaway even from that very short answer. So I appreciate you sharing that. And then earlier when you were talking it kind of sounded like I sensed that and you had also also just you know said it explicitly but it sounded like there was like a gap bit that existed between uh you know academia and industry now that you are in the industry. So what are some ways in which those gaps can be can be filled before it's too late for those people?

before it's too late for those people? Yeah, I think one of the biggest things that I felt moving to industry was uh or like after all of this journey was that people in economics uh they're thrown into uh doing economics uh with if you if you go ahead and do PhD and everything you basically get thrown and you want you they expect you to do economics economic modeling in um any sort of programming language um and you never get a chance to actually learn the fundamentals. It's on you. So that is at a very later stage in life if somebody actually decides to do a PhD. But I think even if you're doing an undergrad or a masters, uh if even if you're at that level, uh right now I would say it is worthwhile to spend some time

is worthwhile to spend some time learning at least one programming language. It can be anything. Uh it can be Python, it can be R, you know. Um and also uh things like using Excel or spreadsheet or whatever I think it is it goes a long way. Uh even though I feel like we are very technical in everything I felt a need where there should have been a compulsory uh course in the fundamentals of computer science. I feel that is very important but that's just my opinion. Some schools have it most of the schools don't even stuff like algorithms and data structures stuff like that. Uh, no, not exactly that. But I mean, so I had to self-e or or I transitioned to

I had to self-e or or I transitioned to data science and I got to learn a lot of those things. But I think sure I'll say data structures. Yes, you I think you should you should understand that but not necessarily algorithms. Uh because some might we might not use them. Uh but I think at least understanding the basics of these things is important. um learning to use a language is really important. So by when I say basics, you should know how to use Python. You should know the uh the libraries that exist in Python. Exactly. You should know u what what economics packages you can use in Python. Let's say that just exposes you to something and it doesn't feel uh so intense at the time when you're actually h you actually have to

you're actually h you actually have to do it. But for most people, it only happens later in the stage when they're in their PhDs. And most people, to be honest, also just get away with STA. They don't actually learn these other things. And STA is only used in the academia world. It's a great tool. Uh but it is only used in the academia. It's not really used in the industry. I think if you're trying to go to the industry, which many people are, it is not going to be very useful because most people in the interviews would want you to know if you know Python or R and it's transferable.

know Python or R and it's transferable. So just spend some time learning it. I think to be honest the schools should provide this and not and not we should we should I I I think I want to say that people should learn to use them but uh if there is somebody who can actually uh who's listening to me and actually has the power to do it every school should have a computer science program in economics within economics at least at the mast's level if not at the undergrad um and yeah I think it is very important to to do that because you're kind of forced to learn it on your own when you are in your PhD, let's say, right? I'm actually really just kind of half blown and half impressed by how quickly

and half impressed by how quickly clearly you were able to pick up all of this stuff because it's funny. Um, I recently had a data scientist on the podcast and well, I'll just say talking to him sounded like so much. like there's just so much and it's just so hard to get into that or you know even break into your first role like he mentioned you need to learn machine learning algorithms you need to do your lead code like you know you need to do your DSA all of that stuff and obviously on the side you have to have your projects and such so with the background that you shared I'm actually very curious to learn a little bit about how you were able to crack your role because it actually sounds very overwhelming to

it actually sounds very overwhelming to me in terms of how much ground you would have had to cover during your masters and I'm referring to the econometrics uh/ data science masters. So what was that experience like for you? Like can you share a little bit around that? Yes, definitely. Yeah, it was overwhelming actually. It was very it was a very busy a year or two when I was actually doing that. Uh it was a lot of different things that I was learning. I was um yes sure they we had a Python class and everything but you have to learn on your own, right? So yeah, I was also doing um a data camp that uh it was helpful for

a data camp that uh it was helpful for me at the time. I had a good bunch of people around me who were able to help me navigate through this new field that I was going into. Although I said that there is overlap and there is overlap but there are a few new things too and it is a transition. It's not exactly the same. same. Um and so yeah with my friend's help and a lot of self-arning I would say with the online resources obviously you're doing a lot of courses too um and it's just a lot of practice uh it's it's a completely different um inter like a podcast topic of how I was able to do it

podcast topic of how I was able to do it but I think in uh like a snapshot of that would be it was a lot of hard work and smart work a lot Um I had a very I was in I was at GSU uh which is a good school in terms of actually giving you one thing which most schools don't and that is uh the internship experience. I applied in hundreds of jobs and then got a few interview calls uh went to the next rounds or levels with a few of them. it narrows down and then you kind of stick with one or two uh and so it was definitely not very easy I would say but is worth it and

easy I would say but is worth it and then if you keep giving interviews you kind of tend to understand what they ask right and where you're lacking right and you can then work on that specific aspect right um and so I think yeah a lot of that and my uh GSU alumni group is very strong so you tend to take advant people should take advantage of that. I think it was very helpful too uh to talk to I by the way uh believe in this concept of talking to people who are two steps ahead of you. So you you're in the right place. Yeah, exactly. You're doing a very nice thing uh uh with the podcast I would say and I personally believe in this that you

personally believe in this that you should talk to people who are two levels above you. you'll kind of get to know a lot of things which you don't know at your level right now. Um, and which is where your alumni networks play a role. And also one thing to add is I think people want to help. If you're not asking if if if your ask doesn't hurt them and if your ask is not too much work, they will help you. So reach out to people with very specific questions and I think yeah people want to help. I love that advice and it's come up throughout the time I've been doing this with really not experts from all walks of life across multiple dimensions fields. Some of them have actually

fields. Some of them have actually nothing to do with academia or industry or studying at all. You know, like I remember an artist sharing this where I think she's something she said something that really struck with me. She was like, um, if you ask somebody for help and if they don't respond, your life is the exact same. Like nothing changes. You just didn't get a response from a stranger. But if they do respond, your life changes forever, you know? So, A, you have nothing to lose here. And B, why are you so afraid of just asking for help when you know that closed mouth don't get pet? So it's literally as simple as make the effort as you said don't waste people's time like don't ask them stuff you could Google obviously that don't expect to get a response to

that don't expect to get a response to that but you're so right in that if you ask specific pointed questions to somebody like what reason would they have of not responding like I love it when people reach out to me with questions sometimes even when they're pretty generic I still I I enjoy answering them like that's just me but love that call out I think that if I mean if you as a listener will take one thing away from this just I feel like that can be that one thing but um to really tie the hatch here um so it does sound like if you if you're an economist it to me it feels like there's a bunch

it to me it feels like there's a bunch of there's a lot of abundance in when it comes to opportunities that come that are around in the you know industry as such. Um firstly is that true and if so what are some of the you know more tailored for uh economists like the roles that are most tailored for economists that exist in the industry right now other than tech or maybe even within tech what would be some of the easy or you know well not easy but low hanging fruit at least that could be applied to by people in economics yeah I would um answer it separately for PhDs and people who don't have a PhD. Perfect. So PhDs have are very technical. Uh they a lot of the

are very technical. Uh they a lot of the economics PhD uh PhD people uh do causal inference. They do a time series forecasting. Most of the people are focused on this. Uh you can answer so many different questions with causal methods. And so uh if you are doing that and you're not a theoretical economist, you can be a theoretical economist too. But people of that sort are rare in general. uh most people are doing applied work and so if you're doing applied work you can easily transition to industry. There are so many companies right now offering uh full-time jobs, internships, research positions to all these different uh through this specific group. Uh if you uh I'll name a few companies but this is just a very small list. There are so many other companies

list. There are so many other companies that are actually looking for people who have done PhD in economics. Amazon is one big employer. They have like a full team of PhD economists. And to be honest, you'll be surprised. They don't care if you know Python. Python. Wow. Okay. Okay. They can they can work with people who know just data. But I think it is only Amazon. But I might be wrong. No, but still even if there is that just one company that you know you can just apply to if you're just good as a PhD that's great. I mean I that's a wonderful call out I think but sorry please please keep going. Yeah they they'll take you uh if you clear their interview levels they'll it doesn't matter to them if you know Python or R

matter to them if you know Python or R just knowing STA is sufficient for them and many economics PhDs just know STA. So that's one thing but it might not work for other companies. Uh but but they're still looking for people like you. So Lyft and Uber and Wayfair, if you go and search at their career websites, they have a full list of jobs for just PhD econ people. Uh but that's for PhD econ because it's very specialized in causal inference and applied methods or time series forecasting. But for other people who are not um PhDs, there's there's still a lot of jobs. I mean I think because it's economics teaches you a lot of transferable skills like I was able to transfer to data science even if you don't exactly transfer in even if you don't study that is why I was actually

don't study that is why I was actually initially mentioning that you should learn a programming language if you have that under your hat and you're applying to jobs in same companies but different roles uh it will not be called economist let's say Amazon calls their role economist but these other places might not call economics but if you go down and look for what they're looking at or what they want from a uh person, they may say MS in economics is preferred.

may say MS in economics is preferred. You know, in um in um Zillow or uh other organizations like that um even in um Wayfair, it's not necessary to have a PhD for some roles, yes, but not for all roles, right? And so there is a lot of opportunity in the industry and the tech world. I know a lot of people who study PhD economics like me initially want to go into policym. Yeah, that's a big one. Agreed. Uh but if at some point you decide to pivot, uh there is a whole new world which you should be aware of. I mean I think it is important to at least be aware of it

important to at least be aware of it that people are looking for your skills and so have an open eye basically and because the industry is looking for you. Um and yeah um yeah though and and one more thing that I want to add is it might be a very good idea to do an internship if you are a PhD student or even if you're doing a master's I mean I think if you can grab an internship position at all of these companies that I just mentioned and these are just a subset of all the companies out there you should be able to land one I mean it's not easy definitely you have to work for it of course but it's there it

work for it of course but it's there it is there and apply for it I mean I I think it um it helps a lot. You get to know how the industry works. You get to know uh that people outside of academia also value your skill sets. So I think one thing that I would want to say is apply to internships. It's useful because in the environment that I was in, nobody actually valued an internship because everybody was very focused on academia, which to be honest is a very it's a bubble of its own and it's very hard u to look outside because you're so busy with all things going on within your own program because it is all the math and exams and tech but you just stretch so thin I bet. Yeah. But yeah, I

stretch so thin I bet. Yeah. But yeah, I mean I think it is worth it to look outside and yeah, spend some time learning programming and applied internship. No, that I think that's wonderful especially with that list you shared. Like often I think it makes such a big difference to be you know kind of trying to look for a needle in a hay stack versus knowing that start here, start with these five. Do go all out with that. If you can get a referral, if you can find somebody that you can network with, get on a coffee chat, whatever the case might be. And then you know if you can concentrate your efforts at one of these I really do agree that

at one of these I really do agree that it makes such a whole huge difference really in terms of um motivation as well because you know that this is a thing that is there for you just need to go out there and grab it. So yeah, I think that's such a wonderful call out and you know just to I guess um conclude things here, I have just always been so drawn towards economics you know just as a concept really more than anything. I won't pretend to understand it. Um but it's something that has been a long-standing interest of mine really for my for most of my formative years and beyond. So I want to thank you for taking the time here today to share your experiences with us. What exactly goes on in the field itself when you're, you

on in the field itself when you're, you know, in the weeds of it trying to go through the gears as one would say with all of it, you know, like again there's just so many takeaways from here. Biggest one for for me personally being that uh it is so technical and so math and statistics heavy and so much uh modeling is involved in that. I would have honestly never guessed that just you know being on the outside but yeah really want to take you really want to thank you for taking the time here today to talk to us Bumika. Thank you Nan I was happy to be here and I appreciate you talking to me. That brings us to the end of that session with Bumika. Thank

end of that session with Bumika. Thank you all for continuing to share these conversations with those that benefit from them. If you would like to support me, the easiest way to do that is by subscribing on YouTube and leaving me up to a fivestar rating on Spotify or any of your favorite podcast apps. Please also share this episodes with your friends and near and dear ones and tell them how you found your favorite new podcast. Catch you all in the next one.

podcast. Catch you all in the next one. New episodes every Wednesday.