6 ms·

My first thought is wondering if it's one of two things: A. The bottom half of PhD Economists are not being trained in the data science/Big Data side of analys

My first thought is wondering if it's one of two things: A. The bottom half of PhD Economists are not being trained in the data science/Big Data side of analysis increasingly needed B. There is less demand for Theory-sided Economists over computationally trained ones

asciident 1y ago
CS and DS people are getting more applied and gaining domain expertise, and can do a lot of economics work now. Academic economists, especially those who primarily do data science / big data, seem to basically be doing Masters-level data science projects for their Ph.D. The hard part in their Ph.D.s is collecting the data, which used to be a very manual job that relied on connections, but more of them are getting them or imputing them from public sources so it's not that impressive anymore. Speaking as someone who has attended 3 economics Ph.D. defenses in the past two years.
- logicalfails 1y ago
  In 2017, I listened to a highly cited Political Economist rant about how "The whole damn field is turning into a bunch of Data Monkeys" Referring to the rise of CS and DS minded economists in the field. His top student was a computer science major...
  - bonsai_spool 1y ago
    > Political Economist rant about how Political economists are explicitly less interested in the quantitative side of economics - which is why they call themselves political economists. Thus, the comment about data makes a lot of sense, and isn't evidence of what is going on with economists
    - logicalfails 1y ago
      His work and department was very quant heavy. I'd say the majority of his students spend most of their time in Python/R cleaning datasets running models
      - bonsai_spool 1y ago
        I'm not disagreeing with you, and I also know political economists from that time who complain that their discipline is changing. It just has very little to do with what this article is discussing.
- 1980phipsi 1y ago
  Data science wasn't even a degree you could get 20 years ago. Twenty years ago if you were interested in what is now called data science, you were getting a degree with some kind of exposure to applied statistics. Economics is one of those disciplines (through econometrics).
  - SJC_Hacker 1y ago
    > Data science wasn't even a degree you could get 20 years ago. It was called statistics
    - mjhay 1y ago
      No. Data science is different than statistics, because it is done on computers. It also uses machine learning algorithms instead of statistical algorithms. These advances, and the shedding of generations of restrictive cruft - frees data scientists to craft answers that their bosses want to hear - proving the superiority of data science over statistics.
      - mathteddybear 1y ago
        yeah, we called that data mining, decision systems, and whatnot... mapreduce was as fresh and hot as the Paul Graham's essays book... folks were using Java over python, due to some open source library from around the globe... essentially, provided you were at a right place in a right time, you could get a BSc in it
        selimthegrim 1y ago
        You might have missed the /s
    - lordnacho 1y ago
      No, I did stats as part of economics around then, and it's nothing like modern DS. It overlaps a fair bit, but in practice the classical stats student is bringing a knife to a gunfight. The practice of working with huge datasets manipulated by computers is valuable enough that you need separate training in it. I don't know what's in a modern stats degree though, I would assume they try to turn it into DS.
narrator 1y ago
The bitter lesson is making it so all these people hand designing features with econometric models are being obsoleted by big models and lots of data.
ActorNightly 1y ago
Its C. the market doesn't follow traditional models anymore. The whole profession was basically centered around putting a dollar amount on risk. For example, lets say I give you a chance of either taking $1k now, or playing a game where you have 1 in 10 chance to win $200k. What would you do? The right answer is "sell" the risk to someone. For example, on the average, if I "buy" the game from 10 people, at a price of $10k each, I can realistically win twice what I spend. Repeat that over x number of steps and more complex games, and that is what the PhDs worked on in terms of pricing. For most of the time it worked ok. In a few instances (most notably the Gaussian Copula that was a large reason for the subprime house market crisis in 2007) it didn't. The problem is that now, its impossible to predict whether orange man is going to throw a hissy fit and cause the market to go up or down, or if large investors are going to artificially prop up stock like they did with Tesla.
- eej71 1y ago
  Its the old axiom best expressed by the philosopher Sir Michael Tyson - "Everyone has a plan until they get punched in the face".
- bigstrat2003 1y ago
  To the extent that a single person can cause economists' predictions to be off, those predictions were never good in the first place. We will always have unstable people in power, it's just human nature. If your predictions can't hold up in the face of that, then you need to refine them.
  - ActorNightly 1y ago
    Look up the concept of computational irreduceability
- mothballed 1y ago
  There was an electric vehicle stock I was watching for awhile (WKHS) hoping for a good time to short. All their reports showed what I believed was an unviable vehicle, there was simply no way to produce it that was anywhere near gross margin positive and they didn't have the money nor access to capital to lose hundreds of millions proving that out. All the economics of the company was going to zero, and they had a ticking time bomb of debt they were about to default on. Shortly before this debt time-bomb went off, Trump magically showed up tweeting in support of the company and alluded to a deal getting pulled off with GM. [] Of course, this ended up being spun off as Lordstown motors, a company that has failed horribly, including Hindenburg Research publishing a video of one of the few trucks they had literally catching on fire on the road while the CEO was simultaneously claiming they had hundreds of millions of dollars in solid orders (later fined by the SEC for that and barred from being an executive of a company for N years). I still don't understand how Trump magically got involved with this penny stock at the 11th hour, but I can tell you I feel something very fishy happened there. https://www.cnbc.com/2019/05/08/trump-tweet-sends-penny-stock-workhorse-surging-by-more-than-50percent-crashes-website.html https://www.cnbc.com/2019/05/08/trump-tweet-sends-penny-stoc...
Eridrus 1y ago
The note about economists and data science in the article felt weird, because data science as a title was invented to get non-CS PhDs to do analyst work because they wanted smarter people doing it. The point of hiring an economics PhD in industry is largely not because they learnt something but because it's a strong and expensive signal.

asciident 1y ago

CS and DS people are getting more applied and gaining domain expertise, and can do a lot of economics work now. Academic economists, especially those who primarily do data science / big data, seem to basically be doing Masters-level data science projects for their Ph.D. The hard part in their Ph.D.s is collecting the data, which used to be a very manual job that relied on connections, but more of them are getting them or imputing them from public sources so it's not that impressive anymore. Speaking as someone who has attended 3 economics Ph.D. defenses in the past two years.

logicalfails 1y ago
In 2017, I listened to a highly cited Political Economist rant about how "The whole damn field is turning into a bunch of Data Monkeys" Referring to the rise of CS and DS minded economists in the field. His top student was a computer science major...
- bonsai_spool 1y ago
  > Political Economist rant about how Political economists are explicitly less interested in the quantitative side of economics - which is why they call themselves political economists. Thus, the comment about data makes a lot of sense, and isn't evidence of what is going on with economists
  - logicalfails 1y ago
    His work and department was very quant heavy. I'd say the majority of his students spend most of their time in Python/R cleaning datasets running models
    - bonsai_spool 1y ago
      I'm not disagreeing with you, and I also know political economists from that time who complain that their discipline is changing. It just has very little to do with what this article is discussing.
1980phipsi 1y ago
Data science wasn't even a degree you could get 20 years ago. Twenty years ago if you were interested in what is now called data science, you were getting a degree with some kind of exposure to applied statistics. Economics is one of those disciplines (through econometrics).
- SJC_Hacker 1y ago
  > Data science wasn't even a degree you could get 20 years ago. It was called statistics
  - mjhay 1y ago
    No. Data science is different than statistics, because it is done on computers. It also uses machine learning algorithms instead of statistical algorithms. These advances, and the shedding of generations of restrictive cruft - frees data scientists to craft answers that their bosses want to hear - proving the superiority of data science over statistics.
    - mathteddybear 1y ago
      yeah, we called that data mining, decision systems, and whatnot... mapreduce was as fresh and hot as the Paul Graham's essays book... folks were using Java over python, due to some open source library from around the globe... essentially, provided you were at a right place in a right time, you could get a BSc in it
      - selimthegrim 1y ago
        You might have missed the /s
  - lordnacho 1y ago
    No, I did stats as part of economics around then, and it's nothing like modern DS. It overlaps a fair bit, but in practice the classical stats student is bringing a knife to a gunfight. The practice of working with huge datasets manipulated by computers is valuable enough that you need separate training in it. I don't know what's in a modern stats degree though, I would assume they try to turn it into DS.

narrator 1y ago

The bitter lesson is making it so all these people hand designing features with econometric models are being obsoleted by big models and lots of data.

ActorNightly 1y ago

Its C. the market doesn't follow traditional models anymore. The whole profession was basically centered around putting a dollar amount on risk. For example, lets say I give you a chance of either taking $1k now, or playing a game where you have 1 in 10 chance to win $200k. What would you do? The right answer is "sell" the risk to someone. For example, on the average, if I "buy" the game from 10 people, at a price of $10k each, I can realistically win twice what I spend. Repeat that over x number of steps and more complex games, and that is what the PhDs worked on in terms of pricing. For most of the time it worked ok. In a few instances (most notably the Gaussian Copula that was a large reason for the subprime house market crisis in 2007) it didn't. The problem is that now, its impossible to predict whether orange man is going to throw a hissy fit and cause the market to go up or down, or if large investors are going to artificially prop up stock like they did with Tesla.

eej71 1y ago
Its the old axiom best expressed by the philosopher Sir Michael Tyson - "Everyone has a plan until they get punched in the face".
bigstrat2003 1y ago
To the extent that a single person can cause economists' predictions to be off, those predictions were never good in the first place. We will always have unstable people in power, it's just human nature. If your predictions can't hold up in the face of that, then you need to refine them.
- ActorNightly 1y ago
  Look up the concept of computational irreduceability
mothballed 1y ago
There was an electric vehicle stock I was watching for awhile (WKHS) hoping for a good time to short. All their reports showed what I believed was an unviable vehicle, there was simply no way to produce it that was anywhere near gross margin positive and they didn't have the money nor access to capital to lose hundreds of millions proving that out. All the economics of the company was going to zero, and they had a ticking time bomb of debt they were about to default on. Shortly before this debt time-bomb went off, Trump magically showed up tweeting in support of the company and alluded to a deal getting pulled off with GM. [] Of course, this ended up being spun off as Lordstown motors, a company that has failed horribly, including Hindenburg Research publishing a video of one of the few trucks they had literally catching on fire on the road while the CEO was simultaneously claiming they had hundreds of millions of dollars in solid orders (later fined by the SEC for that and barred from being an executive of a company for N years). I still don't understand how Trump magically got involved with this penny stock at the 11th hour, but I can tell you I feel something very fishy happened there. https://www.cnbc.com/2019/05/08/trump-tweet-sends-penny-stock-workhorse-surging-by-more-than-50percent-crashes-website.html https://www.cnbc.com/2019/05/08/trump-tweet-sends-penny-stoc...

Eridrus 1y ago

The note about economists and data science in the article felt weird, because data science as a title was invented to get non-CS PhDs to do analyst work because they wanted smarter people doing it. The point of hiring an economics PhD in industry is largely not because they learnt something but because it's a strong and expensive signal.