[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:00.00,0:00:02.55,Default,,0000,0000,0000,,♪ [music] ♪
Dialogue: 0,0:00:03.80,0:00:05.80,Default,,0000,0000,0000,,- [Narrator] Welcome to \NNobel Conversations.
Dialogue: 0,0:00:07.10,0:00:08.10,Default,,0000,0000,0000,,In this episode,
Dialogue: 0,0:00:08.10,0:00:11.57,Default,,0000,0000,0000,,Josh Angrist and Guido Imbens\Nsit down with Isaiah Andrews
Dialogue: 0,0:00:11.57,0:00:14.60,Default,,0000,0000,0000,,to discuss how the field \Nof econometrics is evolving.
Dialogue: 0,0:00:16.10,0:00:18.75,Default,,0000,0000,0000,,- [Isaiah] So Guido and Josh, \Nyou're both pioneers
Dialogue: 0,0:00:18.75,0:00:21.50,Default,,0000,0000,0000,,in developing tools for\Nempirical research in economics.
Dialogue: 0,0:00:21.50,0:00:22.93,Default,,0000,0000,0000,,And so I'd like to explore
Dialogue: 0,0:00:22.93,0:00:25.30,Default,,0000,0000,0000,,sort of where you feel like \Nthe field is heading,
Dialogue: 0,0:00:25.30,0:00:28.08,Default,,0000,0000,0000,,sort of economics, econometrics,\Nthe whole thing.
Dialogue: 0,0:00:28.51,0:00:31.29,Default,,0000,0000,0000,,To start, I'd be interested to hear
Dialogue: 0,0:00:32.20,0:00:35.20,Default,,0000,0000,0000,,about whether you feel like\Nsort of the way in which
Dialogue: 0,0:00:35.20,0:00:38.51,Default,,0000,0000,0000,,the local average treatment \Neffects framework sort of took hold
Dialogue: 0,0:00:38.80,0:00:42.10,Default,,0000,0000,0000,,has any lessons for how \Nnew empirical methods in economics
Dialogue: 0,0:00:42.10,0:00:44.30,Default,,0000,0000,0000,,develop and spread\Nor how they should.
Dialogue: 0,0:00:44.56,0:00:45.96,Default,,0000,0000,0000,,- [Josh] That's a good question.
Dialogue: 0,0:00:46.61,0:00:47.79,Default,,0000,0000,0000,,You go first.
Dialogue: 0,0:00:47.79,0:00:49.46,Default,,0000,0000,0000,,(laughter)
Dialogue: 0,0:00:49.70,0:00:53.18,Default,,0000,0000,0000,,Yeah, so I think \Nthe important thing
Dialogue: 0,0:00:53.18,0:00:58.55,Default,,0000,0000,0000,,is to come up \Nwith good convincing cases
Dialogue: 0,0:00:58.55,0:01:02.21,Default,,0000,0000,0000,,where the questions are clear
Dialogue: 0,0:01:02.40,0:01:05.72,Default,,0000,0000,0000,,and where kind of the methods\Napply in general.
Dialogue: 0,0:01:05.72,0:01:07.56,Default,,0000,0000,0000,,So one thing I--
Dialogue: 0,0:01:08.07,0:01:12.00,Default,,0000,0000,0000,,Kind of looking back\Nat the subsequent literature,
Dialogue: 0,0:01:12.20,0:01:16.70,Default,,0000,0000,0000,,so I really like the regression\Ndiscontinuity literature
Dialogue: 0,0:01:16.70,0:01:19.67,Default,,0000,0000,0000,,[where there were] clearly a bunch\Nof really convincing examples
Dialogue: 0,0:01:19.67,0:01:21.32,Default,,0000,0000,0000,,and that allowed people to kind of
Dialogue: 0,0:01:22.30,0:01:27.20,Default,,0000,0000,0000,,think more clearly, look harder at\Nthe methodological questions.
Dialogue: 0,0:01:27.40,0:01:28.80,Default,,0000,0000,0000,,Kind of do clear applications
Dialogue: 0,0:01:28.80,0:01:30.60,Default,,0000,0000,0000,,that then allow you \Nto kind of think about,
Dialogue: 0,0:01:30.60,0:01:33.60,Default,,0000,0000,0000,,"Wow, do this type of assumption\Nseem reasonable here?
Dialogue: 0,0:01:33.60,0:01:38.00,Default,,0000,0000,0000,,What kind of things do we not like\Nin the early papers?
Dialogue: 0,0:01:38.50,0:01:39.80,Default,,0000,0000,0000,,How can we improve things?"
Dialogue: 0,0:01:39.80,0:01:42.30,Default,,0000,0000,0000,,So having clear applications\Nmotivating,
Dialogue: 0,0:01:43.40,0:01:46.40,Default,,0000,0000,0000,,these literatures, \NI think it's very helpful.
Dialogue: 0,0:01:46.80,0:01:49.30,Default,,0000,0000,0000,,I'm glad you mentioned\Nthe regression discontinuity, Guido.
Dialogue: 0,0:01:49.30,0:01:53.30,Default,,0000,0000,0000,,I think there's a lot of \Ncomplementarity between IV and RD,
Dialogue: 0,0:01:54.70,0:01:57.06,Default,,0000,0000,0000,,Instrumental Variables and\NRegression Discontinuity.
Dialogue: 0,0:02:00.35,0:02:03.26,Default,,0000,0000,0000,,And a lot of \Nthe econometric applications
Dialogue: 0,0:02:03.26,0:02:04.52,Default,,0000,0000,0000,,of regression discontinuity
Dialogue: 0,0:02:04.52,0:02:07.23,Default,,0000,0000,0000,,are what used to be called \N"fuzzy" RD,
Dialogue: 0,0:02:07.23,0:02:11.62,Default,,0000,0000,0000,,where, you know, it's not discrete \Nor deterministic at the cutoff,
Dialogue: 0,0:02:11.62,0:02:14.90,Default,,0000,0000,0000,,but just the change\Nin rates or intensity.
Dialogue: 0,0:02:14.90,0:02:18.74,Default,,0000,0000,0000,,And and the late framework helps us\Nunderstand those applications
Dialogue: 0,0:02:18.74,0:02:21.14,Default,,0000,0000,0000,,and gives us a clear interpretation
Dialogue: 0,0:02:21.14,0:02:25.00,Default,,0000,0000,0000,,for say, something like, \Nin my paper with Victor Lavy,
Dialogue: 0,0:02:25.00,0:02:28.10,Default,,0000,0000,0000,,where we use Maimonides'\Nrule, the class size cutoffs.
Dialogue: 0,0:02:28.60,0:02:30.20,Default,,0000,0000,0000,,What are you getting there?
Dialogue: 0,0:02:30.40,0:02:31.97,Default,,0000,0000,0000,,Of course, you can\Nanswer that question
Dialogue: 0,0:02:31.97,0:02:33.90,Default,,0000,0000,0000,,with a linear \Nconstant effects model,
Dialogue: 0,0:02:34.20,0:02:36.31,Default,,0000,0000,0000,,but it turns out \Nwe're not limited to that,
Dialogue: 0,0:02:36.31,0:02:39.89,Default,,0000,0000,0000,,and an RD is still very powerful \Nand illuminating,
Dialogue: 0,0:02:40.63,0:02:42.10,Default,,0000,0000,0000,,even when, you know,
Dialogue: 0,0:02:42.10,0:02:46.00,Default,,0000,0000,0000,,the correlation between the cutoff\Nand the variable of interest,
Dialogue: 0,0:02:46.00,0:02:48.98,Default,,0000,0000,0000,,in this case class size, \Nis partial,
Dialogue: 0,0:02:48.98,0:02:51.00,Default,,0000,0000,0000,,maybe even not that strong.
Dialogue: 0,0:02:52.00,0:02:54.100,Default,,0000,0000,0000,,So there was definitely a kind of,\Na parallel development.
Dialogue: 0,0:02:54.100,0:02:56.40,Default,,0000,0000,0000,,It's also interesting,
Dialogue: 0,0:02:56.60,0:02:59.78,Default,,0000,0000,0000,,you know, nobody talked about\Nregression discontinuity designs
Dialogue: 0,0:02:59.78,0:03:01.22,Default,,0000,0000,0000,,when we were in graduate school,
Dialogue: 0,0:03:01.22,0:03:05.30,Default,,0000,0000,0000,,it was something that other social\Nscientists were interested in,
Dialogue: 0,0:03:05.80,0:03:09.51,Default,,0000,0000,0000,,and that kind of grew up \Nalongside the late framework
Dialogue: 0,0:03:09.51,0:03:14.79,Default,,0000,0000,0000,,and we've both done work on \Nboth applications and methods there
Dialogue: 0,0:03:14.79,0:03:18.38,Default,,0000,0000,0000,,and it's been very exciting \Nto see that kind of develop
Dialogue: 0,0:03:18.38,0:03:19.80,Default,,0000,0000,0000,,and become so important.
Dialogue: 0,0:03:20.00,0:03:23.70,Default,,0000,0000,0000,,It's part of a general evolution,\NI think, towards, you know,
Dialogue: 0,0:03:24.00,0:03:27.70,Default,,0000,0000,0000,,credible identification strategies\Ncausal effects...
Dialogue: 0,0:03:28.63,0:03:30.20,Default,,0000,0000,0000,,less, you know, making econometrics
Dialogue: 0,0:03:30.40,0:03:33.30,Default,,0000,0000,0000,,more about causal questions\Nthan about models.
Dialogue: 0,0:03:33.70,0:03:38.70,Default,,0000,0000,0000,,In terms of the future, I think one thing\Nthat LATE has helped facilitate
Dialogue: 0,0:03:38.70,0:03:42.70,Default,,0000,0000,0000,,is a move towards more creative\Nrandomized, trials, where,
Dialogue: 0,0:03:42.80,0:03:44.40,Default,,0000,0000,0000,,you know, there's something of interest,
Dialogue: 0,0:03:45.50,0:03:50.70,Default,,0000,0000,0000,,it's not possible or straightforward\Nto Simply turn it off or on
Dialogue: 0,0:03:51.00,0:03:52.90,Default,,0000,0000,0000,,but you can encourage it
Dialogue: 0,0:03:53.20,0:03:58.20,Default,,0000,0000,0000,,or discourage it. So, you subsidize\Nschooling with financial aid, for example,
Dialogue: 0,0:03:59.00,0:04:00.20,Default,,0000,0000,0000,,so now we have a whole
Dialogue: 0,0:04:00.30,0:04:02.30,Default,,0000,0000,0000,,Framework for interpreting that.
Dialogue: 0,0:04:02.90,0:04:03.50,Default,,0000,0000,0000,,And,
Dialogue: 0,0:04:03.60,0:04:06.90,Default,,0000,0000,0000,,and it kind of opens\Nthe doors to randomized,
Dialogue: 0,0:04:06.90,0:04:09.40,Default,,0000,0000,0000,,Trials of things that that maybe would,
Dialogue: 0,0:04:09.70,0:04:10.20,Default,,0000,0000,0000,,you know,
Dialogue: 0,0:04:10.30,0:04:13.80,Default,,0000,0000,0000,,not have seen possible before we've,
Dialogue: 0,0:04:14.50,0:04:19.70,Default,,0000,0000,0000,,we've used that a lot in the work. We do\Non schools in our in the blueprint Lab
Dialogue: 0,0:04:20.00,0:04:26.60,Default,,0000,0000,0000,,at MIT were exploiting random assignment\Nand in very creative ways, I think.
Dialogue: 0,0:04:28.10,0:04:32.30,Default,,0000,0000,0000,,Related to that. Do you see sort\Nof particular factors that make for
Dialogue: 0,0:04:32.40,0:04:34.30,Default,,0000,0000,0000,,useful research and econometrics.
Dialogue: 0,0:04:34.40,0:04:37.40,Default,,0000,0000,0000,,You've alluded to it?
Dialogue: 0,0:04:37.40,0:04:40.30,Default,,0000,0000,0000,,Having a clear connection to\Nproblems that are actually coming up.
Dialogue: 0,0:04:40.30,0:04:45.00,Default,,0000,0000,0000,,And empirical practice is often a good\Nidea. I'll send it. Always a good idea.
Dialogue: 0,0:04:45.70,0:04:50.10,Default,,0000,0000,0000,,I often find myself sitting in an\Neconomy metrics Theory, seminar.
Dialogue: 0,0:04:50.70,0:04:52.50,Default,,0000,0000,0000,,Say the Harvard MIT seminar
Dialogue: 0,0:04:53.40,0:04:57.00,Default,,0000,0000,0000,,and I'm thinking what problem is\Nthis guy solving who has this?
Dialogue: 0,0:04:57.10,0:04:59.80,Default,,0000,0000,0000,,This problem and you know,
Dialogue: 0,0:05:01.60,0:05:04.70,Default,,0000,0000,0000,,sometimes there's an\Nembarrassing silence if I ask
Dialogue: 0,0:05:04.90,0:05:08.30,Default,,0000,0000,0000,,or there might be a\Nfairly contrived scenario.
Dialogue: 0,0:05:08.80,0:05:11.60,Default,,0000,0000,0000,,I want to see where the tool is useful.
Dialogue: 0,0:05:12.50,0:05:14.90,Default,,0000,0000,0000,,There are some purely foundational tools.
Dialogue: 0,0:05:14.90,0:05:17.60,Default,,0000,0000,0000,,I do take the point, you\Nknow, there are people who are
Dialogue: 0,0:05:18.20,0:05:22.50,Default,,0000,0000,0000,,working on conceptual\Nfoundations of you know,
Dialogue: 0,0:05:22.60,0:05:25.30,Default,,0000,0000,0000,,it's more becomes more like\Nmathematical statistics.
Dialogue: 0,0:05:25.80,0:05:27.00,Default,,0000,0000,0000,,I mean, I remember an early example
Dialogue: 0,0:05:27.10,0:05:28.10,Default,,0000,0000,0000,,I believe that that I,
Dialogue: 0,0:05:28.20,0:05:32.50,Default,,0000,0000,0000,,you know, I struggled to understand was\Nthe idea of stochastic Equity continuity,
Dialogue: 0,0:05:32.50,0:05:36.90,Default,,0000,0000,0000,,which my one of my thesis advisors Whitney\Nknew he was using to great effect and
Dialogue: 0,0:05:37.50,0:05:39.90,Default,,0000,0000,0000,,I was trying to understand\Nthat and there isn't really.
Dialogue: 0,0:05:40.60,0:05:45.20,Default,,0000,0000,0000,,It's really foundational. It's not\Nbut an application that's driving that
Dialogue: 0,0:05:46.10,0:05:47.30,Default,,0000,0000,0000,,at least not immediately
Dialogue: 0,0:05:48.60,0:05:53.20,Default,,0000,0000,0000,,but but most things are not like that\Nand so there should be a problem.
Dialogue: 0,0:05:53.80,0:05:56.90,Default,,0000,0000,0000,,And the I think it's on the it's on
Dialogue: 0,0:05:57.10,0:05:59.60,Default,,0000,0000,0000,,On the, the seller of that sort of thing,
Dialogue: 0,0:06:00.10,0:06:04.40,Default,,0000,0000,0000,,you know, because there's opportunity\Ncost the time and attention and effort
Dialogue: 0,0:06:04.40,0:06:07.20,Default,,0000,0000,0000,,to understand things to, you\Nknow, it's on the seller to say.
Dialogue: 0,0:06:07.40,0:06:08.90,Default,,0000,0000,0000,,Hey, I'm solving this problem
Dialogue: 0,0:06:09.40,0:06:12.90,Default,,0000,0000,0000,,and and here's a set of results\Nthat show that it's useful.
Dialogue: 0,0:06:12.90,0:06:15.20,Default,,0000,0000,0000,,And here's some insight that I get.
Dialogue: 0,0:06:16.20,0:06:17.20,Default,,0000,0000,0000,,As you said, Josh, great,
Dialogue: 0,0:06:17.20,0:06:20.70,Default,,0000,0000,0000,,sort of there's been a move in the\Ndirection of thinking more about causality
Dialogue: 0,0:06:20.70,0:06:22.80,Default,,0000,0000,0000,,in economics and empirical\Nwork in economics,
Dialogue: 0,0:06:22.90,0:06:24.80,Default,,0000,0000,0000,,any consequences of sort of the Wilds,
Dialogue: 0,0:06:24.80,0:06:27.00,Default,,0000,0000,0000,,the spread of that view. That\Nsurprised you or anything.
Dialogue: 0,0:06:27.10,0:06:31.40,Default,,0000,0000,0000,,UB was downsides of sort of the way that\Nhe could empirical, economics has gone
Dialogue: 0,0:06:31.50,0:06:32.10,Default,,0000,0000,0000,,sometimes.
Dialogue: 0,0:06:32.10,0:06:38.50,Default,,0000,0000,0000,,I see somebody does Ivy and they get a\Nresult which seems implausibly large.
Dialogue: 0,0:06:38.80,0:06:40.20,Default,,0000,0000,0000,,That's the usual case.
Dialogue: 0,0:06:42.50,0:06:43.90,Default,,0000,0000,0000,,So it might be, you know,
Dialogue: 0,0:06:43.90,0:06:48.60,Default,,0000,0000,0000,,an extraordinarily large causal effect\Nof some relatively minor Intervention,
Dialogue: 0,0:06:49.10,0:06:52.80,Default,,0000,0000,0000,,which was randomized or for\Nwhich you could make a case that
Dialogue: 0,0:06:52.90,0:06:56.90,Default,,0000,0000,0000,,that there's a good design.\NAnd then when I see that,
Dialogue: 0,0:06:57.10,0:06:57.60,Default,,0000,0000,0000,,That and,
Dialogue: 0,0:06:57.60,0:06:58.50,Default,,0000,0000,0000,,you know, I think,
Dialogue: 0,0:06:58.50,0:06:58.90,Default,,0000,0000,0000,,you know,
Dialogue: 0,0:06:58.90,0:07:02.60,Default,,0000,0000,0000,,it's very hard for me to believe that this\Nrelatively minor intervention has such
Dialogue: 0,0:07:02.60,0:07:04.00,Default,,0000,0000,0000,,a large defect,
Dialogue: 0,0:07:04.10,0:07:07.40,Default,,0000,0000,0000,,the author. Well, sometimes\Nresort to the local average,
Dialogue: 0,0:07:07.40,0:07:09.10,Default,,0000,0000,0000,,treatment effects theorem and say,
Dialogue: 0,0:07:09.40,0:07:12.70,Default,,0000,0000,0000,,wow, these compliers, you know,\Nthey're special in some way.
Dialogue: 0,0:07:13.30,0:07:17.60,Default,,0000,0000,0000,,And, you know, they just benefit\Nextraordinarily from this intervention
Dialogue: 0,0:07:18.10,0:07:22.10,Default,,0000,0000,0000,,and I'm reluctant to take that\Nat face value. I think, you know,
Dialogue: 0,0:07:22.10,0:07:24.10,Default,,0000,0000,0000,,often when effects are too big,
Dialogue: 0,0:07:24.30,0:07:26.90,Default,,0000,0000,0000,,it's because the exclusion\Nrestriction is failing. So
Dialogue: 0,0:07:27.10,0:07:31.70,Default,,0000,0000,0000,,Don't really have the right endogenous\Nvariable to scale that result.
Dialogue: 0,0:07:32.00,0:07:35.70,Default,,0000,0000,0000,,And so I'm not too happy to see
Dialogue: 0,0:07:35.80,0:07:38.80,Default,,0000,0000,0000,,you know, just sort of\Na generic heterogeneity
Dialogue: 0,0:07:38.90,0:07:43.80,Default,,0000,0000,0000,,argument being used to excuse something\Nthat I think might be a deeper problem.
Dialogue: 0,0:07:45.30,0:07:47.40,Default,,0000,0000,0000,,I think it played somewhat\Nof an unfortunate roll pin.
Dialogue: 0,0:07:47.40,0:07:52.30,Default,,0000,0000,0000,,The discussions kind of between reduced\Nform and structural approaches where
Dialogue: 0,0:07:52.60,0:07:54.20,Default,,0000,0000,0000,,I feel that wasn't quite
Dialogue: 0,0:07:55.00,0:07:59.30,Default,,0000,0000,0000,,right. The instrumental\Nvariables assumptions are
Dialogue: 0,0:08:00.40,0:08:05.20,Default,,0000,0000,0000,,at the core structural assumptions about\NBehavior. They were coming from economic
Dialogue: 0,0:08:07.10,0:08:09.90,Default,,0000,0000,0000,,thinking about the economic\Nbehavior of agents,
Dialogue: 0,0:08:10.30,0:08:15.00,Default,,0000,0000,0000,,and it's somehow it got\Npushed in a Direction.
Dialogue: 0,0:08:15.10,0:08:17.60,Default,,0000,0000,0000,,That I think wasn't\Nreally very helpful. If
Dialogue: 0,0:08:18.80,0:08:21.70,Default,,0000,0000,0000,,the way I think, initially the
Dialogue: 0,0:08:22.80,0:08:27.30,Default,,0000,0000,0000,,we wrote things up. It was it was describing\Nwhat was happening, there was set of
Dialogue: 0,0:08:27.50,0:08:32.20,Default,,0000,0000,0000,,methods. People were using be\Nclarified what those methods were doing
Dialogue: 0,0:08:32.90,0:08:38.50,Default,,0000,0000,0000,,and in a way that I think\Ncontain a fair amount of insight,
Dialogue: 0,0:08:39.10,0:08:45.00,Default,,0000,0000,0000,,but it somehow it got pushed into a corner\Nthat I think was not necessarily very
Dialogue: 0,0:08:45.10,0:08:48.60,Default,,0000,0000,0000,,or even just the language of\Nreduced form versus structural.
Dialogue: 0,0:08:48.60,0:08:51.10,Default,,0000,0000,0000,,I find kind of funny in\Nthe sense that the right
Dialogue: 0,0:08:51.10,0:08:53.10,Default,,0000,0000,0000,,the local average treatment\Neffect model, right?
Dialogue: 0,0:08:53.10,0:08:55.30,Default,,0000,0000,0000,,The potential outcomes\Nmodel is a nonparametric.
Dialogue: 0,0:08:55.30,0:08:56.20,Default,,0000,0000,0000,,Structural model,
Dialogue: 0,0:08:56.20,0:08:58.60,Default,,0000,0000,0000,,if you want to think about it, as\Nyou sort of suggested, he does.
Dialogue: 0,0:08:58.60,0:08:59.40,Default,,0000,0000,0000,,So, there's something,
Dialogue: 0,0:09:00.10,0:09:03.70,Default,,0000,0000,0000,,there's something a little funny about\Nputting these two things in a position when
Dialogue: 0,0:09:03.80,0:09:06.60,Default,,0000,0000,0000,,yes, well, that language, of\Ncourse, comes from the area, the
Dialogue: 0,0:09:06.90,0:09:09.80,Default,,0000,0000,0000,,70s equations framework that we inherited.
Dialogue: 0,0:09:10.40,0:09:12.40,Default,,0000,0000,0000,,It has the advantage that people seem
Dialogue: 0,0:09:12.40,0:09:15.00,Default,,0000,0000,0000,,to know what you mean\Nwhen you use it, but might
Dialogue: 0,0:09:15.10,0:09:18.20,Default,,0000,0000,0000,,That people are hearing different. Different\Npeople are hearing different things.
Dialogue: 0,0:09:18.30,0:09:20.90,Default,,0000,0000,0000,,Yeah. I think I think veggies\NFarmers had become use
Dialogue: 0,0:09:20.90,0:09:23.20,Default,,0000,0000,0000,,a little bit of the\Npejoratives. Okay? Yeah.
Dialogue: 0,0:09:23.30,0:09:28.30,Default,,0000,0000,0000,,The word, which is not really quite\Nwhat it was originally intended for.
Dialogue: 0,0:09:30.10,0:09:34.10,Default,,0000,0000,0000,,I guess something else that strikes\Nme in thinking about the effects of
Dialogue: 0,0:09:34.10,0:09:38.20,Default,,0000,0000,0000,,the local average treatment effect\Nframework is that often folks will appeal to
Dialogue: 0,0:09:38.20,0:09:41.80,Default,,0000,0000,0000,,a local average, treatment effects\Nintuition for settings. Well, beyond
Dialogue: 0,0:09:42.00,0:09:44.90,Default,,0000,0000,0000,,ones, where any sort of formal\Nresults has actually been
Dialogue: 0,0:09:45.00,0:09:49.70,Default,,0000,0000,0000,,Shhhhht. And I'm curious given\Nall the work that you guys did to,
Dialogue: 0,0:09:49.90,0:09:53.20,Default,,0000,0000,0000,,you know, establish late results in\Ndifferent in different settings. I'm curious
Dialogue: 0,0:09:53.30,0:09:57.80,Default,,0000,0000,0000,,any thoughts on that. I think there's\Ngoing to be a lot of cases where
Dialogue: 0,0:09:57.90,0:10:02.20,Default,,0000,0000,0000,,the intuition does get.\NYou get you some distance,
Dialogue: 0,0:10:02.80,0:10:07.60,Default,,0000,0000,0000,,but it's going to be somewhat limited\Nand establishing formal results. There
Dialogue: 0,0:10:08.40,0:10:12.70,Default,,0000,0000,0000,,may be a little tricky and there may\Nbe only work in special circumstances,
Dialogue: 0,0:10:13.10,0:10:13.50,Default,,0000,0000,0000,,you need.
Dialogue: 0,0:10:14.60,0:10:19.50,Default,,0000,0000,0000,,And you end up with a lot of formality\Nthat may not quite capture the intuition
Dialogue: 0,0:10:19.90,0:10:23.20,Default,,0000,0000,0000,,sometimes I'm somewhat uneasy with them\Nand they are not necessarily the papers.
Dialogue: 0,0:10:23.20,0:10:25.00,Default,,0000,0000,0000,,I would want to ride that the
Dialogue: 0,0:10:25.10,0:10:30.00,Default,,0000,0000,0000,,but I do think something do intuition\Norphaned US capture part of the
Dialogue: 0,0:10:30.20,0:10:31.10,Default,,0000,0000,0000,,of the problem.
Dialogue: 0,0:10:33.10,0:10:36.30,Default,,0000,0000,0000,,I think, in some sense we were\Nkind of very fortunate there
Dialogue: 0,0:10:36.90,0:10:40.50,Default,,0000,0000,0000,,in the way. The late paper go handle. It.\NDon't know if that, actually the editor,
Dialogue: 0,0:10:40.60,0:10:41.70,Default,,0000,0000,0000,,made it much shorter
Dialogue: 0,0:10:42.10,0:10:46.30,Default,,0000,0000,0000,,and that then allowed us to kind of\Nfocus on very clear, crisp results
Dialogue: 0,0:10:47.10,0:10:49.80,Default,,0000,0000,0000,,where if, you know, this,
Dialogue: 0,0:10:50.00,0:10:54.20,Default,,0000,0000,0000,,this is somewhat unfortunate tendency in\Nthe commercialization of having the papers.
Dialogue: 0,0:10:54.60,0:10:58.80,Default,,0000,0000,0000,,Well, you should be able to fix that, man.\NI'm trying to take some time to fix that.
Dialogue: 0,0:10:59.40,0:11:02.70,Default,,0000,0000,0000,,I think this is an example where it's sort\Nof very clear that having it. Be sure.
Dialogue: 0,0:11:02.90,0:11:08.00,Default,,0000,0000,0000,,It's actually impose that no paper can\Nbe longer than the late paper that wow.
Dialogue: 0,0:11:08.80,0:11:14.30,Default,,0000,0000,0000,,Great. At least no Theory. No Theory Pig.\NYeah, and I think, I think they're well,
Dialogue: 0,0:11:14.50,0:11:16.80,Default,,0000,0000,0000,,I'm trying very hard to get\Nthe papers to be shorter.
Dialogue: 0,0:11:16.80,0:11:18.70,Default,,0000,0000,0000,,And I think there's a lot of value
Dialogue: 0,0:11:19.20,0:11:22.60,Default,,0000,0000,0000,,today because it's often the second\Npart of the paper that doesn't actually
Dialogue: 0,0:11:23.70,0:11:26.50,Default,,0000,0000,0000,,Get you much further\Nand understanding things
Dialogue: 0,0:11:27.00,0:11:31.70,Default,,0000,0000,0000,,but and it does make things much\Nharder to read and, you know,
Dialogue: 0,0:11:32.40,0:11:33.70,Default,,0000,0000,0000,,it sort of goes back to
Dialogue: 0,0:11:34.20,0:11:38.50,Default,,0000,0000,0000,,how I think he kind of a trick should\Nbe done to you should focus on the see.
Dialogue: 0,0:11:38.70,0:11:41.30,Default,,0000,0000,0000,,It should be reasonably\Nclose to empirical problems.
Dialogue: 0,0:11:41.50,0:11:43.90,Default,,0000,0000,0000,,They should be very clear problems.
Dialogue: 0,0:11:44.80,0:11:48.90,Default,,0000,0000,0000,,But then often the the theory\Ndoesn't need to be quite so long.
Dialogue: 0,0:11:49.00,0:11:49.30,Default,,0000,0000,0000,,Yeah,
Dialogue: 0,0:11:51.10,0:11:53.40,Default,,0000,0000,0000,,I think they had things have
Dialogue: 0,0:11:53.60,0:11:54.70,Default,,0000,0000,0000,,On a little off track.
Dialogue: 0,0:11:56.40,0:11:58.40,Default,,0000,0000,0000,,The relatively recent change has been a
Dialogue: 0,0:11:58.50,0:12:02.20,Default,,0000,0000,0000,,seeming big increase in demand for\Npeople with sort of econometrics.
Dialogue: 0,0:12:02.20,0:12:04.80,Default,,0000,0000,0000,,Causal effect, estimation\Nskills in the tech sector.
Dialogue: 0,0:12:05.00,0:12:09.00,Default,,0000,0000,0000,,I'm interested either of you have\Nthoughts on sort of how that's gonna
Dialogue: 0,0:12:09.20,0:12:11.60,Default,,0000,0000,0000,,how that's going to interact with\Nthe development of empirical methods,
Dialogue: 0,0:12:11.60,0:12:14.30,Default,,0000,0000,0000,,or Empirical research, and\Neconomics. Going forward, sort of
Dialogue: 0,0:12:14.60,0:12:21.00,Default,,0000,0000,0000,,whether sort of a meta point, which\Nis there's this new kind of employer
Dialogue: 0,0:12:21.80,0:12:26.00,Default,,0000,0000,0000,,the Amazons and the Uber and, you know,
Dialogue: 0,0:12:26.20,0:12:27.60,Default,,0000,0000,0000,,Riser world
Dialogue: 0,0:12:28.00,0:12:29.30,Default,,0000,0000,0000,,and I think that's great.
Dialogue: 0,0:12:29.30,0:12:33.20,Default,,0000,0000,0000,,And I'd like to tell my students about\Nthat, you know, especially at MIT.
Dialogue: 0,0:12:33.20,0:12:37.00,Default,,0000,0000,0000,,We have a lot of computer science\NMajors. That's our biggest major
Dialogue: 0,0:12:37.40,0:12:42.80,Default,,0000,0000,0000,,and I try to seduce some of those folks\Ninto economics by saying, you know,
Dialogue: 0,0:12:43.20,0:12:45.70,Default,,0000,0000,0000,,you can go work for these,
Dialogue: 0,0:12:45.80,0:12:48.40,Default,,0000,0000,0000,,you know companies that\Npeople are very keen to
Dialogue: 0,0:12:48.70,0:12:50.80,Default,,0000,0000,0000,,work for because the work seems exciting,
Dialogue: 0,0:12:52.00,0:12:56.00,Default,,0000,0000,0000,,you know that the skills that you get in\Neconometrics are are as good or better.
Dialogue: 0,0:12:56.10,0:13:01.10,Default,,0000,0000,0000,,Better than than any competing discipline\Nhas to offer. So you should at least
Dialogue: 0,0:13:01.40,0:13:04.20,Default,,0000,0000,0000,,take some econ, take some\Neconometrics. And some econ.
Dialogue: 0,0:13:04.80,0:13:07.00,Default,,0000,0000,0000,,I did a fun project with a uber
Dialogue: 0,0:13:07.60,0:13:12.90,Default,,0000,0000,0000,,on labor supply of Uber drivers and was\Nvery, very exciting to be part of that.
Dialogue: 0,0:13:13.10,0:13:15.40,Default,,0000,0000,0000,,Plus. I got to drive for Uber for a while
Dialogue: 0,0:13:15.90,0:13:20.70,Default,,0000,0000,0000,,and I thought that was fun tonight. I did\Nnot make enough that I was attempted to
Dialogue: 0,0:13:21.10,0:13:25.10,Default,,0000,0000,0000,,give up by a mighty job, but\NI enjoyed the experience.
Dialogue: 0,0:13:25.30,0:13:26.00,Default,,0000,0000,0000,,I see a
Dialogue: 0,0:13:26.20,0:13:30.90,Default,,0000,0000,0000,,Cho challenge to our model\Nof graduate education here,
Dialogue: 0,0:13:31.70,0:13:37.40,Default,,0000,0000,0000,,which is if we're trading people\Nto go work at Amazon, you know,
Dialogue: 0,0:13:37.90,0:13:42.90,Default,,0000,0000,0000,,it's not clear. Why? You know, we should\Nbe paying graduate stipends for that.
Dialogue: 0,0:13:43.20,0:13:45.40,Default,,0000,0000,0000,,Why should the taxpayer effectively
Dialogue: 0,0:13:46.10,0:13:51.40,Default,,0000,0000,0000,,be subsidizing? That our graduate education\Nin the u.s. Is generously subsidized?
Dialogue: 0,0:13:51.40,0:13:56.00,Default,,0000,0000,0000,,Even in private universities. It's\Nultimately there's a lot of public money.
Dialogue: 0,0:13:56.10,0:13:59.30,Default,,0000,0000,0000,,Me there. And I think the\Ntraditional rationale for that is,
Dialogue: 0,0:13:59.50,0:14:03.90,Default,,0000,0000,0000,,you know, we were training, Educators and\NScholars, and there's a great externality
Dialogue: 0,0:14:04.30,0:14:05.70,Default,,0000,0000,0000,,from the work that we do.
Dialogue: 0,0:14:05.70,0:14:09.60,Default,,0000,0000,0000,,It's either the research externality,\Nor a teaching externality.
Dialogue: 0,0:14:10.10,0:14:14.60,Default,,0000,0000,0000,,But, you know, if many of our students\Nare going to work in the private sector,
Dialogue: 0,0:14:16.30,0:14:21.70,Default,,0000,0000,0000,,that's fine, but that maybe their\Nemployers should pay for that.
Dialogue: 0,0:14:22.30,0:14:25.10,Default,,0000,0000,0000,,He says, so different from\Npeople working for a Consulting.
Dialogue: 0,0:14:26.30,0:14:26.90,Default,,0000,0000,0000,,Trust me.
Dialogue: 0,0:14:27.20,0:14:33.00,Default,,0000,0000,0000,,It's not clear to me that the number\Nof jobs in academics has changed.
Dialogue: 0,0:14:33.10,0:14:37.60,Default,,0000,0000,0000,,It's just, I feel like this is a\Ngrowing sector whereas Consulting,
Dialogue: 0,0:14:37.70,0:14:42.10,Default,,0000,0000,0000,,your right to raise that, it might\Nbe the same for for Consulting.
Dialogue: 0,0:14:43.30,0:14:44.40,Default,,0000,0000,0000,,But this,
Dialogue: 0,0:14:44.50,0:14:47.50,Default,,0000,0000,0000,,you know, I'm placing more and\Nmore students in these businesses.
Dialogue: 0,0:14:47.50,0:14:50.40,Default,,0000,0000,0000,,So, it's on my mind in\Na way that I've sort of,
Dialogue: 0,0:14:50.80,0:14:55.50,Default,,0000,0000,0000,,you know, not been attentive to consulting\Njobs, you know, Consulting was always,
Dialogue: 0,0:14:55.60,0:15:00.40,Default,,0000,0000,0000,,It's important and I think they'll so\Nthere's some movement from Consulting back
Dialogue: 0,0:15:00.40,0:15:02.60,Default,,0000,0000,0000,,into research. It's a little more fluid.
Dialogue: 0,0:15:02.90,0:15:03.50,Default,,0000,0000,0000,,The,
Dialogue: 0,0:15:03.90,0:15:05.40,Default,,0000,0000,0000,,a lot of the work in the
Dialogue: 0,0:15:06.40,0:15:09.80,Default,,0000,0000,0000,,in both domains. I have to say,\Nit's not really different but
Dialogue: 0,0:15:10.10,0:15:13.70,Default,,0000,0000,0000,,you know, people who are working\Nin the tech sector are doing things
Dialogue: 0,0:15:13.70,0:15:16.80,Default,,0000,0000,0000,,that are potentially of scientific\Ninterest, but mostly it's hidden.
Dialogue: 0,0:15:17.10,0:15:20.90,Default,,0000,0000,0000,,Then you really I have to say, you know,\Nwhy, why is the government paying for this?
Dialogue: 0,0:15:21.80,0:15:25.50,Default,,0000,0000,0000,,Yeah, although yeah, I mean taquitos point,\NI guess it. There's a, there's a data.
Dialogue: 0,0:15:25.60,0:15:30.00,Default,,0000,0000,0000,,Question here of it has the sort\Nof total nanak. It sort of say
Dialogue: 0,0:15:30.50,0:15:30.90,Default,,0000,0000,0000,,private
Dialogue: 0,0:15:31.30,0:15:34.10,Default,,0000,0000,0000,,for-profit sector employment of econ Ph.D.
Dialogue: 0,0:15:34.10,0:15:37.70,Default,,0000,0000,0000,,Program graduates increased or has\Nit just been a substitution from
Dialogue: 0,0:15:37.90,0:15:40.20,Default,,0000,0000,0000,,finance and Consulting towards tack.
Dialogue: 0,0:15:40.30,0:15:44.30,Default,,0000,0000,0000,,I may be a reaction to something\Nthat's not really happening
Dialogue: 0,0:15:44.40,0:15:48.20,Default,,0000,0000,0000,,so bad. I've actually done some work\Nwith some of these tech companies.
Dialogue: 0,0:15:49.10,0:15:52.30,Default,,0000,0000,0000,,So I don't disagree with Justice\Npoint that we need to think
Dialogue: 0,0:15:52.30,0:15:55.10,Default,,0000,0000,0000,,a little bit about the funding model\Nwhose it was in the end paying for the
Dialogue: 0,0:15:55.60,0:15:59.40,Default,,0000,0000,0000,,It education. But from a\Nscientific perspective.
Dialogue: 0,0:16:00.10,0:16:03.50,Default,,0000,0000,0000,,The only do these places have\Nhave great data and nowadays.
Dialogue: 0,0:16:03.50,0:16:07.10,Default,,0000,0000,0000,,They tend to be very careful\Nwith that for privacy reasons,
Dialogue: 0,0:16:07.50,0:16:08.80,Default,,0000,0000,0000,,but also have great questions.
Dialogue: 0,0:16:10.20,0:16:11.20,Default,,0000,0000,0000,,I find it very
Dialogue: 0,0:16:11.60,0:16:13.30,Default,,0000,0000,0000,,inspiring kind of to listen to
Dialogue: 0,0:16:13.30,0:16:15.80,Default,,0000,0000,0000,,the people there and kind of see\Nwhat kind of questions they have
Dialogue: 0,0:16:15.90,0:16:17.30,Default,,0000,0000,0000,,and often their questions.
Dialogue: 0,0:16:18.20,0:16:20.60,Default,,0000,0000,0000,,That also come up outside of these.
Dialogue: 0,0:16:20.70,0:16:25.40,Default,,0000,0000,0000,,These companies have a couple of\Npapers with the rights in the chat.
Dialogue: 0,0:16:25.80,0:16:26.30,Default,,0000,0000,0000,,And then
Dialogue: 0,0:16:26.50,0:16:31.60,Default,,0000,0000,0000,,as soon as an atheist kind of where we\Nlook at ways of combining experimental data
Dialogue: 0,0:16:31.70,0:16:34.10,Default,,0000,0000,0000,,and observational data, and can it there.
Dialogue: 0,0:16:35.50,0:16:38.60,Default,,0000,0000,0000,,Rights Chetty was interested\Nin what is the effect
Dialogue: 0,0:16:38.70,0:16:44.60,Default,,0000,0000,0000,,of Early Childhood programs on outcomes\Nlater in life? Not just kind of test scores,
Dialogue: 0,0:16:44.60,0:16:48.30,Default,,0000,0000,0000,,but on earnings and stuff, and\Nwe cannot be developed methods
Dialogue: 0,0:16:48.60,0:16:51.50,Default,,0000,0000,0000,,that would help you shed\Nlight on that, on the some,
Dialogue: 0,0:16:52.70,0:16:55.00,Default,,0000,0000,0000,,in some settings and the same problems.
Dialogue: 0,0:16:56.30,0:17:00.50,Default,,0000,0000,0000,,Came up kind of in this\Ntech company settings.
Dialogue: 0,0:17:00.80,0:17:03.70,Default,,0000,0000,0000,,And so for my perspective, it's
Dialogue: 0,0:17:04.40,0:17:07.50,Default,,0000,0000,0000,,the same kind of a stocking two\Npeople doing a protocol work.
Dialogue: 0,0:17:07.60,0:17:11.80,Default,,0000,0000,0000,,I tried to kind of look at these\Nspecific problems and then try to come up
Dialogue: 0,0:17:11.90,0:17:18.30,Default,,0000,0000,0000,,with more General problems that we\Nformulating the problems at a higher level.
Dialogue: 0,0:17:18.50,0:17:22.90,Default,,0000,0000,0000,,So that I can think about solutions\Nthat work in a range of settings.
Dialogue: 0,0:17:23.40,0:17:25.50,Default,,0000,0000,0000,,And so from that perspective, the
Dialogue: 0,0:17:25.70,0:17:30.30,Default,,0000,0000,0000,,His with the the tech companies I\Njust very valuable and very useful.
Dialogue: 0,0:17:30.90,0:17:31.40,Default,,0000,0000,0000,,It's know.
Dialogue: 0,0:17:31.70,0:17:33.70,Default,,0000,0000,0000,,We do have students. Now spent
Dialogue: 0,0:17:33.80,0:17:37.20,Default,,0000,0000,0000,,doing internships there and\Nthen coming back and writing
Dialogue: 0,0:17:37.40,0:17:43.40,Default,,0000,0000,0000,,more interesting thesis, as a\Nresult of their experiences there.
Dialogue: 0,0:17:44.60,0:17:47.80,Default,,0000,0000,0000,,If you'd like to watch more\NNobel conversations, click here,
Dialogue: 0,0:17:48.20,0:17:50.50,Default,,0000,0000,0000,,or if you'd like to learn\Nmore about econometrics,
Dialogue: 0,0:17:50.60,0:17:53.20,Default,,0000,0000,0000,,check out Josh's mastering\Neconometrics series.
Dialogue: 0,0:17:53.70,0:17:55.50,Default,,0000,0000,0000,,If you'd like to learn more about he do.
Dialogue: 0,0:17:55.60,0:17:58.30,Default,,0000,0000,0000,,Josh and Isaiah check out\Nthe links in the description.