♪ [music] ♪
- [narrator] Welcome
to Nobel Conversations.
In this episode, Josh Angrist
and Guido Imbens
sit down with Isaiah Andrews
to discuss and disagree
over the role of machine learning
in applied econometrics.
- [Isaiah] So, of course,
there are a lot of topics
where you guys largely agree,
but I'd like to turn to one
where maybe you have
some differences of opinion.
So I'd love to hear
some of your thoughts
about machine learning
and the goal that it's playing
and is going to play in economics.
- [Guido] I've looked at some data
like the proprietary
so that there's
no published paper there.
There was an experiment
that was done
on some search algorithm.
And the question was...
it was about ranking things
and changing the ranking.
That was sort of clear...
that was going to be
a lot of heterogeneity there.
Mmm,
You know, if you look for say,
a picture of Britney Spears
that it doesn't really matter
where you rank it
because you're going to figure out
what you're looking for,
whether you put it
in the first or second
or third position of the ranking.
But if you're looking
for the best econometrics book,
if you put your book
first or your book tenth,
that's going to make
a big difference
how much how often people
are going to click on it.
And so there you go --
- [Josh] Why do I need
machine learning to discover that?
It seems like could
I can discover it simply?
- [Guido] So in general--
- [Josh] There were lots
of possible...
- You what you want to think about
there being lots of characteristics
of the items
that you want to understand
what drives the heterogeneity
in the effect of--
- But you're just predicting
In some sense, you're solving
a marketing problem.
- [inaudible] it's causal effect,
- It's causal, but it has
no scientific content.
Think about...
- No, but it's similar things
in medical settings.
If you do an experiment,
you may actually be very interested
in whether the treatment
works for some groups or not.
And you have a lot of individual
characteristics,
and you want
to systematically search.
- Yeah. I'm skeptical about that --
that sort of idea that there's
this personal causal effect
that I should care about,
and that machine learning
can discover it
in some way that's useful.
So think about -- I've done
a lot of work on schools,
going to, say, a charter school,
a publicly funded private school,
effectively, you know,
that's free to structure
its own curriculum
for context there.
Some types of charter schools
generate spectacular
achievement gains,
and in the data set
that produces that result,
I have a lot of covariance.
So I have baseline scores,
and I have family background,
the education of the parents,
the sex of the child,
the race of the child.
And, well, soon as I put
half a dozen of those together,
I have a very high dimensional space.
I'm definitely interested
in sort of coarse features
of that treatment effect,
like whether it's better for people
who come from
lower income families.
I have a hard time believing
that there's an application,
for the very high dimensional
version of that,
where I discovered
that for non-white children
who have high family incomes
but baseline scores
in the third quartile
and only went to public school
in the third grade
but not the sixth grade.
So that's what that high
dimensional analysis produces.
This very elaborate
conditional statement.
There's two things that are wrong
with that in my view.
First, I don't see it as...
I just can't imagine
why it's actionable.
I don't know why
you'd want to act on it.
And I know also
that there's some alternative model
that fits almost as well,
that flips everything,
Because machine learning
doesn't tell me
that this is really
the predictor that matters.
It just tells me that
this is a good predictor.
And so, I think
there is something different
about the social science contest.
- [Guido] I think
the [socialized sign] applications
you're talking about,
once were...
I think there's not a huge amount
of heterogeneity in the effects.
- [Josh] There might be
if you allow me
to to fill that space.
- No... not even then.
I think for a lot
of those interventions,
you would expect that the effect
is the same sign for everybody.
There may be small differences
in the magnitude, but it's not...
For a lot of these education
defenses -- they're good for everybody.
It's not that they're bad
for some people
and good for other people,
and that is kind
of very small pockets
where they're bad there.
But it may be some variation
in the magnitude,
but you would need very,
very big data sets to find those.
I agree that in those cases,
they probably wouldn't be
very actionable anyone.
But I think there's a lot
of other settings
where there is
much more heterogeneity.
- Well, I'm open
to that possibility,
and I think the example you gave
is essentially a marketing example.
- No, those have implications for it
and that's the organization,
whether you need
to worry about the...
- Well, I need to see that paper.
- So the sense I'm getting...
- We still disagree on something.
- Yes.
[laughter]
- We haven't converged
on everything.
- I'm getting that sense.
[laughter]
- Actually, we've diverged on this
because this wasn't around
to argue about.
[laughter]
- Is it getting a little warm here?
- Warmed up. Warmed up is good.
The sense I'm getting is, Josh,
you're not saying
that you're confident
that there is no way
that there is an application
where the stuff.
It's useful you are saying
you are unconvinced by
the existing application to date.
Fair enough.
- I'm very confident.
[laughter]
- In this case.
- I think Josh does have a point
that even in the prediction cases
where a lot of the machine learning
methods really shine
is where there's just a lot
of heterogeneity.
- You don't really care much
about the details there, right?
It doesn't have
a policy angle or something.
- They kind of recognizing
handwritten digits and stuff.
It does much better there
than building
some complicated model.
But a lot of the social science,
a lot of the economic applications,
we actually know a huge amount
about the relationship
between its variables.
A lot of the relationships
are strictly monotone.
Education is going to increase
people's earnings,
irrespective of the demographic,
irrespective of the level
of education you already have.
- Until they get to a Ph.D.
- Yeah, there is a graduate school...
[laughter]
but go over a reasonable range.
It's not going
to go down very much.
In a lot of the settings
where these machine learning
methods shine,
there's a lot of [ ]
kind of multimodality
in these relationships,
and they're going to be
very powerful.
But I still stand by that.
These methods just have
a huge amount to offer
for economists,
and they're going to be
a big part of the future.
- [Isaiah] Feels like
there's something interesting
to be said about
machine learning here.
So, Guido, I was wondering,
could you give some more...
maybe some examples
of the sorts of examples
you're thinking about
with applications [ ] at the moment?
- So on areas where
instead of looking
for average cause or effects
we're looking for
individualized estimates,
predictions of cause or effects
and the machine learning algorithms
have been very effective,
Traditionally, we would have done
these things using kernel methods.
And theoretically they work great,
and there's some arguments
that, formally,
you can't do any better.
But in practice,
they don't work very well.
Random causal forest-type things
that Stefan Wager and Susan Athey
have been working on
have used very widely.
They've been very effective
in these settings
to actually get causal effects
that vary be [ ].
I think this is still just the beginning
of these methods.
But in many cases,
these algorithms are very effective
as searching over big spaces
and finding the functions that fit very well
in ways that we couldn't
really do beforehand.
- I don't know of an example
where machine learning
has generated insights
about a causal effect
that I'm interested in.
And I do know of examples
where it's potentially
very misleading.
So I've done some work
with Brigham Frandsen,
using, for example, random forest
to model covariate effects
in an instrumental
variables problem
Where you need you need
to condition on covariance.
And you don't particularly
have strong feelings
about the functional form for that,
so maybe you should curve...
be open to flexible curve fitting,
and that leads you down a path
where there's a lot
of nonlinearities in the model,
and that's very dangerous with IV
because any sort
of excluded non-linearity
potentially generates
a spurious causal effect
and Brigham and I
showed that very powerfully.
I think in the case
of two instruments
that come from a paper of mine
with Bill Evans,
where if you replace it
a traditional two stage
[ ] squares estimator
with some kind of random forest,
you get very precisely
estimated [non-sense] estimates.
I think that's a big caution.
In view of those findings
in an example I care about
where the instruments
are very simple
and I believe that they're valid,
I would be skeptical of that.
So non-linearity and IV
don't mix very comfortably.
No, it sounds like that's already
a more complicated...
- Well, it's IV....
- Yeah.
- ...and we work on that.
[laughter]
- Fair enough.
- As Editor of Econometric [guy],
a lot of these papers
cross by my desk,
but the motivation is not clear
and, in fact, really lacking.
They're not... [we call] type
semi-parametric foundational papers.
So that that's a big problem.
A related problem is that we have
this tradition in econometrics
of being very focused
on these formal [ ] results.
We have just have a lot of papers
where people propose a method
and then establish
the asymptotic properties
in a very kind of standardized way.
- Is that bad?
- Well, I think it's sort
of closed the door
for a lot of work
that doesn't fit it into that.
where in the machine
learning literature,
a lot of things
are more algorithmic.
People had algorithms
for coming up with predictions
that turn out
to actually work much better
than, say, nonparametric
kernel regression
For a long time, we were doing all
the nonparametrics in econometrics,
we were using kernel regression,
and it was great for proving theorems.
You could get [ ] intervals
and consistency,
and asymptotic normality,
and it was all great,
But it wasn't very useful.
And the things they did
in machine learning
are just way, way better.
But they didn't have the problem--
- That's not my beef
with machine learning theory.
[laughter]
No, but I'm saying there,
for the prediction part,
it does much better.
- Yeah, it's a better
curve fitting to it.
- But it did so in a way
that would not have made
those papers
initially easy to get into,
the econometrics journals,
because it wasn't proving
the type of things.
When Brigham was doing
his regression trees
that just didn't fit in.
I think he would have had
a very hard time
publishing these things
in econometric journals.
I think we've limited
ourselves too much
that left us close things off
for a lot of these
machine learning methods
that are actually very useful.
I mean, I think, in general,
that literature,
the computer scientist,
have proposed a huge number
of these algorithms
that actually are very useful.
and that are affecting
the way we're going
to be doing empirical work.
But we've not fully internalized that
because we're still very focused
on getting point estimates
and getting standard errors
and getting P values
in a way that we need to move beyond
to fully harness the force,
the benefits
from the machine learning literature.
- On the one hand, I guess I very
much take your point
that sort of the traditional
econometrics framework
of sort of propose a method,
prove a limit theorem
under some asymptotic story,
story story, story story...
publisher paper is constraining.
And that, in some sense,
by thinking more broadly
about what a methods paper
could look like,
we may [write] in some sense.
Certainly the machine learning
literature has found a bunch of things,
which seem to work quite well
for a number of problems
and are now having
substantial influence in economics.
I guess a question I'm interested in
is how do you think
about the role of...
sort of -- do you think there is
no value in the theory part of it?
Because I guess a question
that I often have
to sort of seeing that output
from a machine learning tool,
that actually a number of the
methods that you talked about
actually do have inferential results
developed for them,
something that
I always wonder about
of uncertainty quantification
and just...
I have my prior,
I come into the world with my view.
I see the result of this thing.
How should I update based on it?
And in some sense,
if I'm in a world
where things are normally distributed,
I know how to do it here --
here I don't.
And so I'm interested to hear
what you think about that.
- I don't see this as sort
of saying, well,
these results are not interesting,
but it's going to be a lot of cases
where it's going
to be incredibly hard
to get those results
and we may not be able to get there
and we may need to do it in stages
where first someone says,
"Hey, I have
this interesting algorithm
for doing something
and it works well by some of the criterion
that on this particular data set,
and I'm visit put it out there,
and maybe someone will figure out a way
that you can later actually
still do inference
on the [sum] condition,
and maybe those are not
particularly realistic conditions,
then we kind of go further.
But I think we've been
constraining things too much
where we said,
"This is the type of things
that we need to do.
And in some sense,
that goes back
to the way Josh and I
thought about things for the
[local average treatment] effect.
That wasn't quite the way
people were thinking
about these problems before.
There was a sense
that some of the people said
the way you need to do
these things is you first say,
what you're interested in
in estimating
and then you do the best job
you can in estimating that.
and what you guys are doing
is you're doing it backwards.
You kind of say,
"Here, I have an estimator,
and now I'm going to figure out
what it's estimating,
and I suppose you're going to say
why you think that's interesting
or maybe why it's not interesting,
and that's not okay.
You're not allowed
to do that that way.
And I think we should
just be a little bit more flexible
in thinking about
how to look at problems
because I think
we've missed some things
by not doing that.
- [Josh] So you've heard
our views, Isaiah.
You've seen that we have
some points of disagreement.
Why don't you referee
this dispute for us?
[laughter]
- Oh, it's so nice of you
to ask me a small question.
So I guess for one,
I very much agree with something
that Guido said earlier of...
[laughter]
- So one thing where it seems
where the case for machine learning
seems relatively clear
is in settings where
we're interested in some version
of a nonparametric
prediction problem.
So I'm interested in estimating
a conditional expectation
or conditional probability,
and in the past, maybe
I would have run a kernel...
I would have run
a kernel regression
or I would have run
a series regression,
or something along those lines.
It seems like, at this point,
we've a fairly good sense
that in a fairly wide range
of applications,
machine learning methods
seem to do better
for estimating conditional
mean functions
or conditional probabilities
or various other
nonparametric objects
than more traditional nonparametric
methods that were studied in econometrics
and statistics, especially
in high dimensional settings.
So you thinking of maybe the propensity
score or something like that?
So exactly, so nuisance functions. Yeah.
So things like propensity scores
things or I mean even objects
of more direct inference
interest, like conditional
average treatment effects, right?
Which of the difference of two
conditional, expectation functions,
potentially things like that.
Of course, even there,
right? We the the theory
for in France or the theory for
sort of how to how to interpret,
how to make large simple statements
about some of these things are
less well-developed depending on the
machine learning, estimator used.
And so, I think there's something
that is tricky is that we
can have these methods, which work a lot,
which seemed to work a lot
better for some purposes.
But which we need to be a bit
careful in how we plug them in or how
we interpret the resulting statements.
But of course, that's a very,
very active area right now. We're
People are doing tons of great work.
And so I exfoli expect and hope
to see much more going forward there.
So one issue with machine learning,
that always seems a danger is, or
that is sometimes a danger
and had some times led to
applications that have
made. Less sense, is
when folks start with a method that are
start with a method that they're very
excited about rather than a question,
right? So sort of starting with
a question where here's the
object I'm interested in here is
the parameter of Interest. Let me
You know,
think about how I would
identify that thing,
how I would recover that
thing, if I had a ton of data,
oh, here's a conditional
expectation function.
Let me plug in an estimator on
machine. Learning estimator for that.
That seems very very sensible.
Whereas, you know, if I
digress quantity on price
and say that I used a
machine learning method,
maybe I'm satisfied that that
solves the in dodging, 80 problem.
We're usually worried
about their maybe I'm not,
but again, that's something where the,
the way to address. It, seems
relatively clear, right?
It's the find your object of interest and
think about, is that just
bringing the economics?
Exactly.
And and can I think about it,
and they denied it, but harnessed
the power of the machine
learning methods for precisely
for some of the components precisely.
Exactly. So sort of, you know, the, the,
the question of interest is the same as
the question of interest is always been,
but we now better methods for estimating
some pieces of this, right? The
the place that seems harder to, uh,
harder to forecast is Right.
Obviously, there's a huge amount
going in going on in the machine.
Learning literature
and the great sort of The Limited ways
of plugging it in that I've referenced
so far are limited piece of that.
And so I think there are all sorts of
other interesting questions about where,
right sort of
where does this interaction
go? What else can we learn?
And that's something where,
you know, I think there's
a ton going on which seems very promising
and I have no idea what the answer is.
No, no. No, it's I so I totally
agree with that but it's no.
That's makes it very exciting.
And I think that's just a
little work to be done there.
All right. So I say agrees
with me there, say that person.
If you'd like to watch more
Nobel conversations, click here,
or if you'd like to learn
more about econometrics,
check out Josh's mastering
econometrics series.
If you'd like to learn more
about he do Josh and Isaiah
check out the links in the description.