00:00:01
I'm Adam Grant I'm here with Steven
00:00:03
Dubner the author of The Freakonomics
00:00:05
Enterprise along with Steven levit
00:00:07
Stephen welcome thanks Adam uh your
00:00:09
books have been fascinating to read
00:00:11
they've obviously created an
00:00:12
international explosion your latest is
00:00:14
Think Like a Freak what motivated you to
00:00:16
write this one well boredom no um
00:00:21
honestly after so we wrote the first one
00:00:23
together which was an accident I was a I
00:00:25
am a journalist I hadd written an
00:00:27
article about Steve levit and his
00:00:29
strange brand of um e economics economic
00:00:33
research and I was working on a totally
00:00:35
separate book about the psychology of
00:00:38
money I was really was and remain really
00:00:40
interested in that so I was interested
00:00:41
in behavioral economics what we now call
00:00:43
that so I wrote about Lev then we
00:00:46
decided after this article uh someone
00:00:49
decided it would be a good idea if we
00:00:50
teamed up and we did and wrote free
00:00:51
economics which was very successful we
00:00:53
didn't plan on it being successful we
00:00:54
didn't plan on working together even
00:00:56
once then we thought do we want to do
00:00:58
another and we took about two years to
00:00:59
to decide if we did and if we had if we
00:01:03
could come up with enough good material
00:01:05
for a second one which we did then for a
00:01:07
third one we were pretty sure we weren't
00:01:09
going to do another because we just
00:01:10
didn't want to you know milk it um
00:01:14
Contra the wishes of our publisher an
00:01:16
agent you know you have to anytime
00:01:18
someone sees a franchise presented to
00:01:20
them they want to take it and exploit it
00:01:23
and we just you know we had slightly
00:01:25
different incentives we felt like we'd
00:01:26
profited and been really lucky to get to
00:01:28
that point and we didn't want to exploit
00:01:31
it uh unless we had material that we
00:01:33
were really proud of so again it took us
00:01:34
a couple years to come up with a
00:01:37
framework for a different book and
00:01:39
that's this this third book Think Like a
00:01:40
Freak and basically what what happened
00:01:43
is we hear from people a lot emails
00:01:46
mostly which is great I mean you know of
00:01:48
all the things that the digital
00:01:49
revolution has produced one of the
00:01:51
coolest simplest ones is you can now
00:01:53
contact people who write books that you
00:01:54
read which used to be you used to have
00:01:56
to write a letter to the publisher and
00:01:57
hope they pass along which they never
00:01:59
did so we hear from people with all
00:02:01
these problems and questions and queries
00:02:03
about the way the world works and we
00:02:05
couldn't answer them all it's hard you
00:02:07
know to answer one email well could take
00:02:11
forget about one day could take months
00:02:12
of research so rather than trying and
00:02:15
failing to answer A Shard of those
00:02:17
questions we thought what if we could
00:02:19
write a book that basically deputizes
00:02:21
the entire world or whoever wants to to
00:02:24
think like we do to to kind of develop a
00:02:26
set of rules a kind of blueprint for for
00:02:30
essentially problem solving um it's not
00:02:32
always problem solving but mostly that's
00:02:34
what we try to do and that's what this
00:02:35
book is it's meant to be a fun engaging
00:02:40
practical way to think about the way the
00:02:42
world really works think about the way
00:02:44
incentives really work the way that
00:02:46
people really respond to incentives
00:02:48
rather than how they say they might and
00:02:50
then if you're trying to solve a problem
00:02:51
big or small in business or government
00:02:53
or in your own family you know how you
00:02:55
can maybe slightly increase your chances
00:02:57
to actually solve it that's the idea
00:02:59
well you you certainly accomplish those
00:03:00
goals and you start with the premise
00:03:03
that there are three words that all of
00:03:04
us should probably utter more often than
00:03:06
we do which are I don't know yeah where
00:03:09
did that come from
00:03:11
um well I think that came primarily from
00:03:15
the fact that levit and I Le Steve levit
00:03:18
my co-author you know he lives in a
00:03:20
world of in the world of Academia where
00:03:22
you are I'm a writer I've been a
00:03:25
journalist you know for my whole adult
00:03:27
life and both of us wouldn't have a job
00:03:30
if we pretended we knew all the answers
00:03:33
all the time the whole premise of what I
00:03:35
do as a journalist is go find people who
00:03:37
know things that are interesting or
00:03:39
worthwhile or hidden and ask them about
00:03:42
it try to find out so you have to
00:03:44
acknowledge what you don't know levit in
00:03:46
Academia and you in Academia you know
00:03:48
what academic what good academic
00:03:50
research is like good medical research
00:03:52
like good uh physics or engineering
00:03:54
research is trying to Sol trying to
00:03:56
figure out questions um where you don't
00:03:59
yet know the answers so once you come in
00:04:01
with that mindset you're going to have a
00:04:03
different approach you're going to
00:04:04
acknowledge what you know which may not
00:04:06
be very much and what you don't know and
00:04:08
then you're going to in order to try to
00:04:09
figure out what you need to know you're
00:04:10
going to develop a framework for
00:04:13
experimentation Gathering feedback and
00:04:15
so on now as totally ridiculously
00:04:18
obvious as that sounds what I just said
00:04:21
there are huge quadrants of modern
00:04:23
society particularly in business and in
00:04:25
government where people are constantly
00:04:28
pretending they know the answer to a
00:04:30
question or the solution to a problem
00:04:33
and I get it I understand the way the
00:04:35
incentives work I understand that
00:04:36
reputational you know reputation Works
00:04:39
nobody wants to be the ignoramus or the
00:04:42
dummy so if I'm a politician and someone
00:04:44
says you know Governor blah blah Senator
00:04:47
blah blah you know we just had this
00:04:49
terrible mass shooting at a school if
00:04:52
you could do anything if if all options
00:04:55
were available to you what would you do
00:04:57
to prevent that fut in the future the
00:05:00
way the world works is I'm I'm going to
00:05:02
tell you I'm going to tell you I'm going
00:05:03
to do these three things and that's
00:05:05
what'll do it do you have any evidence
00:05:06
is there any empirical reason to think
00:05:08
that that actually would work often I
00:05:12
hate to say it no and so you see that in
00:05:14
certain Realms politics and in business
00:05:16
where the incentives are different
00:05:17
there's a big incentive to get it right
00:05:19
in business but there's also a lot of
00:05:22
you know for lack of a a more
00:05:24
sophisticated term peer pressure to be
00:05:27
the gal or guy who knows who has the
00:05:30
plan so you know a really basic um you
00:05:33
know rule of thumb is uh um you know or
00:05:38
a basic Mo that happens very frequently
00:05:41
now is a firm will say we need to come
00:05:43
up with a plan or a solution let's get
00:05:45
our 20 top people get together in a room
00:05:47
for an hour that's 20 person hours and
00:05:49
let's come up with the best one the best
00:05:51
idea and then put all our resources into
00:05:53
that and go what are the odds I mean if
00:05:56
this were science what are the odds that
00:05:57
that would bear a good result almost
00:06:00
none so then there's the counter example
00:06:03
of Someone Like A Google who lets its
00:06:06
Engineers take 20% of their time and
00:06:08
work on their projects on the side the
00:06:10
idea being you know have a lot of ideas
00:06:12
most of them will be bad but let the
00:06:14
triage process work and let people
00:06:17
figure out through scientific or
00:06:19
empirical ways how they can really learn
00:06:21
stuff and then once you've done some
00:06:23
experimentation and some small scale
00:06:25
work then maybe put some resources
00:06:27
behind it so that's something that I
00:06:28
think business needs to do
00:06:30
much better but I think many businesses
00:06:32
are moving in the right direction and
00:06:33
the digital Revolution helps that so
00:06:36
much because it's now so easy and cheap
00:06:38
to gather data and do AB testing or a
00:06:41
through z testing to tell you what's
00:06:43
actually working do you have favorite
00:06:45
tests that you're seeing recently that
00:06:47
kind of represent this revolution in a
00:06:50
positive direction as opposed to you
00:06:52
know we can all name bad decisions that
00:06:54
should have been based on evidence but
00:06:55
weren't are there are there any standout
00:06:57
examples for you I'll tell you one
00:06:59
example Le I don't know how well it's
00:07:00
working out I did some reporting on it a
00:07:02
few years ago I have no idea how well
00:07:04
it's working out I like the idea because
00:07:06
it's uh the federal government doing it
00:07:08
and the federal government has typically
00:07:10
been really bad I mean they're the worst
00:07:12
and if you think about it it makes sense
00:07:14
why they are on the top theoretically in
00:07:17
some ways of 50 state governments and
00:07:18
all those Municipal governments under
00:07:20
them so they're not in a position really
00:07:23
to go micro um and I I understand that
00:07:26
but what they did with this race to the
00:07:28
toop program and education I thought was
00:07:30
a really good idea and again I don't
00:07:31
know how well it's going to work out but
00:07:34
basically they set up first of all a
00:07:35
contest which means that there are
00:07:37
incentives that presumably are going to
00:07:40
work better than no incentives or then
00:07:42
than there's or better than some kind of
00:07:44
negative reinforcement that we're used
00:07:46
to and basically uh you know Arie Duncan
00:07:49
Secretary of Education the president
00:07:50
said to all the states hey we need to
00:07:53
think of ways to improve or rethink our
00:07:56
education system and believe me you know
00:07:59
I could talk about that for years
00:08:01
because education is such a complicated
00:08:05
box with so many inputs and so many
00:08:07
outputs it's it's really easy to look
00:08:09
for Magic Bullets you know hire Pay
00:08:11
Teachers more or get rid of the unions
00:08:13
or smaller class size everybody likes
00:08:15
those Magic Bullets but it's a very
00:08:16
complicated scenario so basically the
00:08:19
Department of Education said we go to
00:08:21
all 50 states each of you we want you to
00:08:23
try to come up with a good program a
00:08:25
good idea a good solution that works and
00:08:29
if works we'll pay you for it and
00:08:31
there's a good chance then we'll run it
00:08:33
up and we'll kind of you know
00:08:34
standardize it that's the right kind of
00:08:38
thinking um think
00:08:40
small don't pretend you know the answers
00:08:43
um experiment get feedback these are all
00:08:46
the premises of Think Like a Freak
00:08:48
really and and there's one example where
00:08:49
even the federal government which we're
00:08:50
not used to thinking that empirically I
00:08:53
think you know gave it a good shot you
00:08:55
have some fascinating examples of the
00:08:56
book uh that probably stretch beyond
00:08:58
what most readers would themselves be
00:09:00
willing to do uh one of which is you
00:09:02
actually got people to agree to let you
00:09:05
randomly assign them to do things like
00:09:07
ask for a raise or quit their job or
00:09:09
even break up with their significant
00:09:10
other uh what was the logic behind that
00:09:12
what you learn so this came about
00:09:15
because of a uh a podcast episode we do
00:09:18
free economics Radio podcast and public
00:09:20
radio show and we did an episode that I
00:09:23
I love it was just to me this was a
00:09:26
great topic because it's a blend of data
00:09:29
and empirical thinking with narrative
00:09:33
storytelling which is my tradition um so
00:09:35
it was called the upside of quitting and
00:09:36
it was basically making an economic
00:09:39
argument to some degree which is
00:09:42
um considering that most of us have been
00:09:45
conditioned to not quit we've been
00:09:47
conditioned to think that quitting is an
00:09:49
equivalent of fail is a failure a form
00:09:51
of failure um how do we know that that's
00:09:53
true and how much of the counter might
00:09:55
be opposite so um you know if you think
00:09:58
about a project a job a war you know all
00:10:01
a relationship all these things that you
00:10:03
might could quit but because of sunk
00:10:04
costs and because of peer pressure and
00:10:07
because of um you know your own moral
00:10:09
position you might not want to quit we
00:10:11
tried to look at you know what is the
00:10:13
upside of quitting and we we argued that
00:10:14
there's a significant upside that people
00:10:16
are really bad at um estimating um
00:10:19
opportunity cost what they could be
00:10:20
doing if they could quit and so on so
00:10:23
but the fact is it's really hard to get
00:10:26
data on this because it's not like
00:10:29
you can go into you know one big school
00:10:32
district and say you know I'm going to
00:10:34
take a thousand kids and I'm going to
00:10:36
totally mix them so their grades are you
00:10:38
know equivalent on either side and I'm
00:10:39
going to randomly Force half of them to
00:10:42
quit school uh don't allow them to go
00:10:44
back to school and then 10 and 20 and 30
00:10:47
years later see how their lives turned
00:10:48
out that's one way to you might do that
00:10:50
experiment but of course we couldn't do
00:10:52
that the people who tend to quit school
00:10:53
tend to be a very different population
00:10:55
than the people who don't quit school so
00:10:56
to compare them after is not equivalent
00:10:59
so what we came up with was a website
00:11:02
called free economics experiments where
00:11:04
we offered that if people had a decision
00:11:06
to make it wasn't necessarily something
00:11:08
to quit although usually it was if they
00:11:10
had a decision to make um should I quit
00:11:13
my job and go back to grad school should
00:11:15
I join the military or stick with my job
00:11:17
should I leave my boyfriend or
00:11:18
girlfriend or husband or wife uh all
00:11:21
different kinds of things should I get a
00:11:22
tattoo or not and if they had a decision
00:11:25
and they couldn't and they really were
00:11:26
on the really truly on the fence then we
00:11:29
offered to help him out and flip a coin
00:11:32
for him and we asked all we asked is
00:11:35
that they fill out a survey beforehand
00:11:37
telling us about it and that then they
00:11:39
tell us whether or not they followed the
00:11:41
coin because we have no power to make
00:11:43
them follow the coin and then we would
00:11:45
follow up with them and and do research
00:11:47
later to find out uh what their outcomes
00:11:51
were and so to many different categories
00:11:55
a variety of outcomes and the research
00:11:57
isn't done yet but the short short
00:11:59
answer is that when people quit
00:12:02
something that they were generally
00:12:05
really worried about quitting their
00:12:07
lives tend to get a little bit better so
00:12:09
even if they didn't get a lot worse you
00:12:11
might argue that it's a a pretty good uh
00:12:13
a pretty good bet so
00:12:16
basically I think we should all consider
00:12:19
quitting as a really good option um but
00:12:22
you know it's hard when you have the
00:12:24
words of Vince Lombardi you know a
00:12:26
quitter never wins and a winner never
00:12:27
quits which was wasn't Lombardi
00:12:29
originally and Winston Churchill telling
00:12:32
people never never never never never
00:12:34
give up in anything large or small
00:12:36
greater Petty you know you have these
00:12:37
great people and that gets in your ear
00:12:39
and it convinces you that oh man if I
00:12:41
start a project I have to see it through
00:12:43
but if you just for five minutes spend
00:12:45
some time thinking about the sunk cost
00:12:48
and thinking about opportunity cost then
00:12:51
you can really get to different places
00:12:53
so that's what we were trying to
00:12:54
accomplish there well I'm the opposite
00:12:56
of quitting I think another one of the
00:12:58
most audacious things that comes out and
00:12:59
think like a freak dates back to Super
00:13:01
Freakonomics I remember reading that you
00:13:03
said that that basically terrorists
00:13:05
should buy life insurance and thinking
00:13:07
there was an awfully interesting way
00:13:09
that you could use that information
00:13:10
which you then went and did tell us a
00:13:12
little so you thought yeah you you were
00:13:13
thinking a step ahead of most people
00:13:15
reading I was just right with you
00:13:16
thinking like a social scientist well a
00:13:18
lot of people were thinking so right in
00:13:20
super fre economics we described this
00:13:22
algorithm that levit worked on with a a
00:13:25
British bank that was trying to catch
00:13:27
terrorists from nothing more than retail
00:13:29
banking data that was for our purposes
00:13:32
anonymized and so there were all these
00:13:33
metrics that seemed to indicate someone
00:13:35
who might be uh you know either we
00:13:39
couldn't tell from their banking dat it
00:13:40
wasn't like we know that they're buying
00:13:42
bomb making materials although that
00:13:43
would have showed up but they're not
00:13:44
that stupid but if they were consorting
00:13:46
with other people um but then there were
00:13:48
other Clues um that would uh raise the
00:13:53
that would move the needle for
00:13:55
indicating that someone is is quite
00:13:57
possibly involved in it but then there
00:13:58
was one argument that we made in super
00:14:00
fre economics that if you really um one
00:14:03
really good indicator is that young men
00:14:06
who are prone to thinking about
00:14:08
terrorism are not going to buy life
00:14:09
insurance from their bank because why
00:14:11
would you because if you if you kill
00:14:14
yourself in the commission of a crime
00:14:15
it's not going to pay out so why would
00:14:16
you waste the money that was the
00:14:18
argument we made so basically we were
00:14:20
saying here's this algorithm we made
00:14:22
here are what here are the metrics that
00:14:24
comprise the algorithm there's one or
00:14:26
two that are too good we're not going to
00:14:27
tell you but here's by the way is
00:14:29
another really good one that we will
00:14:30
tell you which is that you should buy
00:14:31
life insurance from your bank now this
00:14:34
was done for real the algorithm was real
00:14:37
and we were really trying to catch
00:14:38
people and it worked to some degree we
00:14:41
think it worked to some degree but the
00:14:42
life insurance thing was just a red
00:14:44
herring or it was a it was a trap and
00:14:47
the idea was that very few people buy
00:14:50
life insurance from their banks anyway
00:14:52
like very very very few people even
00:14:53
though most banks offer it and the same
00:14:55
is in Britain but by putting that in our
00:14:57
book and then when we went to uh Britain
00:14:59
for book tour what most people were
00:15:01
saying when they interviewed us is you
00:15:04
idiots why would you go to the trouble
00:15:06
to build a a tool to catch terrorists
00:15:09
then tell them how to evade it and then
00:15:11
we were just like well yeah maybe we huh
00:15:14
because that's the tool yeah so um
00:15:17
because that was the tool so then the
00:15:20
algorithm was already in place uh by
00:15:22
that point so anybody that saw us
00:15:24
talking about this or saw journalists
00:15:27
there yelling at us for giving away this
00:15:29
clue what are they maybe now a little
00:15:31
bit more likely to do if they're guilty
00:15:33
to cover their tracks go buy some life
00:15:35
insurance from their bank then the
00:15:37
algorithm being in place we could look
00:15:39
at that data and see who already fit the
00:15:41
profile and now additionally ran out and
00:15:43
bought some life insurance and that
00:15:45
increased the likelihood even a little
00:15:46
bit more that those people were bad guys
00:15:49
that generated a new smaller list that
00:15:52
we then passed on to the authorities
00:15:53
there do you know what was done with
00:15:54
that list so my colleague
00:15:57
Lev presented It To The Head or near
00:16:01
head of uh the Intelligence
00:16:06
Division he presented it it was
00:16:08
literally an envelope with a wax seal
00:16:11
because he'd been given the envelope
00:16:12
from the bank because we were again we
00:16:14
weren't allowed to see any identifying
00:16:17
data on any of the people and it was a
00:16:19
little bit it it was a little bit um
00:16:21
maybe James Bond outtake combined with
00:16:24
uh the very end of Raiders the Lost Arc
00:16:26
where they get the arc and it and it
00:16:27
goes into the warehouse and the camera
00:16:29
pulls back and you see that there are 8
00:16:32
million boxes so we have no idea uh
00:16:36
whether this list that we were quite
00:16:41
sure had some value for anti-terror uh
00:16:45
purposes um whether or to what degree it
00:16:49
was put to use uh that that wasn't part
00:16:51
of the game we were allowed to play in
00:16:52
basically yeah so in closing um other
00:16:55
than saying I don't know when working on
00:16:58
Freakonomics the book the radio the
00:17:00
podcast the movie um through that whole
00:17:02
process what's the biggest lesson you've
00:17:04
taken away about how to think like a
00:17:10
freak oh the biggest lesson
00:17:16
um I I guess I should know that by now
00:17:18
shouldn't I since this is kind of my
00:17:20
thing
00:17:24
um you know I it's amazing that I'm
00:17:27
coming up blank to the question that
00:17:29
should be the first one that I don't
00:17:30
come up blank to um I I guess I really
00:17:35
um so this is more of a philosophical
00:17:37
answer than a a kind of tactical or
00:17:40
strategic answer to me the challenge is
00:17:43
always going to be the blend between the
00:17:47
empirical or scientific or data whatever
00:17:50
you want to call that and the intuitive
00:17:53
or the human or the Humane whatever you
00:17:55
want to call that and what I mean by
00:17:56
that is especially in this era of big
00:17:59
data which is kind of what we've been
00:18:01
doing for a long time now not as
00:18:02
systematically as a lot of firms and
00:18:04
governments are doing it now but you
00:18:06
know we believe in it we believe that if
00:18:07
you get a pile of data representing a
00:18:09
million decisions that that's better
00:18:10
than asking three people what decisions
00:18:12
they made so while I very much believe
00:18:15
that to be true and I very much applaud
00:18:17
all the instincts for all of us to kind
00:18:20
of work with data in aggregate to
00:18:23
distill the biggest truths I also know
00:18:26
that we're humans and that we're
00:18:28
fallible and bi I shouldn't say fallible
00:18:30
although we're fallible too we're biased
00:18:32
in a lot of ways so that even if you
00:18:35
could tell me or I could tell you the
00:18:37
most
00:18:38
foolproof strategic way to reach a
00:18:41
decision or the best decision to make or
00:18:43
the best strategy or the best set of
00:18:45
numbers to embrace there might be a lot
00:18:47
of good reasons why you still won't be
00:18:49
successful and that's because the people
00:18:52
that you are now employing that strategy
00:18:54
on or the people that you're now
00:18:55
offering those incentives to don't
00:18:57
respond the way way you think about the
00:18:59
problem and that requires a lot of
00:19:01
humility um and that's something that
00:19:03
people in government in business in
00:19:06
Academia in journalism uh everywhere
00:19:10
where you know people in all those
00:19:12
fields are really used to like when we
00:19:13
come up with something and we put it
00:19:15
into play we're used to people like
00:19:16
snapping up and saying okay we're going
00:19:18
to do this now that's a lot of power
00:19:20
that's a lot of authority but with that
00:19:22
power and authority comes I think the
00:19:24
need for humility to understand that
00:19:26
when you make decisions like that put
00:19:28
out incentives whatever they are big or
00:19:31
small governmental or non that there are
00:19:33
people on the other end of that and how
00:19:35
it affects their lives the decision
00:19:38
makers don't often think through very
00:19:39
well how they'll respond to the
00:19:40
incentives and so on and so that to me
00:19:42
is the balance to be as scientific as
00:19:45
you can while understanding that even if
00:19:47
I can present a 100 people with the
00:19:49
science that says hey you should really
00:19:51
do this 90 of them might have a really
00:19:53
good reason for not wanting to do it
00:19:55
they might be wrong I might be right but
00:19:57
it doesn't mean I'll win the argument
00:19:59
and being right doesn't win that many
00:20:01
arguments um weirdly enough I mean there
00:20:03
are a lot of people who are right about
00:20:05
a lot of things who who don't get their
00:20:07
way so I think that's really the the
00:20:09
trickiest part I mean I'm working on a a
00:20:11
radio Podcast episode right now about
00:20:13
the flu vaccine very very simple it's
00:20:15
pretty effective the flu vaccine about
00:20:17
60% or so influenza along with pneumonia
00:20:21
is always one of the 10 leading causes
00:20:22
of death in the US which most people
00:20:24
don't think about or don't know and um
00:20:26
and yet very few people a lot of people
00:20:29
who should get the flu vaccine don't why
00:20:31
it's kind of a conundrum so we're going
00:20:33
through all these different layers of
00:20:35
Behavioral and uh PR um and uh um you
00:20:41
know financial decisions to try to
00:20:43
figure out how is something as seemingly
00:20:45
simple as this so hard to accomplish and
00:20:47
that is what I'm constantly reminded is
00:20:49
the people the smart money may be smart
00:20:52
but unless it can deliver on something
00:20:54
that really raises everyone's Behavior
00:20:57
then it's not worth that much thank you
00:21:00
for joining us my pleasure Adam
00:21:05
[Music]
00:21:24
thanks
