talk about unintended consequences and I
consider it to be a primer to critical
thinking the reason why I chose this
topic is because today it is very
fashionable to talk about evidence-based
policymaking and to talk about empirical
realities but I think it is important to
understand that to be able to get any
benefit out of empirical realism as we
put it we have to understand the data
when and wherever we see it it is
important to understand that if two
people see the same things they can come
to very different conclusions right and
it depends crucially on how they have
decided to look at the data you know you
are all business school people many of
you here so you know this very famous
example that is used right two people
were sent to a place to find out if
there was a market for shoes and both
k-max are exactly the same things and
one said there’s a huge market for shoes
and the other said there is no market
for shoes right the observation was
nobody was wearing shoes right based on
the theory that they had about why
people don’t wear shoes they came to
exactly different conclusions so that is
what I mean when I say that you have to
be able to abstract and have a theory to
be able to understand the data and that
is what I’m going to talk about today so
we start with a very simple problem we
are going to go through a thought
experiment in the sense that there is no
data I’m not going to talk about any
data we are going to construct the data
in our own way as if so let’s take two
places a and B and we find that people
travel from A to B every day during rush
hour a is where they stay B is where
they work right so the commute every day
but the important thing that we are
looking at is how much time it takes to
go from A to B so now in so we run this
you know periodically to figure out
whether people are taking more time or
less time so we do this study in 2010
and we find that there are 2,000 cars
that go from A to B and they take 45
minutes to travel from A to B okay so
that’s the commute time we run this
again in 2015 to see if things have
changed and we find 2,000 cars are still
going from A to B but they’re taking 50
minutes so that is the data we get and
we are now trying to figure out why is
it that this has happened people have
gone or they’re driving slow five years
good enough time to get old so they’re
driving slower all of them and hence
it’s taking 50 minutes or is there
something else that is happening so we
find that when we look closer into the
problem we find that there are four
nodal points that are of interest to us
a and B to begin with and C and D to Y
because what was happening in 2010 is
that there are thousand cars out of the
2000 that we started with thousand cars
were going to be via C okay and they
were taking 45 minutes and another
thousand cars were going to be via D and
they were taking also 45 minutes and so
that’s what we figure out when we do a
little bit more detailed analysis of
what was happening in 2010 now we come
to 2015 and we again do the survey we
had again ask people at B that how did
you come why do you take you 15 minutes
how did you come and this and I find
that thousand of those two thousand a
traveling a CD B right so they’re not
doing ACB but they’re going AC dB
they’re not going a DB they’re going AC
DB and that’s the reason why probably
they are taking 50 minutes the other
thousand cars are doing exactly the same
thing so they’re also taking 50 minutes
and thereby hangs a tale that all 2,000
cars are going from A to B earlier they
were breaking up into two groups one
going ACB they are they’re going a DB
now all 2,000 are going AC DB and now we
have to figure out why this is happening
and here is where I bring in the
unintended consequence of increasing
their travel time by five minutes
even though they’re going to the same
place same number of people so what we
are going to do now is we are going to
go back to 2010 right and see what was
happening so people who are going a to
be right via see were following this
route what is so special about this
route that the stretch of the road seat
would be was a highway and what is so
special about a highway what is special
about a highway is that once you enter
the highway right you travel at the same
speed regardless of how many cars there
are there is no congestion but the
stretch a to see had congestion it’s
like a bridge one lane bridge right so
clearly the time taken to go from A to C
depends on the number of cars in that
stretch a to C and we are doing a very
simple model and we are saying that if T
is the number of cars then the amount of
time taken to go from A to C is T over
100 remember thousand cars are going a
Seabee so thousand cars were taking ten
minutes to move from A to C and then 35
minutes from C to B and that was the
reason why they were taking 45 minutes
okay 10 plus 35 what were the people who
were traveling the other way
ADB what were they doing well they also
had a 35 minute highway stretch and they
also had to cross the same bridge right
at a different place of course sorry the
same river they had to cross at a
different place and there they faced
exactly the same congestion problem so
they also took ten minutes to cross it
right so 35 plus 10 a D and B so that
was the reason why they were taking 45
minutes in 2010 now we come to 2015 so
what’s happening in 2015 again we have
these four nodes because we know they
are going ACD B so our interest is in
knowing why is it that they’re going C D
well we find that there is a new highway
that has been created between C and D
okay so there’s a new road that has been
made and one would expect that when you
build more roads the time taken to
commute should go down at least
it should not go up but here we find
that it has gone up to 50 minutes the
only thing that has changed is that
there is a new road between C and D a
new highway between C and D and this
highway takes ten minutes to cross from
C to D okay so it’s really modern
highway concrete roads and all of that
so now what happens suppose we know that
all 2000 cars are going ACD B so what is
happening when they go from A to C 2,000
cars it’s taking 20 minutes earlier it
was taking 10 minutes now it’s taking 20
minutes 10 minutes to go from C to D and
again all 2,000 cars are going D to B
right so that’s the reason why they’re
taking 50 minutes because 20 minutes
plus 10 minutes plus 20 minutes is 50
minutes so this new road has resulted in
this now we need to be able to ask to
understand this completely unintended
consequence right we need to be able to
understand why is it that people are
behaving in this way you know when you
build a river sorry when you build a dam
to change the course of a river right
you do a lot of study and you figure out
exactly how fluid dynamics works right
and then you build a dam and divert the
water now imagine if these water
particles were you and me who figured
out that this is what the engineers are
trying to do right and we love we love
to act contrary unto what others want us
to do so we decide not to divert
ourselves so we jump over the dam think
about it that’s exactly what we do as
people and that’s exactly what
policymakers need to keep in mind that
they’re not dealing with natural
particles they’re dealing with what we
would call strategic elements and
therefore we need to know how they
behave to be able to understand how to
intervene in the marketplace some things
that we don’t do whether in RCT or the
so called random from randomized control
trial experiments all in the
experimental approach that we often for
it’s totally devoid of theory and that
should be avoided all right so that’s
the reason why I’m going through this
example so let’s see what happens I call
it phase one so phase one what happens
there are thousand cars going ACB right
before the road is built then the road
is built now let’s look at these
thousand cars right so if all of them
change to a c DB then they’re going to
take ten minutes to go from A to C ten
minutes to go from C to D and now all
2,000 cars are going D to B yes so
therefore they will take 10 plus 10 plus
20 that’s 40 minutes that is five
minutes less than ACB so they will shift
now when they shift what happens there
are these people who are still going ad
B so now we come to phase two so the ad
be people are saying 35 plus 20 I am
taking 55 minutes all because the
so-called migrants onto D and B have
entered our road and that is increasing
our time right not good but when you
can’t beat them you join them so what if
I were going a CD B so consider the
first person who shifts so I’m a to see
there are thousand cards now it becomes
thousand and one so the time taken
becomes ten point zero one plus ten that
is twenty point zero one plus 20 that’s
40 point 0 that’s a lot less than 35
plus 20 which is 55 so the first card
shifts and now suppose everybody shifts
then the deck 2010 20 so individually
each one is taking an action which is
good for herself but in the aggregate
everybody is damaging everybody else’s
happiness including the first personal
shifted and that’s the unintended
consequence that we have to keep in mind
when we are trying to discuss right how
we should intervene
this focus on get the data first let the
data tell the story is a must is a
necessity but please remember it is not
sufficient I will end with another
example which again you may all know the
reason why I’m giving these examples is
so that you can recall them right and
this appears in your class twelve
textbooks this is about the rat menace
in Vietnam where people try to get rid
of rats the French you know
corresponding to our governor-general
the French authorities at that time to
keep Saigon free of rats offered people
that if you bring in the tail of a rat
then I’ll give you an award so what were
people doing they were farming rats so
all that happened was that instead of
rats with taste you had rats without
tails running around Saigon and actually
the same thing happened in British India
when they tried to control the cobra
population the snake they offered reward
for bringing in the Cobra and so people
started forming cobras so this is what
you need to understand that unless you
can figure out how people are going to
respond to an intervention don’t even
[Applause] [Music]