00:00:00
ladies and gentlemen welcome to the kdi
00:00:02
school World Bank monthly webinar series
00:00:05
my name is yanji Kim from the kdi school
00:00:08
of public policy and management and I
00:00:10
will saler MC for today's
00:00:13
webinar before we get started please
00:00:16
note that today's webinar will be
00:00:18
recorded for knowledge sharing
00:00:20
purposes the recording will also be
00:00:22
uploaded later to the kdi school's
00:00:24
official YouTube
00:00:26
channel this webinar is our joint
00:00:29
venture with the world Bank where
00:00:31
experts from the bank will share their
00:00:33
recent findings with the school
00:00:35
Community we hope you gain exclusive
00:00:38
insights into The Cutting Edge research
00:00:40
by the bank and have time to reflect on
00:00:43
possible
00:00:44
applications during the presentation
00:00:47
please keep your microphone muted if you
00:00:50
have any questions or comments please
00:00:52
wait for the Q&A session or leave your
00:00:55
messages in the
00:00:57
chat now I'm pleased to introd U today's
00:01:00
distinguished speaker Mr RF
00:01:04
CH he is an economist in the economic
00:01:07
policy division of the prosperity unit
00:01:10
at the World Bank group his expertise
00:01:13
includes microeconomic modeling fiscal
00:01:16
policy and analyzing the macroeconomic
00:01:19
and fiscal impacts of climate
00:01:21
change since joining the bank in
00:01:24
2018 he has held concerting roles
00:01:27
contributing to macroeconomic and fiscal
00:01:30
policy analysis in
00:01:32
India he holds a master's degree in
00:01:34
quantitative economics from the Indian
00:01:37
Statistical Institute and bachelor's
00:01:40
degree in economics from H Raj College
00:01:43
Delhi
00:01:44
University everyone please join me in
00:01:47
welcoming Mr rishab
00:01:50
chower Mr chower please share your
00:01:53
slides when you're ready to
00:01:55
begin thank you for the introduction and
00:01:58
having me here very good morning to
00:02:01
everyone I'll share my slides
00:02:14
with please let me know whenever you can
00:02:16
see the
00:02:24
slides um I hope you can see the slides
00:02:28
by now yeah we can see the a s okay
00:02:31
thank you so
00:02:34
uh so this uh work today I'm going to
00:02:37
present uh has been co-authored with my
00:02:39
colleagues zilia and France uh as the
00:02:41
title suggest taxing for growth
00:02:43
revisiting the 15%
00:02:46
threshold uh in this uh paper uh we
00:02:49
revisit this relationship between
00:02:51
Taxation and growth and extend the
00:02:54
analysis to inclusive growth when I say
00:02:56
revisit it means we are building on the
00:02:58
existing work that has been there around
00:03:01
uh exploring this
00:03:03
relationship so I'll uh structure my
00:03:06
presentation in U four broad parts I'll
00:03:09
first start with the motivation around
00:03:12
this paper why we did this there's a
00:03:14
world Bank corporate
00:03:16
scorecard and we did some uh event study
00:03:19
analysis uh before proceeding with more
00:03:21
rigorous approaches uh then I will move
00:03:24
on to uh discussing the modeling uh work
00:03:27
done behind this paper which explores
00:03:29
the relationship between inclusive
00:03:30
growth and
00:03:32
Taxation eventually I'll cover the last
00:03:34
section where I'll explore uh the
00:03:36
possible channels which explains the
00:03:38
relationship we
00:03:41
observe so the first part uh the World
00:03:44
Bank corporate score card well it's an
00:03:47
uh accountability mechanism where uh the
00:03:50
World Bank tracks its uh progress on
00:03:53
development
00:03:54
priorities our president recently
00:03:56
reduced the size of the previous score
00:03:59
cards that has been there to one single
00:04:01
scorecard and uh not just that uh we
00:04:05
have earlier about 150 indicators now we
00:04:08
just have 22 indicators broadly covering
00:04:11
our development priorities on um poverty
00:04:15
climate change uh inclusive
00:04:18
growth um and finally uh on this uh we
00:04:22
have moved from uh more like
00:04:24
harmonization of these results
00:04:26
indicators with other multilateral
00:04:28
institutions which was one of the G20
00:04:32
priority
00:04:33
uh and uh now just to give you another
00:04:37
context within this uh
00:04:40
scorecard uh domestic Revenue
00:04:42
mobilization Still Remains uh one of the
00:04:45
components and domestic Revenue
00:04:48
mobilization is of course the country's
00:04:50
capability to mobilize revenues
00:04:53
domestically U raise fiscal revenues
00:04:57
domestically uh well there are three uh
00:05:00
elements broadly I would say first is uh
00:05:03
what are the vision indicators and then
00:05:07
what are the result indicators and
00:05:09
what's the narrative around those
00:05:10
results so giving a very quick
00:05:13
overview uh we start with looking at the
00:05:16
high level measurements of uh
00:05:18
development progress uh move on to some
00:05:20
aggregate results of uh interventions
00:05:23
and finally uh some narrative stories
00:05:25
around how those reforms were
00:05:27
implemented so uh within the this uh DRM
00:05:31
agenda I will specifically point out the
00:05:34
indicators how we track progress is uh
00:05:38
looking at the country's progress in
00:05:40
passing the tax to revenue 15% threshold
00:05:43
which uh we are going to revisit in this
00:05:46
paper and then of course looking at how
00:05:49
uh the low uh collection countries U
00:05:53
with the World Bank is helping increase
00:05:54
their revenue how they address their tax
00:05:56
Equity systems and lastly on the stories
00:05:59
we look at how the bank supported
00:06:01
reforms help in this uh help these
00:06:04
countries improve revenues and improve
00:06:06
other
00:06:07
outcomes this was a broad summary of the
00:06:10
corporate scorecard in a nutshell we
00:06:12
look at uh the countries with weak
00:06:14
Revenue capacities and see how the banks
00:06:17
reforms have helped them and improve
00:06:19
their development
00:06:22
outcomes now uh uh moving to another uh
00:06:26
before moving to the modeling approach
00:06:28
we do another m exercise behind this
00:06:32
paper which where we try to look at uh
00:06:34
the relationship between tax to GDP and
00:06:37
economic growth or development outcome
00:06:39
using an event
00:06:41
study so what is the objective the
00:06:44
objective here is very clear we want to
00:06:46
see whether that increase in the tax to
00:06:49
GDP ratio of a country preceded a
00:06:51
country's transition to a higher income
00:06:54
group so when countries moving from
00:06:57
lower income to lower upper middle
00:06:59
income or upper middle to high income do
00:07:03
these countries observe an increase in
00:07:05
tax or GDP ratio prior to those
00:07:07
transition
00:07:09
dates we use the World Bank income group
00:07:11
classification data from 1987 to
00:07:15
2021 uh a broad um stats some broad
00:07:18
stats are presented on the right side we
00:07:20
see that when countries move from low
00:07:22
income to lower middle income we have
00:07:25
about 53 countries in this sample period
00:07:28
likewise 67 when they move from lower
00:07:30
middle to upper middle and when they
00:07:32
move from upper middle to high we have
00:07:34
about 41 such
00:07:35
cases however there are limitations on
00:07:37
the tax to GDP rati not every country
00:07:40
during the time of transition had this
00:07:42
data available at least in public uh
00:07:44
domain so this reduces our sample size
00:07:47
so you can see about
00:07:49
27 uh countries have data when they move
00:07:51
from low to Low Middle income so we were
00:07:54
able to only track those in the event
00:07:57
study
00:08:00
and uh another thing maybe I can put
00:08:03
some context to the numbers so low
00:08:04
income class recent number suggest has a
00:08:07
classification of about
00:08:10
$1,145 gni per capita by uh Atlas
00:08:14
method and uh for Middle income the cut
00:08:17
off is like uh you cross 1145 and
00:08:20
between $145 to
00:08:22
$4,500 for upper middle income group the
00:08:25
income range is 4500 to 14,000 and
00:08:27
higher income is above that threshold
00:08:30
this is the recent one but it has
00:08:32
increased over time of course adjusting
00:08:34
for inflation and other
00:08:36
factors now uh coming to this uh chart
00:08:41
where we present findings from our event
00:08:43
study the leftmost chart looks at the
00:08:47
tax to GDP
00:08:49
ratios of the countries which were
00:08:52
transitioning from different income
00:08:54
groups and what we find is that in the
00:08:57
first stack
00:09:00
when countries transition from lower to
00:09:02
lower middle income category at the time
00:09:04
of transition their average tax to GDP
00:09:07
ratio was about
00:09:10
15% and the median which is sky blue
00:09:12
color is about
00:09:15
13% similarly we see that the tax to GDP
00:09:19
ratio from when countries transition
00:09:23
from Low Middle to Upper Middle was
00:09:25
about 23 and correspondingly it was 25%
00:09:28
for transition from upper middle to high
00:09:30
income group so we find that no
00:09:34
will 15% number when countries are
00:09:37
transitioning from this low to lower
00:09:38
middle income group so there's something
00:09:41
about that threshold that just a
00:09:43
motivation and in the right side chart
00:09:45
what we look at is we zoom in our focus
00:09:49
on those countries transitioning from
00:09:51
lower income to lower middle income
00:09:53
category what we see is that t denotes
00:09:56
here on the xaxis the period of
00:09:58
transition
00:10:00
and the solid Orange Line denotes the
00:10:02
median tax to GDP ratio the dash line
00:10:06
denotes the interquartile range of those
00:10:09
countries so what we see is that 10
00:10:12
years prior to the transition the median
00:10:14
ratio was 9% and it increased to almost
00:10:18
133% at the time of transition so the an
00:10:22
increase of about
00:10:24
4% which is interesting that uh
00:10:27
countries have increased their tax
00:10:28
capacity now this can be because of that
00:10:31
they were growing better and better
00:10:34
growth outcomes imply better tax revenue
00:10:36
collection their base was expanding they
00:10:38
have more efficient tax systems
00:10:41
Administration it could also mean the
00:10:43
other way around that you have higher
00:10:45
tax to GDP ratio means that you
00:10:48
are able to do more redistribution in
00:10:51
the economy reduce inequality and do
00:10:54
more uh growth oriented expenditures to
00:10:57
support growth so there relationship can
00:11:00
work either way so far we have just
00:11:02
established a more of a correlation kind
00:11:06
of framework not any caus about the
00:11:08
relationship between Taxation and growth
00:11:10
and what we have observed is that around
00:11:13
uh the transition uh date the tax to GDP
00:11:16
ratio average around
00:11:20
15% there are just some country examples
00:11:22
because so far I reported main median uh
00:11:25
now I just show some country examples
00:11:27
starting uh the top left one is Georgia
00:11:31
where the tax to GDP ratio increased
00:11:34
from 7.3 to 14 uh in the 10 years
00:11:37
preceding this uh income transition from
00:11:39
low to lower middle income uh group
00:11:42
likewise in codia it increased from 8 to
00:11:44
14.6 in India 8.9 to
00:11:48
11.9 uh in maldiv 7.7 to 12.8 so you see
00:11:52
broadly the number is of course not
00:11:53
exactly 15 but there's a range uh around
00:11:56
that number that we see the increase
00:11:58
happened
00:12:01
um and uh so this ends the event study
00:12:04
we have learned some lessons from the
00:12:06
event study that uh tax to DDP ratio
00:12:08
increased 3 to 4 percentage points prior
00:12:11
to the income
00:12:12
transition well it just motivates the
00:12:15
idea that uh there's something that
00:12:17
matters for growth that increase
00:12:19
maybe now we'll revisit this
00:12:22
relationship from a more technical
00:12:24
perspective to establish uh causality
00:12:29
and uh I will start with some motivation
00:12:33
uh again looking at some correlations uh
00:12:36
on the entire data set between tax to
00:12:39
GDP ratio and measures of growth so the
00:12:43
first thing I look at is on the leftmost
00:12:45
chart is whether the tax to GDP ratio
00:12:49
today is correlated with 10e ahead
00:12:53
growth rate GDP growth rate
00:12:56
normal we don't really observe any
00:12:58
strong POS relationship as you can see
00:13:00
in the leftmost chart but uh well the
00:13:03
red line is a predicted polinomial uh
00:13:06
fractional polinomial fit where we do
00:13:09
see that at least in the very initial
00:13:11
phases there is an increase and then of
00:13:13
course relationship is flat so there's
00:13:15
no such relationship that is coming
00:13:18
out now on the right chart I do the same
00:13:22
uh scatter but instead of growth I look
00:13:24
at Prosperity Gap Prosperity Gap is
00:13:27
coming from this create all paper which
00:13:30
is published by World Bank authors so
00:13:33
Prosperity Gap is a indicator which is
00:13:37
measuring how by how many times your
00:13:42
income uh of all the persons in an
00:13:45
economy must be multiplied by a certain
00:13:48
Factor so that you reach this $25 per
00:13:52
day income now that $25 per day is
00:13:57
coming from the fact that it is observed
00:13:59
at uh when the countryes transition from
00:14:03
uh upper middle income to high income
00:14:05
status so it's a
00:14:07
factor uh the lower factor means you are
00:14:10
better off the higher means that you
00:14:11
need to do more to improve the incomes
00:14:14
and have shared prosperity in the
00:14:15
economy so higher uh factor is a bad
00:14:19
outcome uh it means higher inequality
00:14:22
kind
00:14:23
of now what we see is that countries
00:14:26
with higher taxation to GDP ratio have
00:14:29
lower factor which means uh 10 year
00:14:32
ahead uh growth is more inclusive when
00:14:35
you have higher tax or GDP ratios again
00:14:38
these are
00:14:39
correlations but there is some
00:14:41
motivation that there is in relationship
00:14:43
and we can explore further by our
00:14:45
empirical
00:14:49
approach okay so this is our empirical
00:14:53
model now let's go step by
00:14:57
step the left hand side of this equation
00:15:02
denotes the change in the Y variable Y
00:15:06
is our either growth rate or why is why
00:15:10
can be seen seen as a GDP measure or
00:15:13
Prosperity Gap measure and when we look
00:15:15
at the difference of course in
00:15:16
logarithmic terms it becomes a growth
00:15:18
rate so yct denotes the value of
00:15:22
variable Y in for Country C in time
00:15:24
period t+ J and given our objective is
00:15:29
long-term growth impacts we look at J is
00:15:31
10 and 15 years old in our paper we look
00:15:33
how far A little bit far in
00:15:36
time and then we are saying that this
00:15:39
outcome variable depends
00:15:42
on whether your taxation tax to GDP
00:15:46
ratio which is tax of a country C in
00:15:49
time period T tax CT exceeds a threshold
00:15:53
camama so we are saying that if your tax
00:15:57
to GDP ratio exceeds the scamma then
00:16:01
your growth or the outcome variable
00:16:03
which you are interested in on the left
00:16:05
hand side that is higher by an amount
00:16:08
beta 1 for those countries and time
00:16:12
periods when your tax to GDP ratio is
00:16:15
higher than this threshold
00:16:17
level and then there are other factors
00:16:20
as we point out we have tax to DDP
00:16:23
itself which is capturing that higher
00:16:26
taxation is associated with higher
00:16:28
growth the beta 3 coefficient you see
00:16:30
here is capturing that we also control
00:16:33
for a bunch of other factors which are
00:16:35
in
00:16:36
xct X will include factors like uh other
00:16:40
Revenue sources debt to GDP ratio
00:16:43
investment to GDP ratio population
00:16:47
growth and then we have certain Capac C
00:16:50
which denotes country specific
00:16:53
characteristics uh and Capa which you
00:16:55
know time fixed effects for instance in
00:16:57
certain period there were a crisis we
00:16:59
want to control for the fact that uh the
00:17:01
relationship is not affected by those
00:17:04
specific time
00:17:06
periods and then the first is's a error
00:17:08
term around your
00:17:11
estimates but here the challenge is that
00:17:15
we don't know gamma to estimate this
00:17:18
beta 1 we need to know what threshold
00:17:20
are we putting in the model but before I
00:17:23
go there I would like to illustrate this
00:17:26
uh impact in a diagram
00:17:30
so here it's a hypothetical illustration
00:17:33
on the x-axis we have tax to GDP
00:17:36
ratio on the y axis we have uh the
00:17:40
outcome variable of interest which is
00:17:42
let's say
00:17:44
growth the blue line
00:17:47
is for those countries which are it is
00:17:50
for the country when it is below the
00:17:52
threshold and so this is just like a one
00:17:54
country example and then Red Line
00:17:57
denotes the outcome when it is above the
00:17:59
threshold hypothetically we just assume
00:18:02
here that let's say the threshold is
00:18:04
15% so what we are expecting we are
00:18:07
expecting that this outcome variable
00:18:10
will jump around this 15% threshold so
00:18:13
this beta 1 is where you will see that
00:18:18
impact this is the interpretation of
00:18:20
beta 1 you expect a jump around that
00:18:23
threshold which is captured by beta 1
00:18:27
but we don't know the threshold so what
00:18:30
we are going to do is as I mentioned in
00:18:32
the previous slide we will do a grid
00:18:34
search we will run this model for
00:18:37
different different values of gamma
00:18:39
between 7 to 30% of
00:18:42
GDP and sub increase it by a marginal
00:18:45
amount let's say 0.12% of GDP and look
00:18:49
at the values of beta 1 this beta 1 AC
00:18:53
for different values of those uh tax to
00:18:55
GDP ratio and eventually we we will see
00:18:59
where this value is
00:19:01
maximized that is where we will observe
00:19:03
that the threshold has the maximum
00:19:06
effect
00:19:08
so in this
00:19:10
Slide the left hand chart is precisely
00:19:13
doing that what we see is that tax to
00:19:16
DDP ratio is on the x- axis on the y-
00:19:19
axis we have cumulative growth rate so
00:19:21
that's why you see the numbers are
00:19:22
higher because it's a 10-year growth
00:19:24
rate uh
00:19:25
cumulative and uh what you see is
00:19:30
that the the beta one which is capturing
00:19:35
this cumulative growth rate it takes a
00:19:37
maximum value somewhere closer to
00:19:41
133% first and foremost point to note is
00:19:44
that beta 1 here is positive and
00:19:47
statistically significant which is shown
00:19:49
by this shaded band of 90% confidence
00:19:53
interval which means that the effect is
00:19:57
there there is a change there is a jump
00:19:59
but where that jump is maximum now we
00:20:02
search that
00:20:04
point which appears to be closer to 13%
00:20:06
in our uh
00:20:08
chart similarly our goal is due to the
00:20:12
similar analysis for Prosperity Gap but
00:20:16
with prosperity Gap the interpretation
00:20:18
is the opposite because higher
00:20:19
Prosperity Gap would mean that you are
00:20:23
you need to there's more inequality you
00:20:25
need to multiply your incomes by larger
00:20:27
factor to reach that 25 $5 threshold
00:20:31
so here we are interested in minimizing
00:20:34
that beta 1
00:20:36
coefficient that's why in the right
00:20:38
chart what we observe is that the
00:20:40
minimum value of that beta 1 over this
00:20:44
range is achieved again around 133%
00:20:48
threshold now here I have presented and
00:20:51
another thing uh I would like to mention
00:20:53
before that it is statistically
00:20:54
significant the negative number so the
00:20:57
impact is still there there is a decline
00:20:59
in the prosperity Gap indicator around
00:21:02
that
00:21:03
threshold so here I presented the case
00:21:06
for 10 year ahead growth rates and 10
00:21:09
year ahead Prosperity Gap in the
00:21:11
original paper we do have this uh 15
00:21:15
years ahead as well and the range is
00:21:17
broadly the same 12 to 14% for the
00:21:19
threshold for both the
00:21:24
cases now in this slide
00:21:30
we are doing some predictions based on
00:21:32
our model so our goal is to look at
00:21:36
whether when countries move from the
00:21:38
lower growth rate uh lower tax GDP ratio
00:21:42
of let's say 7% to 15% the threshold
00:21:46
which we are vouching for to the what's
00:21:49
the predicted Inc predicted value of the
00:21:51
cumulative roow three so the left track
00:21:54
clearly shows that there is a higher
00:21:58
growth when you have a higher tax DDP
00:22:00
ratio and this is predicted value from
00:22:02
the model where we control for other
00:22:04
factors as
00:22:06
well the orange whiskers are showing the
00:22:09
90% confidence interval around that
00:22:12
growth likewise in the right hand chart
00:22:15
we see that if we look at Prosperity Gap
00:22:17
as an indicator we see that Prosperity
00:22:20
Gap is lower when your tax to DDP ratio
00:22:22
increases from 7 to 15 percentage
00:22:25
points so the impact are visible even in
00:22:29
a more sophisticated model than just
00:22:32
Scatter
00:22:35
Plots so so far I have touched upon the
00:22:39
motivation of
00:22:42
why uh firstly why it matters and the
00:22:45
corporate scorecard the event study and
00:22:48
eventually I moved on to a empirical
00:22:50
model where we established the fact that
00:22:53
tax to GDP ratio indeed
00:22:56
U uh is relevant for explaining higher
00:23:00
growth patterns and the threshold was
00:23:03
broadly observed around 133% not exactly
00:23:06
15 but
00:23:07
13 and that is true for both the cases
00:23:10
for uh GDP growth and for inclusive
00:23:13
growth measure which is captured through
00:23:14
Prosperity
00:23:15
Gap in the next section we will explore
00:23:19
what explains this increase like why
00:23:22
does a higher tax would mean higher
00:23:24
growth or lower
00:23:25
Prosperity so now using our similar
00:23:28
model we are going to explore some
00:23:30
channels well using conventional
00:23:32
knowledge we know that uh some channels
00:23:35
obviously that might appear to our mind
00:23:37
is that these countries which have
00:23:39
higher taxation revenues they are they
00:23:41
are able to better spend on health and
00:23:44
education which is crucial for growth
00:23:47
particularly in low-income
00:23:49
countries also these countries with
00:23:51
higher taxation might be able to
00:23:53
redistribute those revenu so there is
00:23:56
this progressivity aspect I'll explain
00:23:58
this definition as I go
00:24:00
ahead likewise uh these countries where
00:24:04
higher tax revenues are collected you
00:24:07
can probably have lower volatility in
00:24:09
your government spending patterns
00:24:10
because you're not really influenced
00:24:13
when the shocks hit the economy and
00:24:14
hence you are able to better stabilize
00:24:16
the economy
00:24:18
so uh you are likely to take some
00:24:20
counter cyclical policy measures which
00:24:22
will improve your ability to respond to
00:24:25
shocks and hence improve the growth
00:24:27
outcomes
00:24:29
so these are possible channels which we
00:24:31
have explored in this
00:24:33
paper uh let's start with health and
00:24:37
education so in the leftmost
00:24:41
chart we plot tax to GDP
00:24:44
ratio and average future health and
00:24:47
education spending as a percentage of
00:24:50
GDP now these scatters are slightly not
00:24:53
not the same as previously shown like
00:24:56
it's not just a simple country year
00:24:58
scatter but what we are doing is we
00:25:01
create bins
00:25:03
so we have this tax to DDP ratio
00:25:07
associated with one particular range uh
00:25:09
in a very small vicinity let's say 0 to
00:25:11
2% we take the average of all those
00:25:15
countries in that range average of
00:25:16
health and education spending of all
00:25:18
those countries in that range and plot
00:25:20
that this makes the visualization
00:25:22
relatively uh
00:25:25
simpler so what we see is that again as
00:25:28
your tax to GDP ratios are rising You
00:25:30
observe an increase in health and
00:25:32
education
00:25:34
spending and there's no threshold L such
00:25:36
it keeps on increasing you have more and
00:25:38
more fiscal space you're spending
00:25:41
more and we see likewise that uh when we
00:25:44
do this with a model controlling for
00:25:47
fact other factors the predicted values
00:25:49
are increasing although there is some
00:25:51
tapering of when we do this model
00:25:52
approach that the increase is not that
00:25:54
large like from 7 to 15 the increase was
00:25:56
large then it is moderating so there are
00:25:59
limits of course to the public health
00:26:00
and education
00:26:02
spending and there is literature which
00:26:05
suggests that uh and there are growth
00:26:08
models which suggest health and
00:26:09
education spending particularly are
00:26:10
relevant for human capital development
00:26:12
and hence they have positive impact on
00:26:14
growth of course in some cases there are
00:26:17
uh mixed
00:26:18
findings but broadly this growth Theory
00:26:21
will suggest that there are positive
00:26:23
effects on
00:26:24
growth and that's why we believe higher
00:26:27
taxation would increase a country's
00:26:30
fiscal space
00:26:31
to increase those critical spendings and
00:26:35
hence growth effects is
00:26:38
materialized now coming to the next
00:26:40
Channel progressive taxation so uh some
00:26:44
of you may already be aware of this
00:26:46
concept but just touching uh very
00:26:49
briefly so a progressive tax is the one
00:26:52
where this average tax burden will
00:26:55
increase as your income will go up now
00:26:57
it is not that simple that higher income
00:27:00
will of course have that higher taxation
00:27:04
because just because their income is
00:27:06
higher and they'll pay a larger
00:27:07
proportion it's not just that the tax
00:27:10
rate itself has to be higher so when you
00:27:14
make the higher income individual pay a
00:27:16
higher tax rate as you can see in the
00:27:18
left uh in the right uh chart where this
00:27:21
example is
00:27:23
presented so we see that as your income
00:27:25
is going up the tax rates are increasing
00:27:27
now this makes the system more
00:27:30
Progressive the definition might be very
00:27:32
obvious for some of you but just thought
00:27:34
to start with
00:27:38
this now why does it matter so likewise
00:27:44
using the similar Bend scatter plot we
00:27:47
show that as the direct
00:27:50
taxation in the economy increases the
00:27:53
proportion of direct
00:27:55
taxation You observe that the pro in the
00:27:58
leftmost chart the Y AIS is prosperity
00:28:01
Gap this Factor Prosperity Gap it is
00:28:04
declining which means when you have
00:28:06
higher direct
00:28:08
taxation you do observe that your
00:28:11
inequality is
00:28:15
lower now we test this in again the
00:28:19
similar model setup and do the
00:28:21
prediction controlling for other
00:28:23
variables we do observe that somewhere
00:28:26
around 15% this effect is there now 50%
00:28:29
is this Direct Tax to
00:28:31
GDP U share of direct taxes in total
00:28:35
taxes it's not the tax to GDP ratio so
00:28:39
uh it's different than the previously
00:28:41
explained Concepts so what we see is
00:28:43
that you have higher proportion of
00:28:46
direct
00:28:47
taxes um as high as
00:28:49
50% then you are going to see likely the
00:28:52
progressive impacts
00:28:54
of uh your tax system
00:28:58
which means a reduced poverty uh
00:29:00
Prosperity gap which is capturing the
00:29:02
inclusive
00:29:03
growth now also we observed of course
00:29:06
that higher taxation was associated with
00:29:08
higher Direct Tax shares now that is why
00:29:11
we believe that that is one channel
00:29:13
through which we see the prosperity Gap
00:29:16
and is declining and we observe more
00:29:18
inclusive
00:29:22
growth next Chanel is volatility now
00:29:26
firstly the why and how well there is
00:29:29
literature which suggests that higher
00:29:31
volatility can negatively impact
00:29:35
growth uh there are two aspects here one
00:29:38
is of course we look at government
00:29:40
spending volatility and the economic
00:29:43
volatility the volatility and growth
00:29:44
rate itself well government spending is
00:29:47
a component of total GDP so higher
00:29:49
spending uh volatility would mean higher
00:29:52
growth
00:29:54
volatility but not just that even the
00:29:57
growth volatility itself can have a
00:29:59
negative impact on
00:30:00
growth well some literature suggest that
00:30:03
there could be some planning error by
00:30:05
the firms because you are living in a
00:30:07
more uncertain environment and
00:30:08
uncertainty can lead to lower
00:30:11
investments in the economy and hence
00:30:14
influence future
00:30:18
growth again for government spending uh
00:30:22
these economies uh with lower revenues
00:30:26
are premises that they will likely have
00:30:28
more volatile spending which will leave
00:30:31
them with lesser opportunity to borrow
00:30:35
undermining their ability to stabilize
00:30:37
the business cycle and implement the
00:30:38
counter cyclical policy measures when
00:30:40
they are needed particularly in low
00:30:42
growth time periods or volatile physes
00:30:44
of the business
00:30:45
[Music]
00:30:47
cycle so that's clear that there we
00:30:50
expect at least some impact through our
00:30:53
literature survey and theoretical
00:30:56
knowledge but how do we measure this
00:30:59
volatility we have so far used all the
00:31:02
annual taxation data but now to measure
00:31:04
volatility we use quarterly data on GDP
00:31:07
growth and government consumption in
00:31:09
real terms there seasonality we adjust
00:31:12
for the seasonal
00:31:14
factors then we look at the 10 year
00:31:17
ahead standard deviation of this quarter
00:31:19
on quarter growth rate in GDP and
00:31:23
government
00:31:25
consumption well 10 year ahead is just
00:31:27
to
00:31:28
make sure that we are not looking at any
00:31:30
contemporaneous relation and Taxation
00:31:32
today matters for future this also
00:31:34
partially takes care of the endogenity
00:31:37
issue where you might say that it's the
00:31:39
other way around that lower volatility
00:31:41
means better taxation but that's why we
00:31:44
look at 10 year ahead so that tax the
00:31:47
relationship goes from taxation to
00:31:50
volatility and then of course now we
00:31:52
want to link it to our annual data set
00:31:54
so we just pick up the first quarter
00:31:55
measure and Link call it
00:31:58
representative for that particular year
00:32:00
because it's 10 year ahead so it covers
00:32:02
the future
00:32:04
periods so then uh this is our measure
00:32:07
of volatility for both spending and
00:32:09
government spending and economic
00:32:11
growth again similar bin plot suggests
00:32:14
that in the left most chart higher
00:32:17
taxation is associated with lower
00:32:19
government spending volatility you see
00:32:22
it's declining and here one can roughly
00:32:25
say that okay up to 20% the impact is
00:32:27
sharp it's sharp Decline and then onward
00:32:29
starts to taper
00:32:31
off in the rightmost start we present
00:32:35
this commment consumption volatility
00:32:37
against
00:32:39
um future GDP volatility and see of
00:32:42
course GDP was uh government consumption
00:32:45
is a component of GDP and there's a
00:32:47
positive association between the two
00:32:49
higher consumption volatility means that
00:32:51
higher GDP
00:32:55
volatility so there is some evidence
00:32:57
that spending and output volatility with
00:33:00
drop with higher tax revenue collections
00:33:03
which is what we show now in a empirical
00:33:06
model so use in the left most chart uses
00:33:09
the same threshold model which we
00:33:12
presented in the very uh beginning slide
00:33:16
in the third
00:33:17
section what we see is that the beta 1
00:33:20
Co now we have replaced the Y by
00:33:23
volatility measure instead of growth and
00:33:25
inclusive growth what we see is that the
00:33:27
volatil
00:33:28
is minimized at around again 133% of
00:33:33
GDP so and it is negative and
00:33:36
statistically significant impact because
00:33:39
those orange whiskers are 90% confidence
00:33:44
interval so again there's motivation
00:33:46
that this channel is at work around that
00:33:50
threshold and the right hand chart is
00:33:52
just doing the same approach we look at
00:33:54
the predicted values and we see that as
00:33:56
you come uh closer to 13% there's a
00:33:59
sharp decline but afterwards it's more
00:34:02
or less flat the volatility is more or
00:34:04
less the same for countries in that
00:34:06
higher bracket but as you approach 13
00:34:08
then there's a
00:34:13
decline okay so this is my last slide on
00:34:18
conclusion what we have donear today is
00:34:21
that from this paper that government uh
00:34:25
when they collect lower tax revenues we
00:34:27
have observed that there are negative
00:34:29
impacts on growth because their capacity
00:34:32
to invest in critical infrastructure can
00:34:34
be limited which we have not really
00:34:36
shown on the infrastructure side but at
00:34:38
least on the development side human uh
00:34:41
Capital spending and U public spending
00:34:44
on health and education we have seen
00:34:46
that that is that can be limited because
00:34:48
of tax uh capacity constraints which can
00:34:52
further exacerbate inequality and lead
00:34:54
to overreliance on debt now higher debt
00:34:57
can also of course have negative impacts
00:34:58
on growth uh if you just keep on relying
00:35:01
on debt and not on your uh domestic uh
00:35:04
capacity the cut off we found uh is in
00:35:08
the range of 12 to
00:35:09
14% well now you will say why then 15%
00:35:13
the reason is we observed that about
00:35:15
there is the the standard deviation of
00:35:18
this tax to GDP ratio was again around
00:35:20
2% over our sample so then when we are
00:35:25
setting targets for the country it's
00:35:26
better to be at 15% because there'll be
00:35:28
volatility around that number so you aim
00:35:31
a slightly higher
00:35:33
number and that stabilizes your future
00:35:36
fiscal outcomes and the channels we have
00:35:38
explored it was through more productive
00:35:40
spending more Progressive taxes less
00:35:44
spending and output
00:35:45
volatility which further translated into
00:35:49
better growth
00:35:51
outs thank you
