THE 2021 FESTIVAL OF SOCIAL SCIENCE RUNNING 1-30 NOVEMBER 2021
Inequality: what do we know, and why do we care?
What’s on offer?
Each year early career economists at IFS deliver a day of public economics talks, aimed at A-level and undergraduate students who have an interest in economics or might want to pursue a career in public policy research. As part of this year’s Festival of Social Science, we will be live streaming a selection of lectures from the series to anyone who is interested in finding out more about economics. There will be time allowed for Q&A on each topic.
What’s it about?
Why do we care about inequality? What should be done about it? This lecture explores the debate over which inequalities the government should act to reduce, and why. It focuses on the contributions that can be made by economic theory and empirical research, and also introduces some of the tools used to measure different inequalities, applying them to UK income inequality to show how different measures can give us different insights.
Who’s leading the event?
Tom Wernham, Research Economist at IFS
Open to
These webinars are primarily aimed at final year undergraduates studying economics, but should be useful to anyone interested in the subject.
hi good afternoon everyone and welcome
to today's event inequality what do we
know and why do we care
i'm shari shu and i'm a senior research
economist at the ifs and i'll be
chairing today's session
this is um the second in a series of
lectures on public economics which will
run at the same time each week so every
monday from four to five o'clock
next week we'll be looking at the
effects of graduating into a recession
so if you're interested please do
register for that as well and you can
find last week's lecture and pensions on
our website and on my youtube channel
so what will happen today is that my
colleague tom wernham will talk to you
about inequality why we as economists
might care about inequality what policy
can do
what's happened to inequality in the uk
and how different measures of inequality
can tell a different story
tom will talk about for about 45 minutes
and then we'll have 15 minutes for
questions and answers
we'll be taking questions on slider so
please post your questions there and
vote up the questions you'd like to see
answered you can find the slider link on
the website
so without further ado i'm going to pass
on to tom
thank you very much charlie
so as i shall i said this lecture is
going to be on inequality what do we
know and why do we care
and so just give you a brief outline of
what we're going to talk about in part
one we're going to think about what it
is we mean by inequality and we're going
to see that people are interested in
inequalities in a range of different
outcomes between different units of
comparison
i'll give a couple of examples as well
and then in part two we're going to ask
why it is that we care about inequality
specifically what arguments are made for
and against policy intervention in
different qualities and what can
economic research and economic theory
contribute to that debate
and then uh in part three
one of the major contributions
economists can make is contributing to a
measurement of inequality so we're going
to think about how it is we measure
inequality um
and what can different measures tell us
we're going to see that many different
measures and different measures give us
different insights into how inequalities
change
and i think the overall message to take
from this lecture is that inequality is
a really complex topic um
different inequalities have changed in
different ways and
different arguments for and against
intervention
may
apply to some inequalities and not to
others
so let's
jump straight to why
what it is we mean by inequality
as i said people are interested in
plotting a range of different outcomes
so
one of those quite a popular one is
earnings for example the attention given
to the uh gender pay gap or to pay
differentials between
in large firms
on the other hand we might be interested
in disposable income so that's income
from earnings but also other sources
like investments benefits etc and the
taxes taken away and that's because this
take-home disposable income is often
seen as a proxy for living standards
another thing we're often interested in
is is wealth inequality and that's
because that wealth in a good is a
significant a key mechanism by which
other inequalities can be transmitted
through generations
um it's not just these material
inequalities we're interested in though
uh i think with the pandemic in
particular inequalities in health
outcomes or well-being have come into
focus
um and another interesting inequality
which uh gets a lot of attention is
inequality of opportunity and there
might be one strand of thought some
people may argue that inequalities of
outcome don't matter so much as long as
people had equal opportunities in the
first place to achieve those different
outcomes
and of course there are there are many
more uh inequalities that we might be
interested in
but there's also different ways of
different units of comparison when
thinking about these different outcomes
so we might be interested in between
grouping quantities
between men and women or people from
different ethnic backgrounds or
different towns and cities
on the other hand we might be looking at
inequality across the population so so
called interpersonal or into our school
properties
which
raises the question of what populations
we can look at inequalities just in the
uk for example or in the developed world
or in the whole world we get very
different answers depending on what
populations
uh
just a couple of examples to motivate
the rest of the discussion um so start
with a
health one given um the attention that's
received recently so it was the
beginning of the pandemic black people
in the uk were over four times more
likely to die from pronounced than white
people when adjusting or age
other inequalities that have received
potentially pandemic are educational uh
or opportunity
so for example at the beginning of 2020
nine percent of children um in the uk
did not have access to a computer at
home according to offcom
which is
really important for
um equality of opportunity given the
fact that during the first lockdown uh
schools were shut and much education
provision was over the internet
and then an example of the inequality
before the pandemic
so the wealthiest 10 in the uk um and
almost 45 percent of the total world so
not far off half
and i think it's fair to say that
inequality is
a really controversial topic thinking
about some of the previous elections has
been a key dividing line between the
parties and
different inequalities
provide different reactions from people
some some are seen as problematic some
odds and there's also huge disagreement
about what should be done about
them
so
with that in mind let's move on to the
next section of the lecture where we're
going to think about
arguments for one against intervening in
different inequalities and what really
is that
people think is wrong with with with
inequalities um now this is a really
question of what's wrong with inequality
really complicated uh philosophical and
moral question and if you want a good
overview of these different moral
arguments i'd really recommend just
recently by don't process
what is wrong with inequality um it's
impossible to do justice the whole
debate here but i'm going to give a bit
of an overview to some of the key
arguments but what i'm also going to try
and do is bring that back to the
contribution that economic research and
economic theory can make to the debate
so
to start with let's actually think about
why we might not want to reduce
inequality or some inequalities
um
so it's not
obvious it shouldn't be taken for given
that all inequalities are a bad thing
that all inequalities are very
intervention so what are some of these
arguments that i can make well
on the one hand it might be argued that
some inequalities are actually
fair
or meritocratic and
they're just not a problem
so
this is the idea that
people who make unequal contributions
perhaps it should be awarded unequally
and those inequalities that are awesome
that aren't problematic a very simple
case would be um
that
some people choose to work more than
others um even imagine a world where
everyone had the opportunity to earn the
same hourly wage simply because people
have different preferences and some
people prefer
more leisure some people prefer more
consumption and would choose to work
more hours we would still see
inequalities in earnings rising and
maybe that isn't so problematic um
but on the other hand it gets more
controversial when we start to think
about how fair inequalities are uh than
aren't food choice
um because people can't work because
they've got to look after children not
because
you know highly skilled people are paid
much more than uh lower skilled people
and
whilst you know some empirical
contribution can be made um
to that debate looking at where
like whatever inequalities are through
earnings or inheritances and so on this
is inherently a really difficult moral
and
moral question the question of fairness
whether differences in pay we see are
actually fair
on the other hand another sort of
argument while we might not want to
reduce
inequality may accept that
some inequalities are problematic in
some way
but also warned that
measures we might want to take to reduce
them would also have their own negative
consequences
for example
it might be accepted that
some of the large differences in pay
between high-skilled workers and
low-skilled workers
aren't fair necessarily um
but if you wanted to redistribute
incomes you'd have to
finance that through taxes and that
could start to you know drive a wedge
between the wages paid by firms received
by individuals and causing their
inefficiencies um
so
you know the idea is that in a perfectly
competitive equilibrium
all the wages and all the prices are
such that they're diverting numbers of
people to do the right jobs buy the
right things so we get market clearing
outcomes um
start to drive wedges between that and
it causes inefficiency and this is where
this idea of an equity efficiency
trade-off
comes in so perhaps inequality is unfair
but to reduce their unfairness you need
to reduce efficiency
similar arguments might be that some
inequalities motivate innovation or
entrepreneurship and to reduce weight
you know stifle growth
um
now the thing is this this trade-off
will of course exist in some cases but
it's not a given that it will always
exist it's quite possible that some
inequalities might either have been
caused by inefficiencies in the market
or actually go on to cause
inefficiencies in the market we'll talk
about this a bit later but in those
cases it's
possible that we could both increase
um
both reduce inequality and also increase
market efficiency and
answering the question of when that's
possible that is a key contribution that
economists can make as we'll see
so let's move on to
reasons why people do care about
inequality why the government maybe
should
intervene to reduce certain quantities
and i think there's two main categories
of argument if you like um one is that
um some inequalities
are undesirable which should be reduced
because they arise from causes that are
in some way problematic unfair unjust
and so on
on the other hand some inequalities may
warrant intervention because they have
consequences
that are deemed to be undesirable or
problematic in some way and i'm going to
go through some examples of each of
these in turn uh to make a bit clearer
so let's start with one of the causes of
some inequalities which might be deemed
to warrant intervention and that's
historical injustice so this is the idea
that
where inequalities have arisen due to
discrimination or colonialism or
exploitation
that is a reason why it might be
imperative for the government to
intervene and produce so as an example
that satsang might give they argue that
uh wealth inequality in the us today is
to a large extent between equality and
homeownership
and we see significant differences
between
ethnicities there precisely because in
the 20th century african americans
suffered discrimination in access to
housing market and to credit
and so on
now
in cases such as these where
inequalities arise directly from past
injustice
it
probably seems clear-cut that there's a
problem there but even in this case uh
the question of what to do about them
can be uh controversial um
we can think about some policy responses
that might be proposed to these sorts of
inequalities um
uh redistribution uh in the country or
reparations uh if we're thinking about
um dealing with the
impact of colonialism in the past or
things like positive discrimination to
try and
you know equalize the
opportunities where people come from
disadvantaged backgrounds and all these
sorts of policies are um
quite controversial i think part of that
does come down to this idea of equity
efficiency trade-off again if you're
going to finance redistribution or if
you're going to have low entry
requirements people from certain
disadvantaged backgrounds that might
introduce um inefficiencies
there's also another
reason why this is controversial is
because given these injustices are
historical the people who might need to
finance their reparations of
rectification of these injustices aren't
necessarily people who committed them in
the first place
um
and then we move on to another another
cause of uh inequality which might be
into our intervention that's market
failure and a few ways in which market
failure might um
cause equalities
one is market power so
here we have the idea that um
some
large firms um
the seller side might be able to um
charge a markup over their costs because
they have market power they don't face
competition and make excess profits
and these excess profits might
contribute to inequalities in income and
wealth because
um they accrue to the owners and owners
of ownership of businesses is not evenly
spread throughout the existing income
wealth distribution
it's also possible some of these excess
profits will go to
paying the high salaries of
ceos or highly skilled workers but
perhaps not to low-skilled workers and
the reference there that denied out
does find evidence of increasing market
power
in the us and globally which might be
one explanation as to why we've seen uh
international trends of increasing
income inequality
on the other hand we might see market
power on the
uh in the labour market on the um buyer
site so
um monopsony power in the labor market
the economic theory tells us that if you
have a small number of firms hiring
labor and there's they don't face much
competition with other firms they may
have incentives to uh reduce wages and
reduce employment in order to increase
their profits and that may also
contribute to inequality in incomes as
you know
workers have their wages pushed down
towards the bottom of the distribution
um in order to increase the profits of
their company owners
and the reference i've given their
ability to um
look at minoxidil power in the uk labour
market and they do find evidence of
nozzle power reducing wages and
crucially though it looks like as if the
fed is confined to workers who aren't
part of collective bargaining
arrangements such as
trade unions
which which makes sense from
economically perspective
now other sorts of market failures that
might
be into problematic but also cause
inequalities are
asymmetric information which means some
risks such as unemployment risks or
health shocks cannot be insured um
privately um
and therefore in the absence of state
provision of insurance of these risks
they won't necessarily cause some degree
of inequality
now in these cases where we think that
inequalities are caused by these market
inefficiencies
that may lead us naturally to policy
responses that target the underlying
market affections rather than perhaps
you know some simple redistribution so
we might look at um
competition law firm controlling
monopolies minimum wages was in
collective bargaining arrangements or
provision of
health care unemployment insurance which
we see quite widely um
on the other hand if those measures
aren't able to go far enough perhaps
they can't be implemented or perhaps
we're worried about inequalities that
have been already
been caused by past market failures
then we still might want to look at
redistribution in which case we're back
to this equity efficiency trade-off
so that's
some of the causes of inequality that
some might feel more intervention but
now let's think about some of the
consequences um
now i think one of the simplest
arguments could be made here that could
be made here is that inequality is
somehow directly harmful to
people's welfare and there's a few ways
we can think about this
one might be that people simply have you
know pure preferences for equality they
prefer a world in which there is more
equality rather than the world in which
there is less
so i think in this sort of case we can
think about inequality
essentially entering the utility
function so i've put an example here
that inequality a version utility
function from erin smith this is the
utility function representing the
preferences of agent i
and if we imagine the x is our income so
we can see that agent i has
preferences over their own income which
is natural they want more income
but they also have preferences
over the incomes of the other
agents j the other in minus one agents
in this case
so we can see here there is a
um
term here procedured by the alpha
coefficient this is disutility that the
agent i
uh gets whenever other agents have more
income
than them you could potentially think of
it as an envy term in the utility
function
on your home this second
term uh preceded by the beta this is
this utility that agent i gets when um
other agents have less income than
or maybe these two terms together
represent
agenda's preferences for
social justice and so or equality more
you know generally
um because this is there's only one
specific example in inequality you seem
to function in many ways um
but if we want to evaluate whether this
is true this is the sort of thing that
can be researched um could be done
through polling simply asking people
about their preferences on inequality or
you could have um
behavioral experiments and i've got an
example of paper by stephanie
which
reviews some of these sorts of
experiments looking at how people's
preferences for inequality depend on
their incomes or information they have
and so on so this is the sort of thing
that's
uh
testimony
um on the other hand there may be a more
deeper and
more complicated way which inequality
can be harmful for welfare and
one way i think thinking about this is
um
the
capabilities approach to relative
poverty
popularized by a much descendant not
further than others now this is the idea
that
in order to achieve a given
absolute
standard of well-being but the resources
you need to do that depend on the
resources held by the rest of the
population
for example
we can think about
social standards
to
social expectations to
furnish your home in a certain way or
decorate it or
own a certain
model of car and so on um
or we could think about maybe
the
rise in
ownership of um computers and phones and
tablets and so on um
bring us back to our example earlier as
well so if
we
um
if you live in a world where the
majority have access to computers and
society will start to be geared towards
them and those who can't afford um
computers will be will be left behind we
saw that potentially with provision of
education over the internet during
during the first lockdown addis which
you know would be the expense of people
who couldn't afford it in these cases um
the the the people at the bottom of the
distribution who aren't able to afford
um
to meet certain social standards social
expectations um are worse often in a
world where nobody could afford these
things possibly um because they're being
excluded
um
so that's that's the other way in which
inequality may be harmful to welfare um
now i think in both these cases we can
actually
think of inequality as an externality if
i'm going for a high-paying job i'm not
necessarily going to think about
the impact that will have on patterns of
inequality and how that may affect other
people's welfare um so there's an
external effect of my actions that i'm
not considering um and clearly we
couldn't have a private market for
reducing inequalities so this is another
potential argument uh for intervention
and might be a way in which we can
increase uh fairness and efficiency if
we think that
and then i'll just briefly mention a
couple of other things um
another consequence of inequality that
may mean we want to intervene is that it
may harm the democratic process for
example or indirect effect um
so it might be that
you need you know rich people find it
easier to stand for election or that
donations give disproportionate
influence to some people in the
policy-making process
um
and these impacts of the inequality on
the democratic process
may be enough for some people
to warrant intervention to reduce them
another argument and it's mobile
circular argument is that inequalities
today might cause more inequalities
tomorrow
inequalities in income may cause
inequalities in wealth and future
generations um but also um i think it's
an interesting point today's inequality
of outcome could lead to tomorrow's
inequality of opportunity and this is a
a interesting come back to this idea
earlier that it's an equality
opportunity that matters it's hard to
imagine a sustainable
system over time where each generation
had
equality of opportunity and yet there
were vast differences in the outcomes um
because
clearly you know if another receptionist
decides you're accumulating wealth
that's going to influence
the opportunity of the next generation
so
that's the end of my summary of these
arguments um as you can see
so hugely complex uh debates to be had
around inequality and many of them hinge
on you know normal values and judgments
about what is fair and so on
however i think that almost all of these
arguments also
rest on factual claims about how the
world works so about where inequalities
come from or what their consequences are
or
uh how effective policy measures might
be at reducing them
and it's in these
um empirical questions that i think
economists you know have a real
opportunity to contribute to the debate
um
yeah that won't be able to provide us
with an answer to the questions but it
will
allow us to
you know
judge the different arguments on their
lyrics and um
yeah
be avoiding the
focus the argument on the fairness
questions if we can establish the facts
so yeah that's the contribution i think
can be most effectively made by
economists and
in that vein we're going to look at
one of those contributions which is
actually to measure inequality
um we need to measure inequality um
before we can work out
or not um
and i'm gonna start by i'm gonna
introduce a range of methods that
economists and statisticians might
employ to measure inequality in
different sorts of outcomes
so first um but not for long we're going
to think about between group
inequalities
the genders or
ethnic backgrounds and so on
um
so one very simple thing you can do here
is simply compare two groups and compare
the average outcomes between them
for example if we think about to our
clothing mortality example you could
take say
all the over 80s and compare the
mortality rates of black people and
white people
fairly straightforward exercise if you
have the data
um
but we often might want to do more than
this and that's because lots of these
inequalities have many
dimensions to them what do i mean by
that well
it might be that um
black people aren't more likely to die
of chronic violence than white people
and it is uh but there may be other
factors that are also relevant such as
um
household composition poverty where
people are from their age their previous
health all these different things and
these um
may correlate with ethnicity to some
extent
and that doesn't take away from the the
importance of the quality but it does
um
complicate the picture um because these
different inequalities between groups
can overlap and it's hard to disentangle
these effects
the one thing that can be done and was
done by the ons when they calculate that
figure earlier
is to do a regression that includes um
all of these different factors
so for example the ons regressed
the
probability of dying from comet on not
just uh ethnicity if there's two but
also a range of these other controls
which i talked about earlier
and you'll remember i said that if
you're just control for age black people
are four times more likely that uh to
die credibility than white people um but
if you include all these other controls
there's still a large difference and
they still find that black people are
twice as likely to die of chronic virus
as white people and so really what this
regression allows you to do is compare
uh like for like um
so if you
take people of a similar age similar
declaration level similar household
composition um
then what is the uh average difference
um that we will see between
ethnicities
okay
um but now for the rest of this section
i'm actually going to think about um
interpersonal
integrity across the whole of the
population rather than between really
qualities and as an example i'm actually
going to use household income um in the
household disposable income in the uk
and we're going to see how different
measures can tell us different things
about um incompetent quality
so one good way of illustrating
inequality is with the lorenz curve so
let me explain what the errance curve is
i want you to imagine
lining everybody in the population up
from
uh poorest to richest lowest income to
highest income
and
ask what is the total income owned by
burned by all these people everyone in
the population
then i want you to imagine
moving along this line from the lowest
and again to the highest and keeping a
running total of the income as you go
and then if you plot that running total
as a proportion of the overall total
income that gives you the lorentz curve
so this is the the rental curve for
those of you in the uk
uh so how do we read this well
what the loans here tells us is that um
the
25
uh
25 percent of people receive about 10 of
the income
uh whereas the bottom 50 percent of
people
receive
um
about 25 of the income
okay and that's a mathematical
expression of what i just said if you're
interested
now
if you imagine
a perfectly equal society which everyone
had the same income
um then
clearly trivially the poorest 20 percent
would receive 20 income of course 30 30
and so on and our lorentz curve could
actually be
this straight line
from 0 0 to 100 100
this is the perfect equality line
and if we look at how far away our
lorenz curve is from the perfect
quantity line that starts to give us a
way of comparing
inequality between
different distributions already
so the further away the lorentz curve is
from the equality line the more bound
out it is the more inequality we have
just as illustration here this blue line
illustrates
a
society with more income inequality
and actually straight from this graph we
can start to read one of our headline
statistics that we we often see when
thinking about inequality and that's the
genie coefficient
so if you take the area between the
perfect equality line and the lorentz
curve and you call it a
and then talk take the area
between the rex curve and the axis and
call it b
then the genie coefficient is
quite simply the area a over the total
area a plus b
so if we
imagine our perfect equality case where
everyone has the same income then the
area a is
on the perfect equality line and so we
have a genie coefficient of zero for
perfect equality
on the other hand imagine a society
where no one had any income at all
except the richest person who had all
the income and maybe we could call this
our perfect inequality case
um
in that case the area a at um
is the entire area between the
performance unit and the axis and b is
zero and our genie coefficient is one
so one is the case of perfect inequality
and as long as we don't have any
negative incomes our gene coefficient is
bounded between
zero and one
so that is our first
headline measure of inequality um that's
an alternative formulation for it you're
interested it's equivalently could be
thought of as the
[Music]
mean
deviation between all possible pairs you
could extended by the average
now the advantage of genie coefficient
is it's a it's a good measure that takes
account of the whole of the income
distribution um
so and it has a fairly intuitive
derivation if you want to take time to
look at it um but it is only a summary
measure um so you could have um
you know different quite differently
shaped income distributions that gave
you the same coefficient but on the
other hand you can also have
income distributions that are fairly
similar but have quite different genie
coefficients just because there are uh
some changes going on at the very top of
the distribution that's because the
genie collection is very sensitive to
extremes it really is a measure of how
concentrated
income is towards the extreme
and that's
becomes problematic really when we want
to estimate it because often we will
estimate
the genie coefficient using survey data
but if
jeff bezos answers your
income survey one year and not the next
you're going to get very different
answers even though not much has changed
and you you don't want um small changes
just from your sampling to have
drastic influences on the statistics you
calculate
so we do need to make adjustments if
we're going to calculate
so let's look at what the geneva
efficient can tell us about uk dispersed
blink so i'm going to plot the genie
coefficient
since the 1960s
and
what we can see is that um
in that period over the last 60 years or
so there has been a significant increase
in
inequality on the genie measure but that
this really was
concentrated in the 1980s
which as i'm sure everyone knows is a
time of significant economic change in
the uk and internationally
changes to trade unions attacks and
benefits systems large structural
changes to the economy
but since the 1980s there's been very
little change really it
perhaps a small uptick in anything um up
to the financial crisis
and then
not much going on since then either um
so
yes we've got this story of a
significant increase of the junior
coefficient in the last 16 years
primarily concentrated on the in the 80s
and that has taken
that increase has taken the uk towards
the top of the
league table when considering uh
high-income developed countries
but um the genie coefficient is as i say
is not the only measure we should look
at and there are other measures as well
um
now one
which i actually alluded to earlier when
talking about wealth is is calculating
income shares or wealth shares and
this is um something we can illustrate
on the
curve so
if we look
say at the 90 level we can say that
the poorest 90 percent received just
over 70 of the income
and from that obviously the top ten
percent um receive just under thirty
percent of the income and that's what we
mean when we talk about income shares um
we'd also do it for the
top one descendancy they receive nine
percent or so of the income
so
income shares are good for understanding
inequalities specifically at the top
they so they focus on specific parts of
the distribution um so they don't tell
you everything but it may well be
inequality at the top that you are
particularly worried about today that's
what you're
researching or do you think that has
particularly adverse consequences that
other inequalities might not um so it
depends on your purposes but incoming
income shares are one very good way of
summarizing a particular um type of
inequality
uh but again just with us with the ge it
could be quite hard to estimate these if
you're using survey data because of
difficulties in capturing um incomes at
the very top
uh another measure uh which we normally
use to focus on the middle of the
distribution was uh percentile ratios
so i'll show you how we calculate those
and this time i'm going to plot uh
income distribution um in a slightly
different chart so i'm gonna plot income
percentiles
so what do these mean well again imagine
they've got everyone lined up in order
of how much you can learn and you're
moving along the line um once you get uh
10 of the way along you ask whoever is
there uh how much income player they'll
tell you about 250 pounds
50 percent of the way on they'll tell
you a little over 500 pounds and we plot
this on the graph
and
set our ratios
quite simply you take the ratio of the
percentiles that you're interested in so
as i said um
the
10th percentile about 250 pounds per
week in income at the 90th percentile um
a bit over a thousand pounds a week and
that gives us what we call a 90 10 ratio
of for
19 obviously four times the income of
the death
on the other hand if we look at the
fifth 1950 ratio um we see a ratio of
about two
so that's percentile ratios and we can
use these to summarize in the quantity
income distribution you can choose which
percentiles you use um and these are
usually focused
on the middle of the income distribution
i mean middle in quite a broad sense it
could be the middle uh 80 percent but um
looking at what's going on in bulk of
the population
without reference to what's going on at
the extreme ends
and just in case you're interested but
um don't worry about it but you can also
see percentile is on the lorenz curve so
if you take the slope of the lorentz
90th percentile and the slope at the
tenth the ratio of the slopes that also
gives you uh percentile ratios
okay let's see what these measures tell
us well
start with our 90 10 ratio actually not
too different to the
significant increase of the last 60
years
mostly concentrated in the 1980s
perhaps you can
tell a little bit of a story of a bit of
a decrease
since um but really not very much um
most of the action again going on in the
80s
but on the other hand
um if we need to plot the top 1 share we
get quite a different story we still see
stability in
60s and 70s of increase in the 80s
but that increase really carried on um
through the 90s up until the financial
crisis
and that's something that isn't obvious
at all from looking at our diversions of
inequality so while the genius 9010
ratios tell us the story of fairly flat
inequality
since the 80s financial crisis
the top 1 share tells us quite a
different story it's only since the
financial crisis that that measure has
uh started to stabilize ability
and then i'm going to give
one final measure and that's actually
relative poverty so perhaps it's not
always
thought of as a measure of inequality
but it is
and let me show you what we mean by
relative property this is how it's
defined usually so if we take our
percentiles of the income distribution
again
we ask um
what is the median income the 50th
percentile
we see that's
about
[Music]
that gives us 330 pounds
and we ask what proportion of people
receive beneath
that level
and that proportion um
18 13
that gives us our
poverty our relative poverty rate
um
and we we can i showed you just on
disposable incomes before subtracting
housing costs but we can also calculate
this measure after subtracting housing
costs there's not a contract to see
what's the best
but either way uh relative poverty
defines the fraction living below sixty
percent of the contemporaneous median
income is
essentially a measure of inequality and
it's a measure of lower end inequality
so i think this is a it's a good measure
it's an interesting measure because it's
a particular type of inequality that we
might be interested in um
perhaps it gets this idea
that i was talking about earlier with
the capabilities approaches sufficiency
for individual standard living is really
it really depends on
what the rest of the population has so
it's almost a proxy for that sort of
idea
um
but i think there's a
world of warning that needs to be made
here and that's that
really this 60 measures is a
is an arbitrary cut off um and indeed
when we're talking about
top income shares percentile ratios the
numbers we choose they're arbitrary as
well um but it's important to remember
that if you're using these measures to
evaluate the impact of the policy
because you wouldn't want to focus
just on relative
poverty um
relative poverty that
a measure that changes the incomes just
of people who are already below the
poverty line or
of people who are you know further up
the distribution that's good that's
going to have no impact on the rent on
tv even though it might have significant
distribution of consequences including
on the people um that you're interested
in
so
you while renting properties it is a
interesting summary measure of lower-end
inequality it's not something that you
want to focus on in isolation if your
aim
is to evaluate policy
[Music]
and if we look at our relative property
rate in the uk
and this time i have deducted housing
costs and we can see not too dissimilar
pictures many of our other measures um
stability towards before the 80s
significant increase
um
in the 80s bit of a
downward trend um
up to the financial crisis perhaps but
only a partial reversal of the increase
we saw in the 80s and then stability
since then
okay so just to summarize we've seen
that uh various measures of inequality
we can use if we're looking at between
group inequality you can do simple
comparisons of group beverages or you
can run a regression to disentangle the
various different uh groups that we see
any quantities between
whereas if you're looking at
interpersonal inequalities you've got
lots of measures you can use the genie
coefficient
top
uh one percent ten percent shares
percentile ratios uh relative poverty
and
um i gave you the example of income but
really well perhaps with these top three
at least you can apply these to any
outcome you like right to wealth
consumption um
any any outcome that you're interested
in
okay and different measures give us
different insights but on all measures
incoming according to has increased in
the last 60 years
okay so let's just uh wrap up obviously
if people are concerned about a huge
range of inequalities so there are
different outcomes
um and the arguments we can see four are
against reducing inequality and
particularly the state taking action
using the quantity whilst they depend on
moral values and complains about
fairness they're also contingent
on cultural claims about inequalities
causes and consequences and of the
consequences of policy and these can be
evaluated by economists
now there are many ways of measuring
inequalities and different measures
could give us different inequalities
into how a different insights and how
the quantities have changed
um and as we as we've just seen um
inequality has increased in the uk in
incomes in the last 60 years um but
apart from at the very top most of this
change in the quality happened during
the 1980s there's been first now
stability since
except for the top one percent who
really have um
continued to hurt the way all the way up
to the financial crisis
when income started to stabilize
okay um
so that's all from me and i'm now able
to take
questions keep posting your questions
and slide it
so i'll stop sharing i'm back to shower
that's brilliant thanks so much tom
um yeah so we've now got some time for
questions
as tom said please uh do post your
questions on the slider and vote up the
questions you'd like to see answered and
i see that we've got a couple of
questions but i think we should have
time for more than this
and so the top rated question at the
moment is how income inequality in the
uk compares to other countries
internationally
okay yeah so i i did mention this um
briefly and
if we think about the um
gene measures one of our headline
measures um and if we if we focus on
other wealthy developed countries the uk
really is actually quite high it's
towards the top i think it's really only
the uh
us that stands out
above the uk that the uk does stand out
as well um
if we think about the top one percent uh
share
then again the uk is fairly high up up
the uh international league table if we
look at similar countries but i would
say it doesn't stand out quite as much
as still closer to the massive countries
as opposed to the us which really it
does stand out on that measure as well
so
yeah but overall i think it's fair to
say uh looking at similar countries that
incoming equality in the uk
is quite high
great
um another question we have is on data
um so there's a question on um how we
get the data that
feeds into these inequality graphs um
and whether the the quality of that data
has got better or worse over time
that's uh it's yeah it's an interesting
question so
um there's yeah there's
you could
essentially divide the data we could use
into two categories so it's
administrative data and
survey data
so administrative data is data that the
government um
or you know other authorities will hold
for operational purposes
so for example hmrc might know uh
people's earnings um and that data is
very useful in that it is very
comprehensive it should
include everybody or almost everybody um
but uh disadvantages of using that sort
of data are that um
well
there's limits to what we can know from
it they won't hold informational
information characteristics so they'd
only hold the information they need to
know and also it can be very difficult
for researchers to get their hands on
their data sometimes so the alternative
is survey data
now with survey data
the disadvantage is you're relying on a
sample you're hoping that that sample is
going to be representative or you can
adjust it so that it becomes
representative
and you might not get that right however
you can ask a much broader range of
questions and if we're interested in
inequalities between
different groups for example different
characteristics and then that's when
survey data really comes into its own
into um
understanding um
these trends um in terms of how quality
of data has changed over time um
well i'd say that
there's
always there's lots been lots of there's
research going on into how we can use
the data we've got better um so i've
mentioned there's difficulties in
understanding what's going on at the top
to measure top incomes and capture those
so there's lots of research going on to
try and develop better ways
better techniques of dealing with that
issue and
also all the time we get
large new surveys introduced
for example since about 2006 i think
we've had a lot of an asset survey which
has helped us look at
wealth inequality um
so new surveys appearing and that really
helps as researchers as well
that's great yeah i might just add to
that that sometimes um we also try to
combine these data sources so for
example we know that top incomes and
people earning um
a high very high levels are
underrepresented in survey data so we
might use an administrative tax data to
adjust for incomes at the top
um i think in terms of you know trends
over time um one thing that we are aware
of is that
if you use tax data from the uk of
course your your um capturing comes that
reported to uk's tax authorities and if
people are
perhaps increasingly using tax havens to
channel their funds then it might be
that we're missing
um part of that data so i think there's
been some work done at lse
on how much um wealth or income we might
be missing through those channels
um and there's a related question to you
to this question on data on um
informality so i think i think this is
referring to um
public burnings in the informal economy
so the question is
how do we factor in informality when
measuring inequality
yeah that's a very important question
and i think here
here um this is an example of when
administrative data
um is unlikely to be very helpful uh
these are precisely this sort of
arrangements where
for example hmrc won't know about the
earnings that are being received
potentially um
so
in these cases um we're probably going
to rely on surveys of um some
description but even then it it may well
be difficult um
these earnings in informal arrangements
might not be measured very well or
people may not disclose them even even
just in a survey um so yeah i think it
is a valid point that it is quite
difficult to factor in informality when
measuring these
equalities but at the end of the day
if there is a way of doing it it really
it's important to consider it because
you know it's income inequality we're
considering
uh then we need to know about incomes
from as many sources as we can it's
just a matter of overcoming the
informational limitations there are
yeah and perhaps in these these
circumstances and measures consumption
inequality could be used to sort of
supplement the income data
yeah absolutely yes
yeah i think yeah consumption inequality
is i mentioned in the lecture that um
income inequality is often thought of as
oh sorry that incomes are often as a
proxy for um living standards um but
they're not the only property we have
available um
one problem with using incomes is that
incomes can vary quite a lot over the
lifetime people don't tend to have much
income when they're very young for
example or when they
retired um they might have less than
when they were working and that
consumption we might think of as an
alternative property because people
might smooth their consumption a bit
more over time so they may borrow when
they're younger for example as a student
they may save some of their income when
they're working and then
divest their savings in retirement so
yeah these other measures are also using
different proxies for uh living
standards isn't definitely i agree
another way of getting this issue
great um so there's one final question
on uh covert 19 so what covets done to
inequality um so i guess you've already
talked to um you know some results on
inequality and mortality by race but
maybe you could talk to
some other forms of inequality so for
example
what's happened to inequality in incomes
uh maybe between the rich and poor the
young and old men and women that sort of
thing
yes so it's it's a
really obviously topical question and
it's uh
because it's such early days really in
terms of getting data on this it's hard
to answer it definitively um
but really
thinking about um
disposable income the evidence so far is
that there hasn't been
drastic changes to the income
distribution as as a result of cobit um
even though there were obviously
significant shocks to the um labor
market and people's employment and that
uh
between different groups quite
substantially
we've seen huge government spending in
terms of the furlough scheme uh
self-employed income scheme universal
credit
and
um
which really we it does look like have
helped us smooth out some of these
shocks and since then as these schemes
have been wound down we actually haven't
seen the
big permanent shocks to employment that
perhaps we might have been fearing so um
i wouldn't say there's been drastic
changes so far in terms of incomes but
that really is only so far we don't know
what the long-term effects are going to
be
for example you know lots of people were
out of work for long periods during the
pandemic they had less work experience
less obtuse for training and that may
have so-called scaring effects so in the
future they may become less productive
in that way her earnings um
another long-term consequence which i
think is really important is um
education in the qualities and again
very briefly but um children from
different backgrounds have very
different um
educational provision available to them
during the
pandemic and we have seen correlations
between
parental earnings and socioeconomic
background of these children how much um
education remote education they got
during lockdowns and that could have
much longer uh consequences for earnings
inequality
as well um we know that children's
earnings often correlate with what their
parents earned and that could become
even more true if lower earning uh
children
would have had a stronger negative
effect their education is more
condemning than uh
children from higher education
backgrounds so that really here we need
to watch
what policies the government uh uses to
try and
compensate for education last year
if we're going to get an answer to that
question so we've got another question
that's related to um um this idea of
intergenerational transmission so
um it mentions that many of the measures
you've you've used so far relate to
income
but
you know what what evidence is there on
wealth inequalities and in particular
measures related to inherited wealth
yeah so um
there has been work looking at um
in
wealth in general and inherited wealth
and there's lots of mechanisms um by
which um
wealth inequalities persist um between
the
generations clearly direct inheritances
are very big component of that um
but we can also see that um there are
correlations between um
children's earnings and parental wealth
or um
you know where children are educated
savings behavior all sorts of factors as
well um and so i i encourage you ever
ask that question to um search the ifs
website for i think a fairly recent
paper a few months ago which uh covered
some of that in more detail but yeah
there is evidence inheritance is
definitely important but it's not the
only driver of
wealth inequalities and how they persist
over time
yeah absolutely i think um
some ifs work maybe it was the same
paper also found that inheritances are
becoming
an increasingly large um driver of
wealth inequalities um between people
the same generation in that you know
younger people are building up wealth
more slowly so the inheritance they get
from their parents is becoming a larger
share of their wealth and therefore
differences in that the level of that
inheritance is driving inequalities in
the world for more than before
and i would also add on the wealth point
that is that is another uh impact of the
um pandemic that we have been able to
see already um there have been huge
changes to uh
the
wealth distribution um partly driven by
changes to property prices we've seen
big increases in housing prices in
particular um also due to increases in
global equities and i think what the
pattern we've seen there is that the
wealth of the
uh
middle and the top of the pre-economic
health distribution has increased
especially in the middle whereas uh
those at the bottom of the distribution
haven't really seen much gaining wealth
and tools and so
so that is another
pandemic rated inequality which we do
know something about which is increasing
wealth inequality but in particular
that's lower enforcement
fantastic thanks tom well we've run out
of questions on slido and i think we've
just about run out of time as well and
so i think we'll end here
um just a reminder that this is a series
of lectures on public economics and next
week we'll have another lecture um
monday 4-5 p.m on the effects of
graduating into a recession um so don't
forget to register for that if you're
interested
um so i guess i'll leave it here thanks
very much for tuning in and thanks very
much tom for presenting