Commentary: Intergenerational and intragenerational equity.
Portes, Jonathan
The concept of 'intergenerational fairness' was
introduced to the UK policy debate by David Willetts MP (now, somewhat
ironically, the Minister for Universities and Science, and hence
responsible for the new system of student finance) in his book, The
Pinch: How the Baby Boomers stole their childrens' future and how
they can give it back (Willetts, 2010). But it has only taken off in
political discussion over the past couple of years; the idea that
somehow the old are getting a better deal than the young, both before
the financial crisis and during the ensuing period of austerity, has
gained popularity across the political spectrum. As Angus Hanton,
Director of the Intergenerational Foundation, put it (Intergenerational
Foundation, 2013):
"Rising intergenerational unfairness should matter to
everyone. The usual focus on inequality between rich and poor misses the
important inequalities between generations. Poorer young people are
increasingly and systematically financing richer older people. Urgent
action must be taken as the Baby Boom cohort starts to retire."
But what does 'intergenerational fairness' (or its
opposite) mean? And how would we know if it is 'rising'? In
practice, this discussion is bedevilled by confusion, both conceptual
and empirical. The purpose of this article is to set out some of the
possible meanings of the term 'intergenerational fairness'
and, at a very stylised level, what the evidence suggests. I identify
four different, and conceptually distinct, issues:
a) The evolution of the relative incomes of different cohorts. Have
older generations (and in particular pensioners) done particularly well
by comparison to younger generations over the past twenty years or so?
b) The burden of austerity: have the specific policy measures taken
over the past few years to reduce the deficit hit the young particularly
hard, while favouring the old?
c) Are the young 'subsidising' the old via taxation and
the provision of public services?
d) Will this generation 'be the first to be worse off than
their parents'?
What has happened to incomes?
Have pensioners done better than non-pensioners? Fortunately, the
ONS produces a very long series for household incomes. From 1977 to
2011, median equivalised disposable income (that is, adjusted for family
size) for pensioner households rose from 7,090 [pounds sterling] to
19,250 [pounds sterling] (in 2011-12 prices). The rise for non-pensioner
households was from 12,080 [pounds sterling] to 25,700 [pounds sterling]
(ONS, 2013).
So the incomes of pensioner households rose significantly faster on
average than for non-pensioners. But there is no indication that this
process accelerated in recent years; the pace of convergence was roughly
the same in the first half of this period as in the second. And, of
course, pensioner incomes remain considerably lower than non-pensioner
incomes throughout this period, even on an equivalised basis. The
discrepancy would be considerably larger in cash terms, since pensioner
households are smaller on average. The median non-retired household,
with an income of more than 25,000 [pounds sterling], would still be at
about the 80th percentile of the pensioner income distribution; in other
words only 20 per cent of retired households are better off than the
median non-retired household. Most pensioners are not 'rich'
compared to non-pensioners.
[FIGURE 1a OMITTED]
What about the distribution of these gains ? Figure I shows real
income growth by decile for retired and non-retired households, between
1996-7 and 2011-12. It shows that gains were actually reasonably well
spread, with poor non-retired households, and middle-income retired
ones, doing relatively well (although these statistics, based on survey
data, do not show the particularly strong growth for incomes of the very
highest earners--the top 1 per cent and 0.1 per cent--during this
period). Even leaving this aside, it remains the case that inequality
within both the retired and non-retired groups is much greater than
disparities between the two.
[FIGURE 1b OMITTED]
[FIGURE 2 OMITTED]
What are the drivers of this relative improvement in the position
of pensioners? It appears to be a combination of two factors: the
maturing of relatively generous defined benefit pension schemes has
helped middle and upper income pensioners, while the introduction of
Pension Credit and the Minimum Income Guarantee has benefited poorer
ones, with a notable fall in pensioner poverty (as measured by relative
income) over the past decade (DWP, 2013).
It does not, however, represent any increased generosity in the
level of the basic, non-means-tested state pension: even including
ancillary universal benefits such as Winter Fuel Payments (see below)
the value of the basic state pension relative to earnings fell very
substantially in the 1980s and 1990s, and has not recovered
significantly since. As with the income distribution as a whole (Portes,
2011), the 2000s saw an increase in the inequality of original income
(before tax and benefits) that was in large part mitigated by increases
in the redistributive impact of the tax and benefit system, but most of
that redistribution was intragenerational rather than intergenerational.
The burden of austerity
While the above analysis suggests that pensioners have been doing
rather better than non-pensioners--in the sense that their incomes have
been catching up--the current focus of the political debate on
'intergenerational fairness' has been driven by a set of
policy decisions taken by the Coalition government. It is argued that
these have ensured that older people have been 'protected'
from the spending cuts and tax rises imposed to reduce the budget
deficit. These policies include:
* The 'triple lock', guaranteeing that pensions should be
increased by 2.5 per cent, earnings or RPI, whichever is the greater;
* Ancillary to this, the Prime Minister's personal commitment
to preserving other universal pensioner benefits (winter fuel payments,
free bus passes);
* At the same time, a series of spending cuts which appear to fall
most heavily on the relatively young: the abolition of the Education
Maintenance Allowance, the very large increase in university tuition
fees, large cuts to working age benefits and tax credits, while health
spending (which is heavily weighted towards the old) is protected in
real terms.
[FIGURE 3 OMITTED]
On the face of it, it does seem obvious that these policy changes
will, taken together, tend to hurt those of working age rather more than
pensioners. And indeed this is the case. Figure 3, from the IFS Green
Budget 2013, shows the distributional impact of tax and benefit reforms.
Pensioner households do indeed lose considerably less on average
than working age households without children. But in fact working age
households without children do no worse than pensioners. And arguably
more important are the intra- (rather than inter-) generational issues.
Poorer pensioners have indeed been protected; less so richer ones. By
contrast, relatively well off working-age households without children
have lost very little, except at the very top, while poorer childless
households have suffered. Meanwhile, households with children have lost
out across the board. This is likely to reflect the net impact of cuts
to working age benefits, especially disability-related benefits, hitting
poorer working-age households (with and without children); cuts to child
benefit and tax credits, hitting families across the board; the increase
in the personal allowance, benefiting particularly middle-income working
households; and the protection to poorer pensioners afforded by the
triple lock.
In the light of this, it is interesting to consider the
implications of one policy that has become something of a lightning rod
for this debate: the Prime Minister's personal commitment not to
withdraw the Winter Fuel Payment (WFP) for pensioners. Arguably, this is
something of a distraction; while the WFP may be an anomaly, it is very
difficult to argue that over the longer term the level of universal
state provision for pensioners has become more generous, as figure 2
shows, or that means-testing it would make the system either
'fairer' or more rational (Portes, 2012). But, leaving logic
aside, what would be the distributional impact of means-testing WFP? By
definition, the impact would primarily be on middle-income pensioners
(poor pensioners would be protected, while for richer ones WFP is not a
significant income source).
But, as figure 3 shows, middle-income pensioners have actually
already lost more from recent policy changes than either poorer or
richer ones. The chart suggests that if the objective of policy was to
smooth the burden of deficit reduction amongst these groups
("we're all in this together") by targeting a group that
had got off relatively lightly so far, the appropriate target group
would be relatively high income working households (albeit not the very
richest) without children. Ironically, this group benefits significantly
from the government's signature tax policy of raising the personal
allowance (Browne, 2012), suggesting that those who are critical of the
distributional impact of current policy might be better off focusing on
this policy decision rather than that on WFP.
This analysis only looks at taxes and benefits, not at other
spending, so excludes the impact of tuition fees, the ring-fencing of
NHS spending, etc. Distributional analyses of the impact of these
measures suggest that overall they are regressive by income (Horton and
Reed, 2011); but no age-related analysis exists at present, although it
seems likely that the relative protection afforded to health spending
compared to other areas will indeed favour older age groups, while lower
income working-age families with children will be the biggest losers.
One policy where the intergenerational impacts have been much distorted
in public debate, however, is that of university tuition fees. In fact,
taxpayer subsidies to students have tended to rise over recent decades.
What has changed is that the subsidy is distributed over far more
students, as HE participation has tripled since the late 1970s. The
introduction of income-contingent fees--better described as a capped
graduate tax- probably does not reduce expenditure much (the situation
is obscured by the way the government accounts for student loans) while
making the distribution of the subsidy rather more progressive.
Certainly the idea that the pre-tuition fee regime, with far fewer
students receiving higher subsidies, and going on to receive excellent
labour market returns, was in any sense 'fairer' seems hard to
sustain.
It is also important to note that the calculations described above
only represent a snapshot. Clearly protecting pensions at the expense of
working age benefits redistributes from relatively poor people currently
of working age to pensioners in general. But those currently of working
age will eventually become pensioners; the net impact on them will
depend on the extent to which the policy is in fact sustainable and
sustained (or, put another way, how it is financed). As Chris Dillow has
pointed out (Dillow, 2013):
"Rising pensions are better for youngsters than oldsters. This
is simply because they can look forward to many years of rises, whilst
an old person will probably die after only a few years of getting them.
If you're 40 years shy of pension age, a 0.5 per cent annual real
rise in the state pension increases the value of your retirement wealth
by over 25 per cent. This is much more than an 80-year-old will get, who
can anticipate only eight years of rising pensions before he dies."
It follows that a government decision to implement a more generous
pension uprating policy redistributes to the elderly only if some other
assumption is made: for example, that the policy is unsustainable and
will have to be offset in future through tax increases or spending cuts
that will fall on those now younger. In order to assess this argument,
we need to look at tax and spending over time.
Are the young subsidising the old?
The concept of 'intergenerational accounts' as a means to
analyse long-term distributional consequences of tax, transfers and
public spending was developed by Auerbach, Gokhale and Kotlikoff (1994).
Generational accounts are designed to show the net discounted lifetime
contribution, positive or negative, that people, as a function of their
age, are expected to make to the Exchequer. This involves a number of
steps: first, calculating taxes and public spending by age groups over
time (for the past), and then projecting these forward on some
assumption about future policies. Since the government must satisfy its
intertemporal budget constraint--the future discounted value of taxes
must equal the value of future spending plus any existing debt--any gap
that emerges from this must be filled somehow, by adjusting future taxes
or spending.
Of course, the outputs of these calculations--the degree to which
the government is redistributing income between different generations,
past, present, and future--depend crucially on various assumptions, of
two kinds: first, those that determine how large the gap that needs to
be filled is (these include demographic projections, the definition of
the 'unchanged policy' baseline, and macroeconomic variables
such as the growth rate and the real interest rate) and second, how the
gap is filled (on whom does the burden of future tax or spending changes
fall).
The most recent generational accounts for the UK were produced by
McCarthy, Sefton and Weale (2011) and are reproduced here (table 1).
They show net contributions/ receipts for all those now alive, assuming
'unchanged policy' applies to all those currently alive; and
assume that the full burden of any necessary adjustment falls equally on
'future generations'.
These do indeed show a considerable degree of intergenerational
redistribution. Taken at face value, the table suggests that those
currently aged 65 will receive 220,000 [pounds sterling] more in
spending than they pay in tax, over their lifetimes. Those just born
will pay in 68,000 [pounds sterling]; while future generations will each
contribute almost 160,000 [pounds sterling].
But this in itself shows the (necessarily) somewhat arbitrary
nature of the assumptions here. Clearly there is no set of feasible
policy changes that will impose an extra 90,000 [pounds sterling]
lifetime burden on someone born next year ('future
generations') compared to those born last year. Adjustment will in
practice be shared between those currently alive and future generations.
It is also worth noting that the substantial burden on those
currently young, and future generations, arises from the calculation
that current tax and spending policies are very far from sustainable,
with a projected future gap between revenues and expenditures of about
6-6.5 per cent of GDP (in NPV terms). This is much more pessimistic than
the picture shown by the OBR's most recent Fiscal Sustainability
Report, although the key trends and drivers--pension and health spending
in particular--are the same (OBR, 2013). This partly reflects recent
policy changes (such as the commitment to raise the state pension age in
future, and reductions in future public sector pensions) which will
impact those alive now as well as future generations.
A further point bearing on the 'fairness' issue emerges
from looking more closely at table 1. The substantial net payment to
those currently aged 65, and net contribution from those in their
twenties and thirties, emerges not because the older generations are
getting more from the state (in pensions, health care, education and so
on)--if anything the reverse (given the rise in life expectancy, and
substantial real terms increases in health spending over the past few
decades, those who are young now are expected to get considerably more
from the state on average than those now old). The spending reductions
now being implemented are unlikely to change that. Instead, it is
because they are expected to pay much more in tax. And this in turn is
not primarily because tax rates have increased over the past half
century or so, but rather because the later generations have earned more
(and spent more) and will do so (assuming continued economic growth) in
the future, hence paying much more in taxes. This also of course implies
that their lifetime consumption (taking into account both their
after-tax income and their consumption of public services) is likely to
be considerably higher than that of previous generations.
To see this in a very rough and ready way, note that the net
present value of future taxes, on unchanged policies, for those
currently aged 20 is estimated at well over 0.5 million [pounds
sterling], implying lifetime consumption (private and public) of well
over 1 million [pounds sterling] in NPV terms, even after their net
'contribution' of 157,000 [pounds sterling]; while those now
age 50 will pay something over 0.2 million [pounds sterling] in taxes;
even after net 'receipts' of about 105,000 [pounds sterling],
their lifetime consumption will be closer to 0.7 million [pounds
sterling].
So what is essentially happening here is not that the government is
deliberately favouring the currently old in terms of public spending,
but rather that the government is redistributing some of the
'fruits of growth' from those now earning and producing to
those who did so in the past. Since the growth that makes this possible
relies--both in an economic sense (investment, research, etc.) and, more
broadly, in political, institutional and moral terms, on the
contribution of previous generations--this casts such redistribution in
a somewhat different light. Whether the level of redistribution implied
by current policies is 'fair' or not is obviously a subjective
issue, but it is certainly not axiomatic that government should seek to
balance generational accounts of the form shown here.
Will this generation "be the first to be worse off than their
parents"?
Perhaps the phrase that has most resonance in this debate has been
the claim that the generation currently entering the labour market will
be "the first to be worse off than their parents". On the
surface, this seems implausible: UK GDP growth over the past half
century, up to the 2008 crisis, has averaged a little over 2 per cent
per year; this has, by and large, translated into real income growth. As
Dillow (2014) points out,
"Barring an improbable disaster, today's young people
will on average be better off than my generation on average."
[FIGURE 4 OMITTED]
However, a recent IFS analysis (Hood and Joyce, 2013) suggests that
there was at least some truth in this thesis. Figure 4 shows the path of
cohort incomes, and does indeed suggest that those born in the 1970s are
doing worse at this point in their careers than earlier cohorts.
This represents a combination of cyclical and structural factors.
Developments in the UK economy and labour market over the past decade
have clearly disadvantaged younger generations, both in absolute and
relative terms: working-age income growth has slowed (and gone into
reverse since 2008) just at the time when, on the basis of past
experience, those in their thirties would have expected to see rapid
progression. Will this damage prove permanent? This obviously depends on
broader issues, and in particular the pace of future productivity and
real wage growth. However, assuming that UK productivity and hence
output growth does return to something approaching the historical
average, it is difficult to see that real wage growth will not
eventually follow suit.
How the fruits of this growth are distributed is a very different
matter, as the example of the US shows. The US has averaged output and
productivity growth of about 2 per cent per year over the past three
decades. However, incomes for the top 1 per cent of households nearly
quadrupled, while those for the bottom quintile grew by only 18 per cent
over the entire period (CBO, 2011 ). The result was that nearly a third
of all the income growth over the period went to the top 1 per cent.
Although inequality increased in the UK in the first half of this
period (1979-95), this pattern was not replicated in the UK in the years
up to the mid-2000s (Portes, 2011), as wage growth, while favouring
higher earners, was much more evenly spread, and redistributive tax and
benefit policies, especially in the early 2000s, restrained the rise in
income inequality. However, in recent years, UK trends have conformed
more closely to the US example. Median wages have grown significantly
less than per capita GDP (Resolution Foundation, 2013), as an increasing
proportion of the increase in average (mean) incomes accrues to those at
the top of the income distribution. HMRC data suggest that about a
quarter of the growth in real pre-tax incomes between 1999-2000 and
2013-14 accrued to the top 1 per cent of (income) taxpayers; almost as
much as went to the bottom 50 per cent (HMRC, 2013).
And indeed the IFS figures show a sharp rise in intragenerational
inequality. The 80-20 ratio for those aged 40 has gone from 2.2 for
those born in the 1940s, to about 2.5 for those born in the 1950s and
1960s, and to about 2.8 for those born in the 1970s (and again this is
likely to understate the growth in inequality at the very top of the
income distribution). By contrast, even if we take the current data at
face value, those born in the 1970s are only about 10 per cent worse off
than those born in the 1960s. So it is hardly surprising that younger
people on low and middle incomes feel that they are doing rather badly
in relative terms, but this is primarily a result of increases in
intragenerational inequality that have hit them particularly hard. The
circumstances of your birth and early life (who your parents are,
education, gender, ability, effort and just plain luck) still matter far
more--indeed, more so than before--than when you were born.
It is also worth noting that, just as intragenerational inequality
remains considerably more important than intergenerational inequality,
so does intragenerational redistribution. The top decile of non-retired
households pay 30,000 [pounds sterling] more annually in taxes than they
receive in benefits and services, while the bottom decile benefit by
10,000 [pounds sterling]; while the figures are not directly comparable
to those in the generational accounts described above, it seems clear
that over a lifetime this type of redistribution--from richer to poorer,
not from younger to older cohorts--is likely to dominate for most
households. The same is true for changes to policy: the main impact of
the government's changes to benefits will not be to redistribute
from young to old, but from poorer to middle-income and richer
households.
Similar considerations apply to the structural changes set out in
the analysis, in particular lower levels of home ownership and expected
future pension income (both state and private). As the IFS notes, the
fact that those now relatively old have high levels of home ownership,
and the sharp rise in house prices over the past few decades, means that
those now young and middle-aged expect to inherit considerably more than
for past cohorts; but, unsurprisingly, "expected inheritances are
distributed unequally and are higher for those who are already
wealthier": well over two-thirds of property wealth belongs to the
wealthiest third of the population. And, crucially, rising property
prices do not directly raise the consumption of older cohorts unless
they realise the value of their properties via equity release or
downsizing; if they simply translate into larger inheritances, then the
redistributive impact is primarily within rather than between
generations. As Portes (2013) puts it:
"There is no doubt that rising house prices redistribute
wealth, and in a way that is regressive and damaging both economically
and socially. But the main dimension on which this redistribution takes
place is not that of age, but of income and wealth (and to a lesser
extent geography)."
In other words, it is quite wrong to say that inequality between
cohorts makes inequalities within cohorts less important; the opposite
is the case.
Conclusion
This brief survey has attempted to show that it is possible to look
at the issue of 'intergenerational fairness' through a number
of different lenses; this in turn means that any conclusions are at
least in part subjective. However, the simple story of coddled
pensioners and struggling youth does not really add up: both inequality
and redistribution within generations remains much greater than that
between generations--and its impact is growing--while the impact of
recent policy decisions on intergenerational equity is more ambiguous
than is immediately apparent.
REFERENCES
Auerbach, A.J., Gokhale, J. and Kotlikoff, L.J. (1994),
'Generational accounting: a meaningful way to evaluate fiscal
policy', Journal of Economic Perspectives, 8(1), pp. 73-94.
Browne, J. (2012), 'A 10,000 [pounds sterling] personal
allowance: who would benefit, and would it boost the economy?',
Institute for Fiscal Studies, March.
Congressional Budget Office (2011), Trends in the Distribution of
Household Income between 1979 and 2007, October.
Department for Work and Pensions (2013), Households Below Average
Income 2011-12, July.
Dillow, C. (2013), 'The intergenerational question',
Stumbling and Mumbling (blog), October.
--(2014), 'Rising pensions: who gains', Stumbling and
Mumbling (blog), January.
HMRC (2013), Income Tax Uabilities Statistics, September.
Hood, A. and Joyce, R. (2013), 'The economic circumstances of
cohorts born between the 1940s and the 1970s', Institute for Fiscal
Studies.
Horton, T, and Reed, R. (2011), 'The distributional
consequences of the 2010 Spending Review', Journal of Poverty and
Social Justice, 19, 1, February, pp. 63-66.
Institute for Fiscal Studies (2013), Green Budget 2013, February.
Intergenerational Foundation (2013), 2013 Intergenerational
Fairness Index, June.
McCarthy, D., Sefton, J. and Weale, M. (2011), 'Generational
accounts for the United Kingdom', National Institute of Economic
and Social Research Discussion Paper No. 377, January.
Office for National Statistics (2013), Effects of Taxes and
Benefits on Household Income, 1977-2011/12, July.
Office for Budget Responsibility (2013), Fiscal Sustainability
Report, July.
Portes, J. (2011), 'Poverty and inequality:
introduction', National Institute Economic Review, 218, October,
pp. R1-6.
--(2012), 'Why pick on rich pensioners' bus passes',
Independent, June.
--(2013), 'Rising house prices help the well-off",
Guardian, October.
Resolution Foundation (2013), Squeezed Britain 2013, September.
Willetts, D. (2010), The Pinch: How the Baby Boomers Took Their
Children's Future--And Why They Should Give it Back, Atlantic
Books.
Jonathan Portes, National Institute of Economic and Social Research
and Centre for Macroeconomics. I am grateful to Raquel Fernandez for
very helpful comments and to various authors at the Institute for Fiscal
Studies for permission to reproduce two of the charts. E-mail:
j.portes@niesr.ac.uk.
Table 1. Base generational accounts for the United Kingdom, 2008
prices
Present 1 2 3 4
[pounds Total Income Production Social
sterling] taxes taxes security
0 68375 185937 162813 105814
5 79151 208916 177933 118217
10 106868 214902 177112 121688
15 143165 222432 176536 126413
20 157700 210679 163527 119437
25 124486 188883 146281 105726
30 67684 165298 132273 89908
35 29832 140823 117553 74758
40 -11925 116689 105451 61122
45 -58601 92564 93644 47705
50 -105011 68902 81086 34205
55 -155058 45561 68463 20933
60 -198930 26152 55373 9814
65 -223183 14720 43770 4309
70 -206912 9827 33943 2848
75 -177732 7165 25016 2032
80 -143375 5364 17428 1530
85 -114636 3968 12206 1155
90 -86040 2997 8653 886
95 -57721 2085 5491 616
Future
Generations 159668
Present 5 6 7 8
[pounds Other Other Health Education
sterling] current receipts
taxes
0 28065 605 -97286 -84054
5 32206 629 -109472 -80723
10 32941 599 -108462 -64469
15 33806 568 -106613 -47106
20 32333 510 -100363 -20180
25 29869 454 -98709 -6100
30 27998 417 -101299 -3400
35 25441 376 -97370 -1990
40 23139 343 -95838 -1312
45 20919 313 -95601 -895
50 18622 280 -94238 -398
55 16455 250 -92747 -169
60 14004 217 -88205 -96
65 11687 187 -84436 -57
70 9539 157 -78861 -33
75 7438 124 -69391 -18
80 5514 93 -58735 -9
85 3936 69 -48766 -3
90 2851 52 -40058 -1
95 1853 35 -28216 -1
Future
Generations
Present 9 10
[pounds Pensions Social Other
sterling] welfare
0 -99138 -88379 -46002
5 -120423 -95786 -52346
10 -120990 -94232 -52222
15 -118940 -92011 -51921
20 -113558 -85404 -49280
25 -117130 -77117 -47672
30 -125197 -70539 -47774
35 -122741 -61130 -45889
40 -125064 -52625 -43829
45 -128811 -46391 -42049
50 -132239 -41241 -39990
55 -139669 -36364 -37771
60 -150746 -30645 -34799
65 -152334 -27965 -33063
70 -129289 -26270 -28776
75 -103037 -23590 -23472
80 -76743 -20782 -17038
85 -53878 -18082 -15240
90 -37718 -14248 -9454
95 -24879 -10151 -4555
Future
Generations
Source: McCarthy, Sefton and Weale (2011).
