TEDGlobal 2011
Richard Wilkinson: How economic inequality harms societies
Filmed:
Readability: 4.6
3,198,950 views
We feel instinctively that societies with huge income gaps are somehow going wrong. Richard Wilkinson charts the hard data on economic inequality, and shows what gets worse when rich and poor are too far apart: real effects on health, lifespan, even such basic values as trust.
Richard Wilkinson - Public health researcher
In "The Spirit Level," Richard Wilkinson charts data that proves societies that are more equal are healthier, happier societies. Full bio
In "The Spirit Level," Richard Wilkinson charts data that proves societies that are more equal are healthier, happier societies. Full bio
Double-click the English transcript below to play the video.
00:15
You all know the truth of what I'm going to say.
0
0
3000
00:18
I think the intuition that inequality is divisive and socially corrosive
1
3000
4000
00:22
has been around since before the French Revolution.
2
7000
4000
00:26
What's changed
3
11000
2000
00:28
is we now can look at the evidence,
4
13000
2000
00:30
we can compare societies, more and less equal societies,
5
15000
3000
00:33
and see what inequality does.
6
18000
3000
00:36
I'm going to take you through that data
7
21000
3000
00:39
and then explain why
8
24000
2000
00:41
the links I'm going to be showing you exist.
9
26000
4000
00:45
But first, see what a miserable lot we are.
10
30000
3000
00:48
(Laughter)
11
33000
2000
00:50
I want to start though
12
35000
2000
00:52
with a paradox.
13
37000
3000
00:55
This shows you life expectancy
14
40000
2000
00:57
against gross national income --
15
42000
2000
00:59
how rich countries are on average.
16
44000
2000
01:01
And you see the countries on the right,
17
46000
2000
01:03
like Norway and the USA,
18
48000
2000
01:05
are twice as rich as Israel, Greece, Portugal on the left.
19
50000
5000
01:10
And it makes no difference to their life expectancy at all.
20
55000
4000
01:14
There's no suggestion of a relationship there.
21
59000
2000
01:16
But if we look within our societies,
22
61000
3000
01:19
there are extraordinary social gradients in health
23
64000
3000
01:22
running right across society.
24
67000
2000
01:24
This, again, is life expectancy.
25
69000
2000
01:26
These are small areas of England and Wales --
26
71000
2000
01:28
the poorest on the right, the richest on the left.
27
73000
4000
01:32
A lot of difference between the poor and the rest of us.
28
77000
3000
01:35
Even the people just below the top
29
80000
2000
01:37
have less good health
30
82000
2000
01:39
than the people at the top.
31
84000
2000
01:41
So income means something very important
32
86000
2000
01:43
within our societies,
33
88000
2000
01:45
and nothing between them.
34
90000
3000
01:48
The explanation of that paradox
35
93000
3000
01:51
is that, within our societies,
36
96000
2000
01:53
we're looking at relative income
37
98000
2000
01:55
or social position, social status --
38
100000
3000
01:58
where we are in relation to each other
39
103000
3000
02:01
and the size of the gaps between us.
40
106000
3000
02:04
And as soon as you've got that idea,
41
109000
2000
02:06
you should immediately wonder:
42
111000
2000
02:08
what happens if we widen the differences,
43
113000
3000
02:11
or compress them,
44
116000
2000
02:13
make the income differences bigger or smaller?
45
118000
2000
02:15
And that's what I'm going to show you.
46
120000
3000
02:18
I'm not using any hypothetical data.
47
123000
2000
02:20
I'm taking data from the U.N. --
48
125000
2000
02:22
it's the same as the World Bank has --
49
127000
2000
02:24
on the scale of income differences
50
129000
2000
02:26
in these rich developed market democracies.
51
131000
3000
02:29
The measure we've used,
52
134000
2000
02:31
because it's easy to understand and you can download it,
53
136000
2000
02:33
is how much richer the top 20 percent
54
138000
2000
02:35
than the bottom 20 percent in each country.
55
140000
3000
02:38
And you see in the more equal countries on the left --
56
143000
3000
02:41
Japan, Finland, Norway, Sweden --
57
146000
2000
02:43
the top 20 percent are about three and a half, four times as rich
58
148000
2000
02:45
as the bottom 20 percent.
59
150000
3000
02:48
But on the more unequal end --
60
153000
2000
02:50
U.K., Portugal, USA, Singapore --
61
155000
2000
02:52
the differences are twice as big.
62
157000
3000
02:55
On that measure, we are twice as unequal
63
160000
3000
02:58
as some of the other successful market democracies.
64
163000
4000
03:02
Now I'm going to show you what that does to our societies.
65
167000
4000
03:06
We collected data on problems with social gradients,
66
171000
3000
03:09
the kind of problems that are more common
67
174000
2000
03:11
at the bottom of the social ladder.
68
176000
2000
03:13
Internationally comparable data on life expectancy,
69
178000
3000
03:16
on kids' maths and literacy scores,
70
181000
3000
03:19
on infant mortality rates, homicide rates,
71
184000
3000
03:22
proportion of the population in prison, teenage birthrates,
72
187000
3000
03:25
levels of trust,
73
190000
2000
03:27
obesity, mental illness --
74
192000
2000
03:29
which in standard diagnostic classification
75
194000
3000
03:32
includes drug and alcohol addiction --
76
197000
2000
03:34
and social mobility.
77
199000
2000
03:36
We put them all in one index.
78
201000
3000
03:39
They're all weighted equally.
79
204000
2000
03:41
Where a country is is a sort of average score on these things.
80
206000
3000
03:44
And there, you see it
81
209000
2000
03:46
in relation to the measure of inequality I've just shown you,
82
211000
3000
03:49
which I shall use over and over again in the data.
83
214000
3000
03:52
The more unequal countries
84
217000
2000
03:54
are doing worse
85
219000
2000
03:56
on all these kinds of social problems.
86
221000
2000
03:58
It's an extraordinarily close correlation.
87
223000
3000
04:01
But if you look at that same index
88
226000
2000
04:03
of health and social problems
89
228000
2000
04:05
in relation to GNP per capita,
90
230000
2000
04:07
gross national income,
91
232000
2000
04:09
there's nothing there,
92
234000
2000
04:11
no correlation anymore.
93
236000
3000
04:14
We were a little bit worried
94
239000
2000
04:16
that people might think
95
241000
2000
04:18
we'd been choosing problems to suit our argument
96
243000
2000
04:20
and just manufactured this evidence,
97
245000
3000
04:23
so we also did a paper in the British Medical Journal
98
248000
3000
04:26
on the UNICEF index of child well-being.
99
251000
4000
04:30
It has 40 different components
100
255000
2000
04:32
put together by other people.
101
257000
2000
04:34
It contains whether kids can talk to their parents,
102
259000
3000
04:37
whether they have books at home,
103
262000
2000
04:39
what immunization rates are like, whether there's bullying at school.
104
264000
3000
04:42
Everything goes into it.
105
267000
2000
04:44
Here it is in relation to that same measure of inequality.
106
269000
4000
04:48
Kids do worse in the more unequal societies.
107
273000
3000
04:51
Highly significant relationship.
108
276000
3000
04:54
But once again,
109
279000
2000
04:56
if you look at that measure of child well-being,
110
281000
3000
04:59
in relation to national income per person,
111
284000
2000
05:01
there's no relationship,
112
286000
2000
05:03
no suggestion of a relationship.
113
288000
3000
05:06
What all the data I've shown you so far says
114
291000
3000
05:09
is the same thing.
115
294000
2000
05:11
The average well-being of our societies
116
296000
2000
05:13
is not dependent any longer
117
298000
3000
05:16
on national income and economic growth.
118
301000
3000
05:19
That's very important in poorer countries,
119
304000
2000
05:21
but not in the rich developed world.
120
306000
3000
05:24
But the differences between us
121
309000
2000
05:26
and where we are in relation to each other
122
311000
2000
05:28
now matter very much.
123
313000
3000
05:31
I'm going to show you some of the separate bits of our index.
124
316000
3000
05:34
Here, for instance, is trust.
125
319000
2000
05:36
It's simply the proportion of the population
126
321000
2000
05:38
who agree most people can be trusted.
127
323000
2000
05:40
It comes from the World Values Survey.
128
325000
2000
05:42
You see, at the more unequal end,
129
327000
2000
05:44
it's about 15 percent of the population
130
329000
3000
05:47
who feel they can trust others.
131
332000
2000
05:49
But in the more equal societies,
132
334000
2000
05:51
it rises to 60 or 65 percent.
133
336000
4000
05:55
And if you look at measures of involvement in community life
134
340000
3000
05:58
or social capital,
135
343000
2000
06:00
very similar relationships
136
345000
2000
06:02
closely related to inequality.
137
347000
3000
06:05
I may say, we did all this work twice.
138
350000
3000
06:08
We did it first on these rich, developed countries,
139
353000
3000
06:11
and then as a separate test bed,
140
356000
2000
06:13
we repeated it all on the 50 American states --
141
358000
3000
06:16
asking just the same question:
142
361000
2000
06:18
do the more unequal states
143
363000
2000
06:20
do worse on all these kinds of measures?
144
365000
2000
06:22
So here is trust from a general social survey of the federal government
145
367000
4000
06:26
related to inequality.
146
371000
2000
06:28
Very similar scatter
147
373000
2000
06:30
over a similar range of levels of trust.
148
375000
2000
06:32
Same thing is going on.
149
377000
2000
06:34
Basically we found
150
379000
2000
06:36
that almost anything that's related to trust internationally
151
381000
3000
06:39
is related to trust amongst the 50 states
152
384000
2000
06:41
in that separate test bed.
153
386000
2000
06:43
We're not just talking about a fluke.
154
388000
2000
06:45
This is mental illness.
155
390000
2000
06:47
WHO put together figures
156
392000
2000
06:49
using the same diagnostic interviews
157
394000
2000
06:51
on random samples of the population
158
396000
2000
06:53
to allow us to compare rates of mental illness
159
398000
3000
06:56
in each society.
160
401000
2000
06:58
This is the percent of the population
161
403000
2000
07:00
with any mental illness in the preceding year.
162
405000
3000
07:03
And it goes from about eight percent
163
408000
3000
07:06
up to three times that --
164
411000
2000
07:08
whole societies
165
413000
2000
07:10
with three times the level of mental illness of others.
166
415000
3000
07:13
And again, closely related to inequality.
167
418000
4000
07:17
This is violence.
168
422000
2000
07:19
These red dots are American states,
169
424000
2000
07:21
and the blue triangles are Canadian provinces.
170
426000
4000
07:25
But look at the scale of the differences.
171
430000
3000
07:28
It goes from 15 homicides per million
172
433000
3000
07:31
up to 150.
173
436000
3000
07:34
This is the proportion of the population in prison.
174
439000
3000
07:37
There's a about a tenfold difference there,
175
442000
3000
07:40
log scale up the side.
176
445000
2000
07:42
But it goes from about 40 to 400
177
447000
2000
07:44
people in prison.
178
449000
3000
07:47
That relationship
179
452000
2000
07:49
is not mainly driven by more crime.
180
454000
2000
07:51
In some places, that's part of it.
181
456000
3000
07:54
But most of it is about more punitive sentencing,
182
459000
2000
07:56
harsher sentencing.
183
461000
2000
07:58
And the more unequal societies
184
463000
2000
08:00
are more likely also to retain the death penalty.
185
465000
4000
08:04
Here we have children dropping out of high school.
186
469000
5000
08:09
Again, quite big differences.
187
474000
2000
08:11
Extraordinarily damaging,
188
476000
2000
08:13
if you're talking about using the talents of the population.
189
478000
3000
08:16
This is social mobility.
190
481000
3000
08:19
It's actually a measure of mobility
191
484000
2000
08:21
based on income.
192
486000
2000
08:23
Basically, it's asking:
193
488000
2000
08:25
do rich fathers have rich sons
194
490000
2000
08:27
and poor fathers have poor sons,
195
492000
2000
08:29
or is there no relationship between the two?
196
494000
3000
08:32
And at the more unequal end,
197
497000
2000
08:34
fathers' income is much more important --
198
499000
3000
08:37
in the U.K., USA.
199
502000
3000
08:40
And in Scandinavian countries,
200
505000
2000
08:42
fathers' income is much less important.
201
507000
2000
08:44
There's more social mobility.
202
509000
3000
08:47
And as we like to say --
203
512000
2000
08:49
and I know there are a lot of Americans in the audience here --
204
514000
3000
08:52
if Americans want to live the American dream,
205
517000
3000
08:55
they should go to Denmark.
206
520000
2000
08:57
(Laughter)
207
522000
2000
08:59
(Applause)
208
524000
4000
09:03
I've shown you just a few things in italics here.
209
528000
3000
09:06
I could have shown a number of other problems.
210
531000
2000
09:08
They're all problems that tend to be more common
211
533000
2000
09:10
at the bottom of the social gradient.
212
535000
2000
09:12
But there are endless problems with social gradients
213
537000
5000
09:17
that are worse in more unequal countries --
214
542000
2000
09:19
not just a little bit worse,
215
544000
2000
09:21
but anything from twice as common to 10 times as common.
216
546000
3000
09:24
Think of the expense,
217
549000
2000
09:26
the human cost of that.
218
551000
3000
09:29
I want to go back though to this graph that I showed you earlier
219
554000
2000
09:31
where we put it all together
220
556000
2000
09:33
to make two points.
221
558000
2000
09:35
One is that, in graph after graph,
222
560000
3000
09:38
we find the countries that do worse,
223
563000
2000
09:40
whatever the outcome,
224
565000
2000
09:42
seem to be the more unequal ones,
225
567000
2000
09:44
and the ones that do well
226
569000
2000
09:46
seem to be the Nordic countries and Japan.
227
571000
3000
09:49
So what we're looking at
228
574000
2000
09:51
is general social disfunction related to inequality.
229
576000
3000
09:54
It's not just one or two things that go wrong,
230
579000
2000
09:56
it's most things.
231
581000
2000
09:58
The other really important point I want to make on this graph
232
583000
3000
10:01
is that, if you look at the bottom,
233
586000
2000
10:03
Sweden and Japan,
234
588000
3000
10:06
they're very different countries in all sorts of ways.
235
591000
3000
10:09
The position of women,
236
594000
2000
10:11
how closely they keep to the nuclear family,
237
596000
2000
10:13
are on opposite ends of the poles
238
598000
2000
10:15
in terms of the rich developed world.
239
600000
2000
10:17
But another really important difference
240
602000
2000
10:19
is how they get their greater equality.
241
604000
3000
10:22
Sweden has huge differences in earnings,
242
607000
3000
10:25
and it narrows the gap through taxation,
243
610000
2000
10:27
general welfare state,
244
612000
2000
10:29
generous benefits and so on.
245
614000
3000
10:32
Japan is rather different though.
246
617000
2000
10:34
It starts off with much smaller differences in earnings before tax.
247
619000
3000
10:37
It has lower taxes.
248
622000
2000
10:39
It has a smaller welfare state.
249
624000
2000
10:41
And in our analysis of the American states,
250
626000
2000
10:43
we find rather the same contrast.
251
628000
2000
10:45
There are some states that do well through redistribution,
252
630000
3000
10:48
some states that do well
253
633000
2000
10:50
because they have smaller income differences before tax.
254
635000
3000
10:53
So we conclude
255
638000
2000
10:55
that it doesn't much matter how you get your greater equality,
256
640000
3000
10:58
as long as you get there somehow.
257
643000
2000
11:00
I am not talking about perfect equality,
258
645000
2000
11:02
I'm talking about what exists in rich developed market democracies.
259
647000
4000
11:08
Another really surprising part of this picture
260
653000
5000
11:13
is that it's not just the poor
261
658000
2000
11:15
who are affected by inequality.
262
660000
3000
11:18
There seems to be some truth in John Donne's
263
663000
2000
11:20
"No man is an island."
264
665000
3000
11:23
And in a number of studies, it's possible to compare
265
668000
3000
11:26
how people do in more and less equal countries
266
671000
3000
11:29
at each level in the social hierarchy.
267
674000
3000
11:32
This is just one example.
268
677000
3000
11:35
It's infant mortality.
269
680000
2000
11:37
Some Swedes very kindly classified a lot of their infant deaths
270
682000
3000
11:40
according to the British register of general socioeconomic classification.
271
685000
5000
11:45
And so it's anachronistically
272
690000
3000
11:48
a classification by fathers' occupations,
273
693000
2000
11:50
so single parents go on their own.
274
695000
2000
11:52
But then where it says "low social class,"
275
697000
3000
11:55
that's unskilled manual occupations.
276
700000
3000
11:58
It goes through towards the skilled manual occupations in the middle,
277
703000
4000
12:02
then the junior non-manual,
278
707000
2000
12:04
going up high to the professional occupations --
279
709000
3000
12:07
doctors, lawyers,
280
712000
2000
12:09
directors of larger companies.
281
714000
2000
12:11
You see there that Sweden does better than Britain
282
716000
3000
12:14
all the way across the social hierarchy.
283
719000
3000
12:19
The biggest differences are at the bottom of society.
284
724000
2000
12:21
But even at the top,
285
726000
2000
12:23
there seems to be a small benefit
286
728000
2000
12:25
to being in a more equal society.
287
730000
2000
12:27
We show that on about five different sets of data
288
732000
3000
12:30
covering educational outcomes
289
735000
2000
12:32
and health in the United States and internationally.
290
737000
3000
12:35
And that seems to be the general picture --
291
740000
3000
12:38
that greater equality makes most difference at the bottom,
292
743000
3000
12:41
but has some benefits even at the top.
293
746000
3000
12:44
But I should say a few words about what's going on.
294
749000
4000
12:48
I think I'm looking and talking
295
753000
2000
12:50
about the psychosocial effects of inequality.
296
755000
3000
12:53
More to do with feelings of superiority and inferiority,
297
758000
3000
12:56
of being valued and devalued,
298
761000
2000
12:58
respected and disrespected.
299
763000
3000
13:01
And of course, those feelings
300
766000
2000
13:03
of the status competition that comes out of that
301
768000
3000
13:06
drives the consumerism in our society.
302
771000
3000
13:09
It also leads to status insecurity.
303
774000
3000
13:12
We worry more about how we're judged and seen by others,
304
777000
4000
13:16
whether we're regarded as attractive, clever,
305
781000
3000
13:19
all that kind of thing.
306
784000
3000
13:22
The social-evaluative judgments increase,
307
787000
3000
13:25
the fear of those social-evaluative judgments.
308
790000
4000
13:29
Interestingly,
309
794000
2000
13:31
some parallel work going on in social psychology:
310
796000
4000
13:35
some people reviewed 208 different studies
311
800000
3000
13:38
in which volunteers had been invited
312
803000
3000
13:41
into a psychological laboratory
313
806000
2000
13:43
and had their stress hormones,
314
808000
2000
13:45
their responses to doing stressful tasks, measured.
315
810000
4000
13:49
And in the review,
316
814000
2000
13:51
what they were interested in seeing
317
816000
2000
13:53
is what kind of stresses
318
818000
2000
13:55
most reliably raise levels of cortisol,
319
820000
3000
13:58
the central stress hormone.
320
823000
2000
14:00
And the conclusion was
321
825000
2000
14:02
it was tasks that included social-evaluative threat --
322
827000
3000
14:05
threats to self-esteem or social status
323
830000
3000
14:08
in which others can negatively judge your performance.
324
833000
3000
14:11
Those kind of stresses
325
836000
2000
14:13
have a very particular effect
326
838000
3000
14:16
on the physiology of stress.
327
841000
3000
14:20
Now we have been criticized.
328
845000
2000
14:22
Of course, there are people who dislike this stuff
329
847000
3000
14:25
and people who find it very surprising.
330
850000
3000
14:28
I should tell you though
331
853000
2000
14:30
that when people criticize us for picking and choosing data,
332
855000
3000
14:33
we never pick and choose data.
333
858000
2000
14:35
We have an absolute rule
334
860000
2000
14:37
that if our data source has data for one of the countries we're looking at,
335
862000
3000
14:40
it goes into the analysis.
336
865000
2000
14:42
Our data source decides
337
867000
2000
14:44
whether it's reliable data,
338
869000
2000
14:46
we don't.
339
871000
2000
14:48
Otherwise that would introduce bias.
340
873000
2000
14:50
What about other countries?
341
875000
2000
14:52
There are 200 studies
342
877000
3000
14:55
of health in relation to income and equality
343
880000
3000
14:58
in the academic peer-reviewed journals.
344
883000
3000
15:01
This isn't confined to these countries here,
345
886000
3000
15:04
hiding a very simple demonstration.
346
889000
2000
15:06
The same countries,
347
891000
2000
15:08
the same measure of inequality,
348
893000
2000
15:10
one problem after another.
349
895000
3000
15:14
Why don't we control for other factors?
350
899000
2000
15:16
Well we've shown you that GNP per capita
351
901000
2000
15:18
doesn't make any difference.
352
903000
2000
15:20
And of course, others using more sophisticated methods in the literature
353
905000
4000
15:24
have controlled for poverty and education
354
909000
2000
15:26
and so on.
355
911000
3000
15:30
What about causality?
356
915000
2000
15:32
Correlation in itself doesn't prove causality.
357
917000
3000
15:35
We spend a good bit of time.
358
920000
2000
15:37
And indeed, people know the causal links quite well
359
922000
2000
15:39
in some of these outcomes.
360
924000
2000
15:41
The big change in our understanding
361
926000
2000
15:43
of drivers of chronic health
362
928000
2000
15:45
in the rich developed world
363
930000
2000
15:47
is how important chronic stress from social sources
364
932000
4000
15:51
is affecting the immune system,
365
936000
2000
15:53
the cardiovascular system.
366
938000
3000
15:56
Or for instance, the reason why violence
367
941000
2000
15:58
becomes more common in more unequal societies
368
943000
3000
16:01
is because people are sensitive to being looked down on.
369
946000
5000
16:06
I should say that to deal with this,
370
951000
3000
16:09
we've got to deal with the post-tax things
371
954000
2000
16:11
and the pre-tax things.
372
956000
2000
16:13
We've got to constrain income,
373
958000
3000
16:16
the bonus culture incomes at the top.
374
961000
2000
16:18
I think we must make our bosses accountable to their employees
375
963000
3000
16:21
in any way we can.
376
966000
3000
16:24
I think the take-home message though
377
969000
3000
16:27
is that we can improve the real quality of human life
378
972000
4000
16:31
by reducing the differences in incomes between us.
379
976000
3000
16:34
Suddenly we have a handle
380
979000
2000
16:36
on the psychosocial well-being of whole societies,
381
981000
2000
16:38
and that's exciting.
382
983000
2000
16:40
Thank you.
383
985000
2000
16:42
(Applause)
384
987000
6000
ABOUT THE SPEAKER
Richard Wilkinson - Public health researcherIn "The Spirit Level," Richard Wilkinson charts data that proves societies that are more equal are healthier, happier societies.
Why you should listen
For decades, Richard Wilkinson has studied the social effects of income inequality and how social forces affect health. In The Spirit Level, a book coauthored with Kate Pickett, he lays out reams of statistical evidence that, among developed countries, societies that are more equal – with a smaller income gap between rich and poor -- are happier and healthier than societies with greater disparities in the distribution of wealth.
While poverty has long been recognized as an indicator for such social ills as crime, obesity, teen pregnancy, Wilkinson and Pickett have demonstrated that societal well-being bears no relation to per capita income. They’ve also found that the symptoms of inequality trouble all levels of society. Across the board, mental health, levels of violence and addiction, even life expectancy are affected by the psycho-social stress caused by income gaps and status anxiety.
In the UK, The Spirit Level won support from politicians both left and right. Wilkinson, who is Professor Emeritus of Social Epidemiology at the University of Nottingham, also cofounded The Equality Trust, a nonprofit that aims to reduce income inequality by educating and engaging the public while supporting political commitment to address the problem.
He says: "While I'd always assumed that an equal society must score better on social cohesion, I never expected to find such clear differences between existing market economies."
NEW: Read the TED Blog's Q&A with Richard Wilkinson >>
Richard Wilkinson | Speaker | TED.com