1. No category

Debate on Inequalities in Health
MAKING SENSE OF INEQUALITIES:
A RESPONSE TO PETER TOWNSEND
Rudolf Klein
This article addresses the continuing controversy generated by the Black Report
on Inequalities in HeaZth, published in Britain in 1980, in response to thc defense
offered by Professor Peter Townsend. The author argues that Townsend’s riposte to
the critics of the Black Report is flawed in at least two respects. First, Townsend fails
to acknowledge that the Black Report was as much an exercise in policy advocacy as
in scholarly analysis, making rather large assumptions about the links in the reasoning leading to its recommendations for a massive program of income redistribution.
Second, Townsend’s defense of Black’s use of social class as its main tool for
analyzing health inequalities dismisses too easily much of the evidence; for example,
the effects of social mobility and the historical dimension. Moreover, by concentrating on social class, a heterogeneous category, analysis may ignore what is most
relevant for policy-making: i.e., specific factors associated with specific forms of
deprivation, located within social classes or particular geographical communities. It
would therefore be more constructive if scholars were to accept and research this
complexity, rather than defending the Black Report as though it were a definitive
(not to say sacred) text.
The Black Report on Inequalities in Health (1) has achieved, both nationally and
internationally, a peculiar status since its publication in Britain in 1980: it has become
something of an intellectual icon. I am therefore grateful to Professor Peter Townsend,
and to the editor of this Journal, for the opportunity presented by Townsend’s recent
article (2) to try to demystify the debate. As Townsend’s article suggests, with its
charges of “unreasoned criticism” and “misrepresentation,” the Black Report tends to
evoke highly emotional, defensive responses among its progenitors and partisans. In
what follows, I will therefore first try to explain why this is so-to “deconstruct” the
controversy, as it were-before moving on to deal with some of the more specific,
technical issues raised. For what makes the debate about the Black Report so interesting
in my view is precisely that it offers an example of a controversy in which the
protagonists have different agenda and use different analytical paradigms with the result
that they tend to talk past each other rather than engaging in constructive, scholarly
argument.
International Journal of Health Services, Volume 21, Number 1, Pages 175-1 81, 1991
0 1991, Baywood Publishing Co., Inc.
175
doi: 10.2190/D6H4-79KQ-NADE-QU9N
http://baywood.com
176 I Klein
PARADIGMS AND AGENDAS
Townsend’s own agenda is clear and explicit. His concern, reflecting his lifelong
dedication to social reform, is to use evidence about inequalities in health to reinforce or
justify the case for “a massive assault on material deprivation,” to quote his concluding
paragraph. This would involve, as the Black Report argued, a quite radical redistribution
of resources to those on low incomes. If the Black Report arouses passion, it is therefore
‘becauseit is part of a political program. It is as much an exercise in persuasion, designed
to prod government along a particular path of action, as a neutral “scientific” inquiry;
hence, of course, the very long and detailed menu of specific policy recommendations
put forward in the Report. It is an example, in this respect, of policy advocacy disguised
as, or at least overlapping with, policy analysis.
My own agenda, in the article that so offended Townsend that he chose to pick on it
rather than the many other papers critical of the Black Report published over the years,
was also clear and explicit (3). It was to try to sort out the notion of “inequality,” and its
role in the policy-making process, drawing largely on the work of political theorists and
using the Black Report to illustrate my arguments. So, for example, I asked whether
health policy should be exclusively driven by a concern about inequality-the unargued
and unexamined premise of the Black Report-r
whether maximizing the population’s
health might not be an equally desirable policy objective, if with rather different
implications. Finally, I advanced the heretical proposition-in Blackian eyes-that the
case for greater social and economic equality should be argued on its own terms, as
intrinsically desirable, rather than using the health issue to justify it. Where the Black
Report offered faith, I urged the case for agnosticism. None of this is, of course,
addressed in Townsend’s riposte; however, it may help to explain his reaction and his
use of my article as a reason for replying to the more expert and technical literature
critical of the Black Report.
There is another difference between Townsend’s approach and mine that requires
noting: we are using different paradigms of the policy process. Professor Townsend’s
assumption appears to be that if you demonstrate large societal effects-such as growing inequalities in health status between social classes-this, in itself, justifies largescale, heroic policy remedies such as the redistribution of income: which is, of course,
why he is so sensitive to any suggestion that the inequalities may not be widening. My
own assumption is that demonstrating large, general effects is unhelpful as a guide to
policy-making, unless it is also possible to establish direct causality and to show how
specific policy interventions will have specific outcomes. Indeed the Black Report
(much more so than Townsend) hedged its bets between these two paradigms: it
combined some highly specific, and sensible, micro-recommendations with its much
more questionable macro-proposals for income redistribution.
THE CENTRAL ISSUE
So we come to what is perhaps the central issue in the controversy between Peter
Townsend, as the champion of the Black Report, and the critics. This is, first, whether it
is sensible to use social class as an analytical category and, second, what (if anything) it
tells us about inequalities over time. As Townsend observes, the Black Report itself was
Making Sense of Inequalities 1 177
conscious about the problems of using the conventional social class categories and drew
attention to the need for more research. But one of the characteristics of the Black
Report, in this as well as in other respects, is precisely that it is extremely meticulous in
identifying the weaknesses of the evidence while ignoring its own reservations when
drawing conclusions; no doubt the mixed membership of the committee that produced it
may help to explain this. (It would not be difficult to identify the authorship both of the
reservations and of the recommendations.) Acknowledging possible objections to one’s
arguments while subsequently brushing them aside is one of the classic devices of
rhetoric or the art of persuasion, and the Black Report offers some prime examples.
The Registrar-General’s classification of occupations provides, in the words of one of
its historians “an extremely crude, but generally accepted, index of social inequality”; its
use may therefore be pragmatically justified by “those hoping to influence opinion and
policy with their findings.” However, “Sociologists and social historians must distinguish clearly between such expedients and those empirical definitions of social class
which are of analytical value to their academic disciplines” (4, p. 539). For the
Regiistrar-General’s classification is based on an “obsolete conceptual framework”
embodying a set of highly questionable propositions, notably that occupation is the most
reliable single empirical indicator of the status and attributes of individuals. If our
concern is with understanding society, rather than trying to influence policy, we therefore. have to be extremely cautious in drawing any conclusions based on the RegistrarGeneral’s categories.
This point applies with special force, of course, to drawing conclusions about trends
over time, i.e., about whether or not inequalities between social classes are widening. If
we don’t think that the conventional occupational classifications are robust analytical
tools at any one point, then these reservations will apply with even greater force over
time. And they will certainly not be dispelled by reanalyzing the data, however
ingenious and heroic the investment of effort. Reanalysis cannot cope with the underlying conceptual problems. Nor can it cope with the problems first identified by Illsley
(5): that, over time, the relative size of classes changes, i.e., that the significance and
social meaning of inequalities change. The point can be simply illustrated. In 1951
semiskilled and unskilled manual workers represented 38.5 percent of the workforce in
Great Britain; by 1981 the proportion had fallen to 29.4 percent (6, p. 82). Conversely,
the proportion accounted for by the “service class” (7)-employers, managers, administrators, professionals, and technicians-rose from 18.7 to 29.4 percent. Hence the
conclusion of Carr-Hill, in the most authoritative overview of the post-Black controversy, that: “By ignoring changes in class composition and size over time, one can point
to apparent increases in inequality over time. However, the claim that, IN GENERAL,
health differences associated with socio-economic position have increased, cannot be
maintained” (8, p. 535).
There are other difficulties about using social class as an analytical tool. There is the
welll-known problem that women are classified on the basis of their husband’s occupation; there is, too, the equally well-known unreliability of the classification of the
over-65s. The Black Report solved the latter by effectively ignoring the over-65s in its
main analysis, so improving the validity of the figures but impoverishing their meaning.
Townsend is disingenuous on this point. If one turns to the relevant chapters of the Black
Report-as instructed by Townsend-ne
finds that there is, indeed, a discussion of
178 1 Klein
infant mortality and of the elderly. But the key tables illustrating the supposed widening
of inequalities over time deal only with the population between 15 and 64.So Illsley’s
point-that the Black Report ignored the real improvement in the health of the working
class over time, i.e., the fact that a far higher proportion of its members survive beyond
the age of 65-remains untouched. To claim that there “can be no scientific doubt about
the direction of the trend” in inequalities, as Townsend does, is therefore quite simply
wrong. The position of Illsley-who accepts (as we all do) that there are serious
inequalities in health between the prosperous and the poor, but remains agnostic about
trends over time-is the only possible one. Progress in science surely depends on the
existence of doubt: on the willingness to probe, challenge, and test. When there is no
doubt, as in Professor Townsend’s mind, there can also be no science. Only religious
movements (whether of the ecclesiastical or political variety) rest on uncontestable
propositions and pronounce anathema on doubters.
DEPRIVATION AND HEALTH
Townsend’s insistence that inequalities are growing is all the odder because, in fact,
it is largely irrelevant to his main concern, which he shares with his critics. This,
presumably, is to illuminate the relationship between specific forms of social and
economic deprivation and health. That there is such a relationship is, of course, well
recognized. In many respects the Black Report was only echoing and updating
Chadwick’s 1842 Report on the Sanitary Condition of the Labouring Population of
Great Britain (9): here, too, we find tables comparing the mortality of the upper and
lower classes. And social policy for the last 150 years in Britain has largely been based
on the recognition of this fact: hence, for example, the dramatic decline in working-class
deaths from those conditions reflecting poor housing, poor nutrition, and other circumstances amenable to policy intervention.
In all these cases, success has largely rested on developing our understanding of
precisely what conditions are linked to particular aspects of health. The trouble about
using social class as an analytical can-opener in this respect is that it may encompass a
whole variety of different forms of deprivation (or none): family income, housing,
working conditions, environment, and so on. And it is not self-evident which factorsor cluster of factors-are most directly related to health, and where, therefore, policydirected change can have most impact. Consider, for example, geographical variations
in standardized mortality ratios within social classes. Thus regional inequalities (10,
p. 247) within social clasJes IV and V are almost as great as national inequalities
between these and social classes I and 11. Workers in East Anglia are almost as
healthy-to the extent that standardized mortality ratios are any guide-as professionals
in Scotland. Similarly, higher death rates in the poorer socioeconomic groups is not
automatic evidence of deprivation. For example, most male deaths in the 15 to 64 age
groups used by Black actually occur between 45 and 64,and these are primarily due to
lung cancer and ischemic heart disease. Rates for these diseases are certainly highest in
classes IV and V but the causes-cigarette smoking, diets high in fat, lack of exercise
leading to obesity-are not the result of material deprivation as conventionally defined,
and policies based on that assumption would be counterproductive.
Making Sense of Inequalities 1 179
So, clearly, the relationship between social class, deprivation, and health is extremely
complex. The point is well caught in the following quotation reporting on a microanalysis of mortality in the United Kingdom (11, p. 142):
Occupational class is clearly of major importance in explaining inequalities in health
but, just as clearly, there is a need to go beyond occupational class (at least as
ordinarily interpreted) in explaining those inequalities. Broadly speaking, either
there are factors independent of occupational class which contribute in substantial
measure to any further explanation of excess deaths, or occupational class includes
systematic variations in the experience of material deprivation which need to be
revealed if a further large number of the observed excess deaths are to be explained.
The case for caution in the use of social class as an instrument for gaining understanding
about the relationship between deprivation and health could hardly be better put. And it
is all the more convincing since it comes from Townsend himself.
The point should, in any case, hardly need arguing by now. It was made more than 50
years ago by M’Gonigle’s and Kirby’s (12) celebrated study of two housing estates
where, counter-intuitively, health was poorer on the new one; the improvement in
hou.sing conditions had been more than offset, it turned out, by higher rents and
con:sequentially lower food budgets. There could be no better warning against embarking on social engineering without a clear understanding of the circumstances and causal
relationships. More recently, a study investigated why three socially similar towns in
England differed considerably in their health record (13). The explanation offered was
that these differences could be traced back to “the environmental influences that determined past differences in child development.” While this remains no more than an
interesting hypothesis, it offers a warning against attempts to explain health inequalities
excllusively in terms of recent changes in the socioeconomic environment: the reasons
for the inequalities may be found less in the record of the present Government than in
the performance or failures of its predecessors. The historical dimension, clearly, is
crucial for any understanding.
PRESCRIPTIONS AND CONCLUSIONS
So one could go on: the temptation to reply to Townsend point by point, reference by
reference, is strong. Instead I have sought here to address mainly the issues or principles
und.erlying the dispute between Townsend and those who think that the Black Report
was less than perfect. In conclusion, however, two specific points need to be
addressed4ne because Professor Townsend makes much of it, the other because he
ignores it totally.
The first is that of whether, and to what extent, selective social mobility can explain
eithier remaining or widening (depending on the view one takes of this) inequalities
between social classes. That there is such selective mobility, with the healthiest traveling upward on the social scale, was shown by Illsley (14) 35 years ago. That, in theory,
such mobility could explain all the observed inequality has more recently been demonstra.tedby Stern (15) in an elegant modeling exercise. That the expansion of the “service
cla:js” has meant a large degree of upward mobility is equally well established. To this
Townsend replies that the Black Report did acknowledge selective mobility as a
180 J Klein
possible explanatory factor. However, as he also points out, the Report in fact put all its
money on “materialist” or “structural” explanations. So there was no attempt to try to
disentangle the effects of all the different possible contributory factors: to isolate in
particular, the contribution of selective social mobility as distinct from particular forms
of deprivation. Yet it is surely only by a painstaking process of elimination that it is
possible to reach any sensible conclusion about the role of “materialist” or “structural”
factors; otherwise the emphasis on these factors is simply a dogmatic, ex cathedra
pronouncement that offers no sort of guide to policy prescription.
The second point is about the dog that did not bark. One of the headline-catching
conclusions of the Black Report, which has since passed into the folklore of the
literature, was that access to Britain’s National Health Service through general practitioners was unequally distributed: that the general practitioner consultation rate of social
classes IV and V was lower than would be expected on the basis of their higher
morbidity-i.e., they did not get their fair share. With a colleague, I challenged this in
1980 by reanalyzing the General Household Survey data used in the Black Report (16).
I was duly, and severely, rebuked by the Black faithful and by Townsend himself in the
introduction to his first edition of the Report (17). But he no longer takes issue on this.
Given his sensitivity to any suggestion of fallibility this silence would be surprising,
were it not that our original findings have since been confirmed in a series of other
studies (18-20). Equity of access to general practitioner services has been achieved,
quantitatively if not necessarily qualitatively. So it seems that even the Black Report did
not get everything right.
Perhaps the saddest aspect of this controversy is that it still revolves around a report
published ten years ago. Concern about the relationship between deprivation and health
had a long history before the Black Report and will rightly continue to attract
researchers for a long time to come. The contribution of the Black Report was that it
gave new salience to the issue; the price paid for this, however, was that it gave
policy-makers an excuse for inaction by overdramatizing and oversimplifying a complex issue. The best way forward now; surely, is to accept that complexity-and to
grope forward toward a deeper understanding by engaging with criticism rather than
dismissing it.
Acknowledgments - I am grateful to my colleagues Raymond Ilsley and Patricia Day
for their comments on the first draft of this article.
REFERENCES
1. Department of Health and Social Security. Inequulifies in Heullh: Report of a Research
Working Group (Chairman, Sir Douglas Black). DHSS, London, 1980.
2. Townsend, P. Widening inequalities of health: A rejoinder to Rudolf Klein. Inr. J. Heulfh
Sew. 20: 363-372,1990.
3. Klein, R. Acceptable inequalities. In Accepfubles Inequalities? Essuys on the Pursuit of
Equuliy in Heulfh Cure, edited by D. Green, pp. 3-20. Institute of Economic Affairs Health
Unit, London, 1988.
4. Szreter. S. R. S. The genesis of the Registrar-General’s social classification of occupations.
Br. J. Sociol. 35: 522-546. 1984.
Making Sense of Inequalities I 181
5. Illsley, R. Professional or Public Health? The Nuffield Provincial Hospitals Trust, London,
1980.
6. Hamnett, C., McDowell, L., and Same, P. The Changing Social Structure. Sage, London,
__
1989.
7. Goldthorpe, J. H. Social Mobility and Class Structure in Modern Britain. Clarendon Press,
Oxford, 1980.
8. Cam-Hill, R. The inequalities in health debate: A critical review of the issues. J. Soc. Policy
16: 509-542,1987.
9. Flinn, M. W. (ed.). Report on the Sanitary Condition of the Labouring Population of Great
Britain. Edinburgh University Press, Edinburgh, 1%5.
10. Whitehead, M. The health divide. In Inequalities in Health, edited by P. Townsend and
N. Davidson, pp. 221-381. Penguin Books, Harmondsworth, Middlesex, 1988.
11. Townsend, P., Philimore, P., and Beattie, A. Inequalities in Health in the Northern Region:
An Interim Report. Northern Regional Health Authority, Newcastle, 1986.
12. M’Gonigle, G. C. M., and Kirby, J. Poverty and Public Health. Gollancz, London, 1936.
13. Barker, D. J. P., and Osmond, C. Inequalities in health in Britain: Specific explanations in
three Lancashire towns. Br. Med. J. 294: 749-752,1987.
14. Illsley, R. Social class selection and class differences in relation to stillbirths. Br. Med. J.
2 1520-1524,1955.
15. Stem, J. Social mobility and the interpretation of social class mortality differentials. J. SOC.
Policy 12: 27-49, 1983.
16. Collins, E., and Klein, R. Equity and the NHS: Self-reported morbidity, access and primary
care. Br. Med. J. 281: 1111-1115,1980.
17. Townsend, P., and Davidson, N. (eds.). Inequalities in Health: The Black Report. Penguin
Books, Harmondsworth, Middlesex, 1982.
18. Collins, E., and Klein, R. Self-Reported Morbidity. Socio-Economic Factors and General
Practitioner Consultations. Centre for the Analysis of Social Policy, Bath, 1985.
19. Puffer, F. Access to primary health care.J. SOC.Policy 15: 293-315,1986.
20. Office of Population Censuses and Statistics. The General Household Survey for 1984. Her
Majesty’s Stationery Office, London, 1986.
Direct reprint requests to:
Rudolf Klein
Centre for the Analysis of Social Policy
School of Social Sciences
University of Bath
Claverton Down
Bath BA2 7AY
England