Copyright Sociological Research Online, 2000


Peter Hodgkinson (2000) 'Who Wants to be a Social Engineer? A commentary on David Blunkett's Speech to the ESRC'
Sociological Research Online, vol. 5, no. 1, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 10/5/2000      Accepted: 24/5/2000      Published: 31/5/2000


This article is a response to a speech addressed to the Economic and Social Research Council which was made, in February this year, by the UK Secretary of State for Education and Employment, David Blunkett. The speech was entitled 'Influence or Irrelevance: can social science improve government?' . Blunkett's programme for engaging social science in the policy process is far from unique and many of the arguments have been heard before. However, the curiosity of the speech lies in the fact that the conception of social science which Blunkett advocates mirrors the approach New Labour itself has to politics and government. This raises some rather interesting difficulties for social scientists. How do we engage in a debate about the role of social scientific research in the policy process when our own conception of the discipline may be radically at odds with that of the government? Furthermore, New Labour's particular conception of the relationship between social and policy-making means that we not only have to contest their notion of what it is we do, but also challenge their conception of the policy process. We cannot ignore this engagement, even if we wanted to. The challenge is to address it and to do so, moreover, in terms which Blunkett might understand. This article is an attempt to start this process.

Applied Social Science; New Labour; Policy; Positivism; Social Engineering; Social Research


The debate on the role of social science in relation to government is as old as social science itself Mills (1959), Coleman (1978), Gouldner (1971), Bulmer (1982), Hammersley (1995)). Since its origins in the Enlightenment, social science has cohabited with government, although the relationship has not always been looked upon by either side as particularly productive. This is because, from the perspective of social science, the demands of government constantly outstrip its ability to deliver the type of knowledge which government requires. Furthermore, social science, for its part, has also outgrown the shackles, which the rather narrow government agenda has placed upon it. Therefore periodically the two sides enter into a round of mutual recrimination followed by an inevitable truce. Then, as surely night follows day, the ritual starts all over again only with new terms of engagement. In other words, the two sides have always existed in something of a symbiotic relationship. In this context, the fact that the Secretary of State for Education, David Blunkett, was to give a speech to the Economic and Social Research Council (ESRC) entitled Influence or Irrelevance: Can Social Science Improve Government? really came as no surprise. The question is duly provocative and of course demands a full response. This commentary attempts to do this. However, in the course of addressing the main arguments put forward by Blunkett it became apparent that there is in fact something new about New Labour's approach to its relationship with social science. In particular, its conception of social science bears an uncanny resemblance to many, if not all, the value assumptions that underpin the New Labour project itself. (although it may well be argued that they do not have a project). Blunkett's speech was not therefore just another call for social science to 'get on message'. It goes further. It is in fact more an expression of the desire of the New Labour government to be 'social science'. In other words, the government sees in one particular mode of doing social science - piecemeal social engineering - a role model for its own political practice. Blunkett's speech can therefore be seen as an inchoate attempt to move towards this model.


Blunkett began his speech by saying :

Some of you may find the title I have chosen for this lecture a little surprising: can the social science community have a major influence in improving government or is it destined to be ever more detached from and irrelevant to the real debates which affect people's life chances?(Blunkett, 2000)

In research terms this might be considered a rather good example of a 'leading question'. Much like the 'when did you stop kicking the cat?' question, one wants to answer it in both the negative and the affirmative. Of course, we are not actually being given a choice. If we answer yes, we are locked into relationship, which is highly problematic and contentious. If we answer no, we may be referring to the first part of the question, rather than the second part, which is rather mischievously constructed in terms of a conflation of 'detachment' and 'irrelevance'. Heads Blunkett wins, tails we lose. As an opening gambit it is obviously designed to put us on the defensive. Is this also a threat? If one is a little paranoid, as surely most of us Sociologists are by now, it can only be read as one. Quite simply, we either knuckle down and help improve his government - a task which may appear rather awesome - or we will not only be cast aside but, far more sinister, also be prevented from engaging in the 'real' debates which matter. It is also interesting to note that Blunkett attempted to make initial contact with his audience by invoking a classic sociological concept, 'life-chances'. Was he, by attempting to speak our language, simply being polite? Or did he not know or recognise the origins of the concept? Did the use of the concept not strike him as rather strange, even perverse? After all, here we have a clarion call to the social sciences to be more engaged with the policy process, being made by using one of the key concepts of a sociologist, Max Weber, who is renowned for his insistence on the separation of (social) science and politics. Was this supposed to be an example of postmodern irony? I fear not. It might however be a good example of the pervasiveness of social science in public discourse - or what Giddens refers to as the 'double hermeneutic' (see below) - and Blunkett, in fact, fails to see that an answer to his question is actually contained in the question itself.

Blunkett's basic contention is that it is 'self-evident' that government policy needs to be informed by 'sound evidence' and, as such, the key role for the social scientist is to help 'decide our overall strategies'. According to Blunkett,

Too often in the past policy has not been informed by good research: a former permanent Secretary once ruefully described the old Department of Education and Science as a "knowledge free-zone"

This is also a rather curious anecdote, since it suggests, by his own admission, that the fault appears to lie with the policy-makers rather than the research - which one assumes was deemed to be 'good' at some time or another. Blunkett nevertheless proceeds to make four specific charges against social scientific research:

Blunkett's idea of the 'research we need' is that which:

i) leads to a coherent picture of how society works: what are the main forces at work and which of these can be influenced by government ...[and]

ii) research to evaluate specific policy initiatives...

We need to be able to rely on social science and social scientists to tell us what works and why and what types of policy initiatives are likely to be most effective.

More categorical still, Blunkett knows what he doesn't want:

We're not interested in worthless correlations based on small samples from which it is impossible to draw generalisable conclusions. We welcome studies which combine large scale, quantitative information on effect sizes (sic) which allow us to generalise, with in-depth case studies which provide insights into how processes work.

Presumably the days of the focus group are numbered? The astonishing thing here is the 'one-size fits all' mentality of the approach to social research. It is as if the whole history of the methodological debates within the social sciences had never happened.

For someone who is interested in evidence-based inquiry, there follows remarkably little in this speech to substantiate each or indeed any of these charges. Hence this plea of not guilty. For example, according to Blunkett, there has been research (not named) which has endeavoured to prove that homework makes no difference to subsequent learning and 'life chances' '...when what most parents (and teachers) want to know is how homework can best help their children'. Presumably Blunkett already knows that homework (all, some?) always helps children (all, some?). It is more than a moot point to suggest that 'proving' something is now, in any case, epistemologically suspect. Instead, it behoves us to actually encourage the 'falsification' (Popper, 1959) of hypotheses - or conjectures - such as that which insists on the efficacy of homework. In this, as in so many other respects, the argument used by Blunkett can be turned back on itself. If the Secretary of State, or whoever wrote this speech, had really done their homework, they would know that the relationship between policy and research is far from being as unproblematic as it might appear. Indeed, if anything, the current state of knowledge suggests that social scientific research cannot deliver better policy - especially in the way that Blunkett conceives it. We can, in other words, falsify the conception of the relationship Blunkett calls for. This does not mean to say however that social science is 'useless' or irrelevant. Quite the contrary. A more informed, mature understanding of the relationship between policy and research could itself be seen as a very 'practical' application of social science.

Knowledge and Control

For Blunkett, the entire enterprise of social science is therefore encapsulated in the following: how does society 'work' and 'how can we best 'engineer' it? And in order to answer both of these questions a certain type of knowledge is called for:

One of our prime needs is to be able to measure the size of the effect of A on B. This is genuine social science and reliable answers can only be reached if the best social scientists are willing to engage in this endeavour [emphasis added].

What is being demanded here is unmistakably a 'positivist' conception of social science. Implicit in this conception is the assumption that 'scientific' knowledge of social conditions can contribute, positively, to interventions that assist the course of human development. It is therefore a conception of social science that entails a particular relationship between the acquisition of knowledge and its potential use. From the origins of social science, with Condorcet, Saint-Simon and Comte, such a positivist view has long held the fascination of social reformers. According to Comte:

'The aim of every science is foresight [prévoyance]. For the laws established by observation of phenomenon are generally employed to foresee their succession. All men, however little advanced, make true predictions, which are always based on the same principle, the knowledge of the future from the past ...Manifestly, then, it is quite in accordance with the nature of the human mind that observation of the past should unveil the future in politics, as it does in astronomy, physics, chemistry, and physiology. The determination of the future must even be regarded as the direct aim of political science, as in the case of the other positive sciences. Indeed, it is clear that knowledge of what social system the elite of mankind is called to by the progress of civilisation - knowledge forming the true practical object of positive science - involves a general determination of the next social future as it results from the past.' (in Kumar, 1978: 23-4)

Hence Comte's famous pronouncement that, 'from Science comes Prevision, and from Prevision comes Control'. It can therefore be argued that the division of labour which Blunkett envisages between social science and government is reflected in the ambition of the positivist conception. That the task of social science is to 'discover' the knowledge required by politicians to enact the necessary interventions. Ostensibly, it might appear that there is no necessary relationship between the two activities. For example, Blunkett makes great play of the fact that politicians e.g. the late Sir Keith Joseph, have in the past 'sought to expunge the very words social science'and carried out 'a deliberate assault on the very existence of the discipline'. By implication this is supposed to demonstrate the autonomous nature of the relationship between social science and political control. However, that these two things - social science and policy - are not simply or contingently related can be shown by looking at the nature of the relationship between causal explanation and prediction. In other words, it can be demonstrated that the search for causal explanations of the nature A causes B, Blunkett's preferred mode of social scientific explanation, necessarily entails a 'will to power', a desire for knowledge which will enable control.

According to Fay (1975), causal explanation necessarily implies prediction and, by extension, the desire to control, because they share a 'structural identity':

'In the case of a prediction the statements of the relevant general laws and the particular facts are given and the statement describing a particular event not yet known to have occurred is deduced from them; in the latter case [causal explanation], the event is known to have occurred, and the general laws and statements about particular facts from which it can be deduced are sought. A prediction is thus simply the obverse of a causal explanation.' (Fay, 1975:33)

Whilst Fay recognises that there are a number of objections to this thesis of structural identity, he maintains that the 'basic truth still stands'. Therefore he goes on to argue that:

'...precisely because scientific explanations are the obverse of predictions, they lay the foundation for the instrumental control of phenomenon by providing the sort of information which would enable one to manipulate certain variables in order to bring about a state of affairs or prevent its occurrence. It is thus no accident that social science positivistically conceived is historically linked with a social engineering viewpoint, for it is in the very nature of the sort of understanding given us by science that it underlies such a viewpoint.' (ibid: 38)

Thus Blunkett's understanding of science has to be seen as intrinsically connected to the very nature of the political interventions which he seeks to enact. In other words, Blunkett's conception of what actually counts as 'science' is that which gives him control. Hence Blunkett's delineation of 'genuine' social science as that which tells him the effect of A on B. Which is why, presumably, the 'best' social scientists are those who are able and willing to measure the effect of (A) 'government' on (B) 'society'! Such is our destiny.

'From the government of men to the administration of things'

In this so-called age of post-modernity, it would not be too difficult to make a case against Blunkett's reaffirmation of the Enlightenment principle that rationality and reason -'science' - be the measure of all things. However, taking Blunkett on his own terms, the wish to understand 'what works and why' actually conflates two quite distinct questions. To say that something 'works' is generally taken to mean that it is an appropriate means to a given end. In which case, the issue is the relationship between means and ends, whereas the 'why' question is concerned with the efficacy of particular means. Is this a case of Blunkett the social scientist manque? Blunkett certainly seems to confuse the business he is in with that of his audience. Politics and government is, if it is anything, all about the former rather than the latter. That is to say, Blunkett's profession is concerned, above all else, with the relationship between goals and means: that policy is the 'operationalisation of values'.

The positivist conception of social science, which Blunkett seeks, had its origins in eighteenth century France and was borne out of the collapse of traditional moral commitments. In the same way, New Labour can be seen to have been borne out of the perceived collapse of the traditional Left/Right antinomies. And just as French Sociological Positivism sought refuge in developing an amoral stance as the moral rule in relation to the Restoration (Gouldner, 1971), New Labour seeks to impose a form of value-neutrality found in social science as its dominant value position in relation to the new world order. New Labour therefore apes Positivism in seeking to suspend the task of radically transforming society until such time as its ability to predict and control the future is attained. Of course, if we have discovered one thing about the nature of the social, it is that that day will never come. The 'pragmatism' of New Labour therefore requires an assault on 'ideology' and 'dogma', 'political philosophy' and even on having an 'essential' political standpoint. It is akin to Comte's attempt to eradicate metaphysics with positive thought. Similarly, in Comte's terms, New Labour can be seen to represent a new, 'positive' stage of development - succeeding the 'theological' and 'metaphysical' stages of its origins - in Socialism and Social Democracy respectively. These former stages must be considered the 'prehistory' of New Labour. Thus, New Labour represents a 'New age', a 'New Britain' and a 'New' Millennium, with Post-Fordism as its material condition and Thatcherism as its political progenitor. And in the manner of the dialectic, New Labour is a synthesis of Old Labour and Thatcherism which were both necessary stages in its evolution. For example, one can see in New Labour's attempt to forge a new moral consensus the rediscovery and repackaging of the moral basis of both its predecessors. This is basically a heady mixture of the moral conservatism that was so evident in patriarchal Welfarism of the post-war state, and the authoritarianism which masqueraded as libertarianism in the New Right of the 1980s. However, the formation of this moral consensus is predicated on its legitimation by advances in secular rationalism and an appeal to 'modernisation' and the future. That is, Order - represented by the imposition of a new moral consensus and authority - has to be reconciled with Progress which, for New Labour, entails being seen to embrace new technology, new institutions, new methods of governance and meritocracy i.e. inequality. Hence Blunkett's task is to complete the process of moving to this new positive stage of development by turning New Labour into social science. In the words of Comte's influential predecessor, Saint-Simon, to move from 'the government of men to the administration of things'. This involves the sublimation of politics and its replacement with technical knowledge. However, like Saint-Simon's rather bizarre attempt to fashion a 'New Christianity' and Comte's 'Religion of Humanity', one can foresee in New Labour the eventual reduction of the role of 'social science' to one of servicing the new moral enterprise, which of course comes complete with its own high priesthood.

Things can Only get Better : New Labour and Social Engineering

There is, therefore, more than an elective affinity between Blunkett's reformist politics and his conception of social 'science'. This is further evidenced by the fact that the use of social science to maintain the status quo, or marginally transform it, is seen as more than just acceptable to Blunkett, it needs to be positively encouraged. Whereas, the use of social science for wholesale re-engineering is seen as unreasonable, 'ideological' or 'utopian', and most definitely dangerous. This is also, of course, a perspective which was famously expounded by Karl Popper in his distinction between 'utopian' and 'piecemeal' social engineering:

'...the social engineer conceives as the scientific basis of politics something like a social technology opposed to the historicist who understands it as a science of immutable historical tendencies.' (Popper, 1962: 22)

Popper argued that social engineers do not concern themselves with history or long-term destiny. Historical forces count for very little. Instead, our ability to transform our condition by collecting information on a rational basis means that institutions can be modelled in our own image. That which is made can be unmade, through the application of the scientific method to political life. Social engineering is therefore seen as directly contributing, in an unmediated fashion, to the process of producing solutions to social problems.

Popper did not rule out rational planning, but it could only be incremental and 'piecemeal'. He rested this argument on the logical premise that the consequences of wholesale or sweeping change are impossible to predict and control:

'At present, the sociological knowledge necessary for large-scale engineering is simply non-existent.' (ibid:162)

As a result, Popper proposes 'piecemeal social engineering' which, whilst concerning itself with social welfare, has no ambition to redesign or transform the entire social system. Instead, it proceeds by small adjustments and continuous fine-tuning. Furthermore, Popper suggested that piecemeal engineering should be aimed at the immediate 'concrete evils' and 'concrete miseries' that confound society, rather than abstract ultimate social ends which are beyond the reach of scientific agreement. How these social evils and miseries were to be determined was however elided by Popper. It was not for the engineer to identify the nature of social problems; that task properly belongs to practitioners or politicians. The social researcher is simply the under-labourer or technician whose task it is to provide the necessary information for understanding the issues. Similarly, he was unable to clearly demarcate what would constitute 'piecemeal' engineering. In other words, he failed to answer questions such as, 'Could a series of piecemeal reforms constitute holistic change?' And one reason for this uncharacteristic uncertainty in Popper is the fact that he disavowed any underlying theoretical framework that would have enabled him to discern the possibility of a relationship between concrete evils. In much the same way, whilst Blunkett identifies a string of 'good' i.e. useful, research projects in his speech, there is no over-arching framework or theorisation of the relationship between the parts. There is no historical destination and no projected future by which we can evaluate the contribution that each project makes to the overall betterment of our social condition. Therefore, when Blunkett asks whether it is:

surprising that applied research useful to government is more common in the USA than here and also carries more prestige relative to theory?

Quite simply, Secretary of State, it is not. It is precisely the lack of an 'ideological' perspective, in fact an insistence on its absence for the most part, which explains why the prolific explorations of US social science fails to add up to effective 'social engineering'. The US has a strong tradition of pragmatism, which involves a more laissez-faire relationship with research. Like all markets, there are strong steers towards work, which is best rewarded. And like all markets, there is also a tendency to provide what the customers think, or are told, they want. This does not always equate with what is actually needed or in the interests of the society as a whole. For all the research into 'social problems' which has been conducted in the US over the years, the amelioration of social miseries still appears to be at a premium. As a result, much of what passes for applied research is facile and undemanding of the politicians who sponsor it. These of course may be the very reasons for Blunkett citing it as a potential model for the relationship between government and social science in Britain.

Blunkett is therefore part of an administration which, while it is calling for 'joined-up' government (a euphemism for centralisation?), at the same time advocates more 'piecemeal' social science. The ruling out of holistic social engineering in favour of incremental, measured-and-tested approaches also fits comfortably with Blunkett's conception of 'science' as the accumulation of knowledge. (And this is somewhat ironic, given that Popper would have been one of the first to deny that knowledge develops through a process of accretion.) Piecemeal reform is therefore an adjunct of New Labour's rather touching belief 'that things can...' in the words of their own campaign anthem, 'only get better'. In both Popper's and now Blunkett's estimation, 'progress' is inevitable and the task of social science is to ease the path to the better future by careful and necessarily relatively minor alterations to the general course of human development. Therefore when Blunkett claims that:

This Government has given a clear commitment that we will be guided not by dogma but by an open-minded approach to understanding what works and why.

He is claiming, not just that he will be guided by reasoned argument, but that his approach is in fact more 'scientific'. Hence his charge that:

sometimes, when it [social science] does try to be directly relevant to the main policy and political debates, ..[it is] seemingly perverse, driven by ideology paraded as intellectual inquiry or critique...

The assertion of open-mindedness by Blunkett has therefore be seen to 'work', discursively, as both a claim to legitimate his 'scientific', as opposed to 'ideology-driven', approach to government and as a justification for 'piecemeal', incremental reform rather than holistic approaches to social engineering. And on both these counts it can be seen to be wanting. Science is not now generally thought to be an ideology free-zone, and the adoption of piecemeal reform as the only approach to addressing social evils is not open-minded or 'scientific'. It is in fact dogmatic and may be extremely limiting in terms of the amelioration of social problems.

Getting Real and Doing your Homework

Blunkett also accuses social scientists of failing to be sufficiently 'street-wise' about their approach to, and drawing of conclusions from, research. He suggests that we have sometimes not taken into account the 'reality of people's lives'. Without accepting or even denying either of these charges, it cannot go without comment that, for a professional politician, Blunkett has a remarkably naive conception of the relationship between policy and research. Indeed, it might be argued that it is he who fails to take the reality of the policy process into account and it is he who has a not particularly 'street-wise' understanding of the ways in which research informs policy. Blunkett appears to conceive of the relationship between policy and research in terms of the 'rational' or 'linear' model. That is, the policy process involves a linear succession of stages:

In the policy analysis literature this model is universally rejected as having any basis in reality (Weiss, 1979). It is an idealised model of both the policy process and the role of social research. Furthermore it is one which does not accord with how government officials actually make use of research. Rather than provide specific problem-solving information or solutions to policy issues, research has been shown to be employed by policy-makers more as a foundation for formulating ideas, concepts and strategies. At best, research provides an orientation and a means of conceptualisation, which enables policy-makers 'to think' of issues. At worst, it is used to legitimate existing policy action - or inaction (Tizard, 1990). Weiss (1983) suggests that:

'...public policy positions taken by policy actors are the resultant of three sets of forces; their ideologies, their interests (e.g. in power, reputation, financial reward) and the information they have' , and the third of these, to which social science is but one contributant, is usually outweighed by the first two, which carry 'higher emotional loadings.' (Weiss, 1983:221-2)

It is widely recognised in the literature that the policy process is inherently 'political' and can only be understood as an interaction between policy makers and 'scientists' (and indeed, other interested parties)(Lindblom, 1979; Tizard, 1990; Booth, 1988; Weiss, 1979; Rein, 1980). All these authors see the outcomes of the policy process as products of negotiation and bargaining between policy-makers and researchers, to the extent that :

'...the very 'facts' themselves may be the outcome of the overall processes of mutual accommodation between all parties, scientists included.' (Bartley, 1996: 23)

Bartley's work is within a 'social constructionist' perspective, which argues that what is taken as valid knowledge is itself, the outcome of social negotiation, especially between the disciplines, professionals and the state. Bartley's examination of the 'micro-politics' of the relationship between policy and research in the area of unemployment and health shows quite convincingly the extent to which:

'...'social-problem processes' are central to policy-making, and usually involve the making of both knowledge-claims and value-claims by various groups pursuing a range of interests. Each process may go on to give rise to either new facts, or new policies, or both or neither. But in many cases, it is the success of value-claims in changing policy which leads to the acceptance of knowledge-claims as fact, rather than the other way round.' (ibid:.20)

That politicians and policy-makers pander to their own idols, Blunkett recognises only too well:

I too may let my prejudices over-ride the legitimate empirically based evidence (politicians have a tendency to believe research when it reinforces their own view).

Quite apart from the political realities of wanting to get re-elected, there are in fact a raft of other reasons which have been found to explain why research does not impact on policy in the manner which Blunkett's linear model suggests. For example, it is well-known that policy-makers adhere to Departmental traditions, customs and rules: officials do what the agency has always done. That rather than seeking new approaches or solutions, officials will often simply adapt that which they know and have tried before. Policy is also made as a consequence of trading between Departments and 'mutual adjustment'. Favoured solutions can also precede problem definition and policy is made on the basis of finding opportunities to enact ones preferences. None of this is new knowledge. Therefore Blunkett's failure to even recognise the real problems associated with the role of research in the policy process has to be seen as a display of either complete ignorance or, more likely, contempt for the evidence (not to mention his audience). From the actual content of this speech, which we must assume was largely compiled by his officials, it appears that the DfEE is still a largely 'knowledge-free zone'. This failure to take into account the evidence relating to the role of research in the policy process is itself confirmation of the findings of research on the topic!

'Do as I say and not as I do'

Blunkett is fully aware that an 'anti- intellectualism' has pervaded the corridors of Whitehall and has contributed to the denigration of social research in the policy process:

Previous administrations have not been prepared to be receptive to what research can tell them. There has, in the past, been a seam of anti- intellectualism running through government both at the political level and amongst officials which has served to alienate academia. I acknowledge that we still have some way to go before we can overcome this mind set within government and the defensive reaction it inevitably provokes. This has been one of the reasons research has itself become too inward looking.

What evidence has Blunkett for this last assertion? An equally plausible explanation could be that social researchers 'turned away' from policy related research because of the selective nature of its appropriation rather than its general dismissal. Social researchers may also choose not to research policy areas when they have fundamental doubts regarding the value-base of certain policies. Thus the reason why many researchers turned away from policy research - to become more 'inward looking' - under the Conservatives may not have been the anti-intellectualism of government, as much as their wish not to be associated with a government and policies which they found unpalatable. No doubt many will feel the same way about the current regime , especially when it insists that:

Policy makers need to make clear for researchers what they see as the key questions and why these are interesting and important.

In the same breath, Blunkett speaks about the building of 'trust' between researchers and policy-makers, as well as 'partnership' and 'interchange'.Thus Blunkett goes on to argues that:

Researchers can also help by being more open and balanced about the strengths and potential weaknesses of their data. As I said earlier, whether they are inside or outside government, researchers should make sure they are not seduced by the demands of the press into reducing complex and qualified conclusions into a set of simplistic and misleading messages which can do much harm.

Coming from one of the high priests of the 'government of spin', this does sound rather hollow. Of course, in the interests of building 'trust' one would expect to be able to juxtapose the terms 'researcher' for 'Secretary of State' in the above quote. Unfortunately, he fails to heed his own warning and proceeds to cite the following 'evidence' to support his claims :

For researchers [ not named] to tell my constituents that perpetrators of vandalism, neighbour nuisance, all night parties, and health destroying noise and intimidation, are really the victims rather than the perpetrators does very little for the public credibility of social science!
some research [ not named] ..reminds me of the investigations a few years ago by a certain newspaper which, perversely, set out to prove that there was no such thing as AIDS!

Amidst the sound-bite nature of this criticism there is the rather serious and substantial point that social science has not been particularly successful at disseminating its relevance and potential contribution to current debates on social issues. However, Blunkett's argument is undermined by the fact that opportunities for researchers to disseminate their findings (criticisms) relating to government sponsored research are, under New Labour, becoming more limited by the day. Evan Davies, the Liberal Democrat education spokesperson has recently conducted a survey of restrictions on the reporting of research which government departments have sponsored. He has been reported as finding that:

(Cohen, 2000)

Putting Science Back in after the Social has Bolted

One of the basic problems with Blunkett's conception of social science is that, not only does he have a rather outdated notion of 'science', he also thoroughly misunderstands the nature of the 'social'. More than that, he has no conception whatsoever of the alternative tradition, interpretivism, and its centrality to debates within social science. Perhaps the most ready source of insight into the interpretive tradition for Blunkett would be a seminar with one of the few sociologists who appears to have (had?) the ear of the Prime Minister, Anthony Giddens.

Giddens argues that the social sciences are unlike the natural sciences in so far as they stand in a very different relation to their subject matter. According to Giddens (1984), the natural sciences have a 'technological' relation with their objects of study, such that they enable the information they generate to be used as a 'means' to transform the selfsame world of objects and events. Furthermore, the relationship between the findings of the natural scientist and lay beliefs is such that the latter has few, if any, consequences for the former. Whereas the social sciences have both more than, and at the same time less than, a 'technological' relation to their subject matter. More than, in the sense that human beings create the object of social scientific inquiry, the social world, by and through their knowledge and social skills. And less than, in that 'technological' information does not provide the insights into the nature of the social world in terms which lay persons readily comprehend and can use in a practical sense. Concepts such as 'life-chances' do however filter into the construction of social worlds beyond those of the social scientist. Even David Blunkett finds himself using this term to frame an understanding of the efficacy of homework with respect to: '... subsequent learning and life chances....'. The permeation of such concepts into the world, which is the object of inquiry of the social scientist, and its reflection back into social scientific explanation is what Giddens refers to as the 'double hermeneutic'. It is this type of interaction which separates the natural from the social sciences and undermines the very possibility of the type of explanatory statements which Blunkett demands. Giddens would therefore be more than capable of explaining to Blunkett that much of the disillusionment with social scientific research in terms of its relevance to policy making, lies at the door of those, like Blunkett, who have misconstrued the nature of both natural and social science. That social science can never achieve the level of generality and 'objectivity' of the natural sciences has not been understood. Instead, we have to contend with the double hermeneutic and a lack of a unified conceptual structure. And it is this which can be seen to account for Blunkett's frustration with social science which fails to conform to his positivistic conception.

Knowledge for its own Sake

Having extolled the virtues of the 'genuine' social science and having all but ruled out anything which does not relate to immediate policy concerns, Blunkett feels that he needs to assuage those (Anthony Giddens?) who might think that he has no place for 'theory':

I have put a great deal of emphasis so far on research with practical applications. But if we are to encourage a more open debate of ideas there must also be a place for the fundamental 'blue-skies' research which thinks the unthinkable. We need researchers who can challenge fundamental assumptions and orthodoxies and this may well have big policy effects much further down the road. Civil servants, politicians, journalists and others do not have the time for this kind of work. If academics do not address it, then it is difficult to think of anyone else who will. We must recognise its importance.

Even in the midst of 'blue-skies' Blunkett cannot let go of his instrumentalist '...big policy effects further down the road'...approach to assessing everything we do. Of course, if we academics are not in bed with policy-related research, we should have the time to lie back, peer skyward and contemplate the 'unthinkable' e.g. not all knowledge is 'useful' or policy relevant? And we must 'recognise its importance' - presumably by mentioning it, as Blunkett does, in one paragraph out of 63 in this speech! This type of 'blue-skies' research appears to conform to what Hammersley (1995) identifies as the 'disciplinary model' of social research:

'Here the goal of research is to contribute to knowledge in a particular discipline, with abstract theoretical knowledge being given priority. While research is seen as ultimately making a contribution to practice, that contribution is not intended to be very immediate or specific. What is involved is the production of general- purpose knowledge, which is valued as much for its own sake as for any instrumental value it has. Findings are simply put into the academic public domain for others to use: as, when and if they wish.' (Hammersley, 1995:125)

Thus the central Enlightenment assumption that a rational comprehension of the social world is possible still underpins this model, but without the onus of showing, as in the Engineering model, how that knowledge directly informs the policy process. However, the negative aspect of this model is that there is no immediate return on the investment i.e. practical payoff. Hammersley argues that the castigation of 'theorising' is due to the instrumentalism which informs the dominant understanding of the theory/practice relationship. That is to say, theorising is seen as having no value in and of itself. It must be seen in relation to its 'other', practice, and the possibility this has for social transformation. As a result, instrumentalist thinking declares that the pursuit of knowledge for its own sake is both pointless and decadent. Hammersley however makes a strong case for the disciplinary mode :

'To claim that knowledge can be of value for its own sake is not to imply that any knowledge is of value, or that we should value knowledge according to its uselessness. It means that we should value knowledge about matters relevant to human life above ignorance about such matters. And such knowledge has value in terms of its relevance to our lives irrespective of whether it can be shown to have a direct and powerful influence on our practical activities. In other words, it is not to be judged, and the pursuit of knowledge does not need to be guided, by some sort of calculation of the degree of instrumental value it has. The fact that the practical benefit of disciplinary knowledge is not calculable does not make it worthless.' (ibid:142)

Therefore Blunkett's association of policy- unrelated research with the 'micro', narrowness and esotericism has to be seen as quite deliberate. The very possibility that social research which fails to address policy can still be 'useful' i.e. valued in some way, needs to be discounted if Blunkett's whole case regarding 'relevance' is to stand scrutiny . And, no doubt, we should be able to surmise from this that instrumentalist knowledge is, by its very nature, both more 'macro' and accessible. Of course the logic of this argument is deeply flawed. This conception of research utility does however appear to mirror quite accurately the 'populist' tendency of New Labour. Just as books are to be judged by their covers and policies by focus groups, research should be judged by its scale and/or utility to policy-makers. Blunkett's ire can then, in his view, be legitimately aimed at those researchers who have 'turned away' from policy-related research:

preferring to work on questions of little interest to those outside the research community. There is a danger of too much concentration on the micro level - what is the point of research which becomes narrower and narrower, with small groups of people responding to each other's writing in esoteric journals, read only by themselves and relevant only to themselves? This is a dangerous turning which we must try and address.

One might well ask of Blunkett, how does he think new areas of social research actually begin? How are new fields of research ever to become established if they are not to be seeded in minority interests? How would any minority interest area ever be studied given his rules of engagement? It is perhaps particularly distressing to hear this demand for research to be justified on grounds of its populist appeal coming from someone with direct experience of social exclusion. When one thinks how difficult it is to find funding for research on issues which have only a 'minority interest', the criticism of concentrating on the 'micro level' and on research which has a narrow interest is quite worrying. If the writing is esoteric and the journals are only being read by the minority interest, it would seem that, under New Labour, there is no warrant for public support. Taken to its logical extreme, one would expect next to find Blunkett arguing that libraries should be emptied of books which are not best-sellers! In fact the 'dangerous turn' here is the attempt to 'technicise' politics, to turn it into the administration of things, because it is precisely this approach which will narrow political engagement and participation. Politics will indeed come to resemble small groups of people responding to each other in esoteric language, addressing only themselves, on matters which most people neither understand or care about.

Reversing the Charges

It would be in no-one's interest to deny the positive contribution that social science could and should make to public policy making. Most people would prefer evidence-based policy to 'dogma'. However, before 'social science' contributes to improving government, it has to be recognised that it encompasses a range of approaches that do not conform to the positivist conception held by Blunkett and others. As such, the contribution cannot and should not be that envisaged by the Secretary of State. Therefore the charges which Blunkett levels at social science are misguided and unhelpful in establishing a new rapprochement. If New Labour wants to be a 'thinking government' it needs to be more reflexive and heed the following:


Thank you to friends and colleagues for all their advice and encouragement - especially Ann Lahiff, Dave Phillips and Michael Marinetto.


BARTLEY, M., (1996) 'Probably, Minister...' the 'Strong Programme' Approach to the Relationship between Research and Policy', in Samson, C., and South, N., (eds.) The Social Construction of Social Policy, London: Macmillan.

BLUNKETT, D., (2000) 'Influence or Irrelevance: can social science improve government?', in Research Intelligence, British Educational Research Association, No.71, March 2000.

BOOTH, T., (1988) Developing Policy Research, Aldershot: Avebury.

BULMER, M., (1982) The Uses of Social Research, London: Allen and Unwin.

COHEN, N., (2000) 'With our money, they hide the truth', New Statesman, 20th March,pp17-19.

COLEMAN, J., (1978) Sociological Analysis and Social Policy, in Bottomore, T., & Nisbet, R., (eds.) A History of Sociological Analysis, London: Heinemann.

FAY, B., (1975) Social Theory and Political Practice, London: George Allen and Unwin.

GIDDENS, A., (1984) The Constitution of Society, Cambridge: Polity Press.

GOULDNER, A., (1971) The Coming Crisis of Western Sociology, London: Heinemann.

HAMMERSLEY, M., (1995) The Politics of Social Research, London: Sage.

KUMAR, K., (1978) Prophecy and Progress: the Sociology of Industrial and Post-Industrial Society, London: Allen Lane.

LINDBLOM, C., (1979) 'Still Muddling, not yet Through', Public Administration Review, 39: pp.517-26.

MILLS, C.W., (1959) The Sociological Imagination, New York: Oxford University Press.

POPPER, K.R., (1959) The Logic of Scientific Discovery, London: Hutchinson.

POPPER, K.R., (1962) The Open Society and its Enemies, Vol. 1: Plato,London: Routledge and Kegan Paul.

REIN, M., (1980) 'Interplay between Social Science and Social Policy', International Social Science Journal, 32 : pp. 361-8.

TIZARD, B.,(1990) 'Research and Policy: Is there a Link?' The Psychologist, 10 : pp.435-40.

WEISS, C., (1979) 'The Many Meanings of Research Utilisation', Public Administration Review, 39: pp.426-32.

WEISS, C., (1983) 'Ideology, Interests and Information; the basis of policy decisions', in Callahan, D., and Jennings, B., (eds.) Ethics, the Social Sciences, and Policy Analysis, New York: Plenum Press.

Copyright Sociological Research Online, 2000