Keynes, Beveridge and Beyond

Tony Cutler, John Williams, Karel Williams

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781136521287
• Edition: 1
• Topic: Business
• Subtopic: Business General
• Index: Business

Presenting a coherent interpretation of the development of economic and social policy in Britain since 1945, this book analyses the political assumptions underlying post-war economic policy. It traces these assumptions through the classic texts of Keynes and Beveridge, the architects of limited, non-socialist state intervention to secure the welfare state and full employment. Topics covered include:
* 'Private saving' versus company pensions
* The level and composition of employment in Britain

1 Beveridge and Keynes: defining liberal collectivism

This chapter compares two key texts by Beveridge and Keynes. The analysis of the social in Social Insurance and the Allied Services and the theory of the economy in the General Theory are, of course, different in many respects. Our thesis is that, despite these differences, the key texts are similar in that they share a common political a priori. Before we establish and develop this thesis, it is useful to begin by outlining the very different character of the two texts.

Social Insurance and the Allied Services (Beveridge, 1942, hereafter SIAS) was a government White Paper on what would now be called ‘the reform of social security’. It was resolutely empirical in its approach: ‘The plan is based on a diagnosis of want. It starts from facts, from the condition of the people as revealed by social surveys between the wars’ (SIAS, p. 8). Beveridge culled two main conclusions from these local surveys of social conditions in London and some major provincial towns. First, a substantial minority of the working-class population lived in ‘want’, below the subsistence poverty lines set in these surveys. Second, there was one major cause of such poverty; three-quarters or more of it was caused by ‘interruption or loss of earning power’ (SIAS, p. 7).

Beveridge’s report proposed to abolish such poverty; ‘the Plan for Social Security … takes abolition of want after this war as its aim’ (SIAS, p. 8). SIAS also envisaged that ‘the provision of an income should be associated with treatment designed to bring the interruption of earnings to an end as soon as possible’ (SIAS, p. 120), and the corollary of treatment was most strongly developed in Beveridge’s proposal for a comprehensive National Health Service (SIAS, pp. 157–63). But, given the emphasis on the income deficiency of those below the poverty line, it was quite logical that Beveridge was primarily concerned with proposals for income maintenance.

Beveridge’s position on state income maintenance was distinctive since he envisaged that ‘compulsory social insurance’ could and should do most of the work. The pre-1914 Liberal reforms had introduced state health and unemployment insurance which were then developed in the inter-war period. SIAS proposed to abolish poverty by raising insurance contributions and paying out substantially higher benefits; ‘social insurance should aim at guaranteeing the minimum income needed for subsistence’ (SIAS, p. 14). At the same time, SIAS proposed a more comprehensive social insurance scheme which would eliminate the gaps in the existing schemes. The whole population (including self-employed, high-paid and dependants of various kinds) would be covered in the new scheme (SIAS, pp. 9–10). All the major misfortunes and contingencies which constituted ‘primary causes of need’ would be covered by some kind of insurance benefit; the list included unemployment, disability, loss of livelihood, retirement, marriage needs, funeral expenses, childhood, physical disease or incapacity (SIAS, p. 125). The basic insurance benefits were to be indefinite in duration; a weekly cash allowance would be paid as long as the interruption of earnings lasted (SIAS, p. 11, pp. 57–9). None of the insurance allowances were to be means-tested; a weekly cash allowance would be paid automatically to all who could establish their status as sick, unemployed, etc. and the fact of contribution (SIAS, p. 120).

The General Theory (Keynes, 1936, hereafter GT) is a text on what Keynes in the preface called ‘difficult questions of theory’. The policy recommendations are underdeveloped and take the form of throw-away asides, as in the concluding paragraphs of chapter 10, supplemented by the ambiguities of the famous final chapter 24. Most of the text elaborates a ‘pure theory of what determines the actual employment of the available resources’ in a (more or less) closed capitalist national economy. Such theory necessarily addresses abstract questions at a very rarefied level and in the case of the GT this tendency is reinforced by an explicit retreat from the realm of the observable and measurable. In chapter 4, for example, Keynes insists that a theoretical understanding of the variation in employment and output is possible although he cannot solve practical problems about conceptualising or measuring the quantity of capital or the general price level. Or, again, Keynes makes new investment depend on the ‘marginal efficiency of capital’ (business expectations of prospective yield). In many ways this is the key independent variable in the Keynesian system and, significantly, it is not measurable or directly observable.

Keynes offers a theory of demand-led fluctuations in employment and output which explains how aggregate effective demand is usually deficient and why consequently unemployment is the norm in capitalist national economies. This theory of deficient demand analyses demand into two components – consumption and investment expenditure. Consumer expenditure is chronically deficient because the marginal propensity to consume is less than one; ‘when aggregate real income is increased, aggregate consumption is increased, but not by as much as income’ (GT, p. 27). This kind of steady-state underconsumption requires the compensation of new investment expenditure if aggregate demand sufficient for full employment is to be maintained. But new investment expenditure does not steadily compensate because it switches on and off according to how the marginal efficiency of capital (or volatile business expectations of profit from new investment) fluctuates above and below a conventional rate of interest. To sum up,

a relatively weak propensity to consume helps to cause unemployment, by requiring and not receiving the accompaniment of a compensating volume of new investment which, even if it may sometimes occur temporarily through errors of optimism, is in general prevented from happening at all by the prospective profit falling below the standard set by the rate of interest. (GT, p. 370)

The presentation of the GT’s theory of unemployment is complicated because the text is written partly in the form of a critique of ‘classical economics’. The main theoretical object of attack is the marginalist theory of unemployment (as developed in the 1930s by A. C. Pigou) which attributes the phenomenon of unemployment to sticky money wages. On this view, unemployment exists because organised workers have demanded and obtained a wage which is greater than their marginal product justifies. The GT makes a double attack on this marginalist explanation. Negatively, in chapters 2 and 3, Keynes dissects and criticises the logical confusions and conceptual imprecisions of the marginalist explanation of unemployment. Positively, Keynes advances an alternative explanation which attributes unemployment to deficient aggregate demand which arises, as we have seen, because a general propensity to underconsumption is not steadily compensated for by new investment. As chapter 19 makes clear, this alternative explanation cannot be reduced to marginalism because the causal relations between the old marginalist variables are changed at the same time as the Keynesian theory introduces new variables which are not conceptualised in the marginalist scheme (GT, p. 278). Thus, in the GT, real wages are identified as a dependent variable proximately determined by the volume of employment, while Keynes introduces new system-determining independent variables in the form of the ‘marginal propensity to consume’ and the ‘marginal efficiency of capital’.

The accounts of the GT and SIAS so far only demonstrate the radical difference of the two texts. It will now be argued that the texts are similar insofar as they share a common political a priori. Beveridge and Keynes are separately engaged on the common task of re-inventing liberalism; they reject the old liberal laissez-faire position of general hostility to state intervention in a market economy. With the presumption against state intervention suspended, the task of the new liberal is to make a reasoned demarcation of the proper sphere of state action. In making this demarcation, liberal collectivists like Beveridge and Keynes take up certain characteristic positions which constitute the fundamental features of the a priori.

(i) Certain basic conditions of social life have to be ensured. Poverty and economic insecurity, for example, are dysfunctional and may ultimately threaten the existence of the capitalist order.

(ii) The maintenance of these conditions of social life is impossible on the free market if a policy of strict laissez-faire is pursued. The control of poverty, for example, requires some form of state social security and so does the curbing of insecurity.

(iii) State intervention is not undesirable but imperatively necessary. This is because norms on the basic conditions of social life cannot be maintained without some form of state intervention.

(iv) State intervention should be minimal and confined to what is necessary. Intervention should be circumscribed in nature and extent so that there is as little interference as possible with the valuable political and economic freedoms of capitalism.

(v) State intervention should be directed so that it suppresses basic capitalist freedoms only so as to re-establish the conditions of their effective functioning (elsewhere). Classically, for example, with state regulation established in one sphere of the economy, unregulated market choice can flourish elsewhere.

Ultimately, the new liberal position is distinctive because it recommends large slabs of collectivism in the interests of preserving as much as possible of capitalist individualism. The liberal collectivists accepted the traditional philosophical justification of capitalism as a political and economic system whose virtue is that it leaves a man free to do what he wills with his own. At the same time, the liberal collectivists accepted that the consequences of this freedom on the market were entirely unacceptable and these consequences would have to be curbed by state intervention in some respects if capitalist freedoms were to survive more generally.

With the main features of the political a priori sketched, the next major task is to demonstrate that this a priori runs through the problem definitions and policy recommendations in SIAS and the GT. Although the a priori has been defined above mainly in terms of policy recommendations, it can only be established in the economic or the social on the basis of an appropriate problem definition; the initial criticism of the capitalist system has to be limited so that the problem which intervention has to solve is not fundamental to, and inherent in, the capitalist order. As for policy recommendations, here it is simply a question of tracing how the texts propose politically congenial interventions which are the minimum necessary to deal with the diagnosed problems.

The political a priori in SIAS

As we have noted, Beveridge’s aim was to abolish poverty. In proposing this aim, SIAS depended on an object and on an operational measure of poverty which were borrowed from the inter-war social surveys. Both the object and the operational measure are of considerable political significance.

Beveridge needs the social scientific object of poverty because it demarcates a limited sphere of state action. Poverty focuses attention on the issue of a deficiency of income at the bottom end of the range of distribution and it diverts attention from the issue of inequality and the whole distribution of incomes from top to bottom. More specifically, poverty allows Beveridge to take up a series of classic liberal collectivist positions. Poverty represents a problem of maldistribution of income which cannot be rectified on the free market through the mechanism of rising real incomes,

growing general prosperity and rising wages diminished want, but did not reduce want to insignificance. The moral is that new measures to spread prosperity are needed. The plan for social security is designed to meet this need. (SIAS, p. 166)

At the same time, poverty sets up a distributional problem which can be rectified without upsetting the whole range of income inequalities. Indeed, according to SIAS, inter-class redistribution would be diversionary,

correct distribution does not mean what it has often been taken to mean in the past – distribution between the different agents of production, between land, capital, management and labour. Better distribution of purchasing power is required among wage earners themselves, as between times of earning and not earning and between times of heavy family responsibilities and of light or no family responsibilities. (SIAS, p. 167)

Beveridge quite explicitly stopped short of recommending that redistribution should be confined to the working class but he could not resist making the point that redistribution could be confined to the working class; the social surveys of the 1930s showed that the income surplus of working-class families above the poverty line was very much larger than the income deficit of those below the poverty line and the implication was that ‘want could have been abolished before the present war by a redistribution of income within the wage-earning classes’. (SIAS, p. 165)

To sustain the poverty problem definition and carry it through to policy recommendations, Beveridge needed an appropriate operational measure of poverty. SIAS therefore privileged the kind of calculation of minimum necessary income which was pioneered by Rowntree in his 1899 York Survey and which was then reworked by Rowntree and others in the inter-war poverty surveys. In the poverty surveys this calculation defined a poverty line below which households lacked the minimum necessary income; in SIAS the calculation specified a target up to which the state income maintenance system would provide support.

The necessary income calculation costed rent, food, clothing, heating, household necessaries and sundries and then added a margin to allow for inefficiency and waste. The calculation left major problems unresolved. It did not represent a bare subsistence minimum because the calculation made concessions to cultural patterns of consumption which pushed up expenditure; in SIAS, for example, expenditure on food was based on League of Nations and British Medical Association dietaries which included meat although vegetarianism was a more economical alternative for the working class. But the concessions were never taken to the point where they accommodated actual consumption patterns. This is most clear in the SIAS treatment of rent and ‘waste’ where the concessions are arbitrarily limited. Beveridge proposed a rent allowance (10/- or 50p a week in 1938 prices) which was pitched slightly below the average for all industrial households as revealed in a Ministry of Labour survey of 1937–8. SIAS ignored the marked local and regional variation in the rents actually paid; in the highest rent area of London industrial households paid rents which were over 50 per cent higher than the national average (SIAS, p. 78). The treatment of ‘wasteful’ expenditure was equally striking. SIAS proposed an allowance of 2/- (10p) a week for a man and wife in 1938 prices to cover inefficient purchasing and diversion of expenditure on to non-essentials (SIAS, p. 87). This was rather half-hearted when the average working-class family in the 1930s spent 7/- (35p) per week on drink alone (Rowntree, 1942, p. 493).

As an attempt to define an adequate minimum income level, Beveridge’s calculation is thoroughly incoherent and unsatisfactory; SIAS never establishes that it is possible to provide realistically for a family on such an income. But to abuse the calculation on this account is to miss the point because the calculation has a political rationale. The procedures of the poverty surveys do not function as a scientific point of departure but as a politically congenial point of support for liberal collectivism. A (scientifically unspecifiable) minimum income level is politically important to SIAS because it defines the target at which state income maintenance should be aimed.

The question of how the state reaches this target is equally important. Indeed, the political a priori is most strongly developed in Beveridge’s discussion of the form which income maintenance should take. As we have already noted, SIAS envisaged a social security system which was dominated by ‘compulsory social insurance’. Non-contributory assistance was to be a ‘subsidiary method’ (SIAS, p. 103) which would deal with the relatively small and diminishing number of ‘special cases’ (SIAS, pp. 120–1) who did not qualify for insurance benefit. Beveridge was an insurance freak of long standing whose enthusiasm is epitomised by the title of his 1924 pamphlet, ‘Insurance for all and everything’. Why did SIAS recommend the social insurance technique?

Compulsory social insurance was fiscally expedient because it represented a way of recovering ‘a substantial part of the cost of benefit as a contribution’ (SIAS, p. 12). The alternative was a non-contributory social security system which was financed entirely o...

Table of contents

• Cover
• Title Page
• Copyright Page
• Table of Contents
• Note and acknowledgments
• Introduction
• 1 Beveridge and Keynes: defining liberal collectivism
• 2 ‘Private saving’ versus company pensions
• 3 The level and composition of employment in Britain
• 4 New terrains of policy
• Bibliography
• Index