Income in Interwar South Wales

Chapter 1

Obtaining an income: Wages, benefits and household strategies. Chapter 1: Income

His father’s toil had been so excessive as to make him stoop like a victim of curvature. That had been just as well, because his father’s wages were so low it would have been impossible to count them standing up straight. It had always been necessary to get them very close to the floor to get them in the right perspective. Gwyn Thomas.1

1 Gwyn Thomas, Sorrow For Thy Sons (London, 1986), p.67.

18 Chapter 1: Income

In any study of interwar Britain, economic depression and its attendant consequences, unemployment and poverty, understandably occupy a central position. Dole queues, soup kitchens, hunger marches, rickety children and hollow eyed adults have dominated popular perceptions of these ‘wasted years’. And yet, as revisionist historians have pointed out, the interwar period was also characterised by impressive economic growth, the development of new light industries and the emergence of new patterns of consumption. While the presence of poverty and hardship are undeniable it has also been pointed out that the unemployed never constituted more than a minority of the British population during the whole of the period. For those in work the interwar years were a period of rising living standards as real wages increased and as a whole range of consumer durables came within the purchasing power of ordinary people.2 If this central paradox of the interwar years has served to enliven British historiography it has occasioned much less controversy among Welsh historians. As K. O. Morgan wrote in 1981: Some historians of late have tended to paint a more cheerful picture of the thirties than once used to be prevalent . . . in south Wales, this verdict cannot possibly be accepted. The thirties were a time when a whole society was crucified by mass unemployment and near-starvation.3

The length and severity of economic depression, the massive volume of unemployment and the dire poverty occasioned by it seems to have led Welsh historians to offer a reductionist explanation of the interwar years that sees a whole society ‘crucified’. Histories of interwar south Wales are, more or less, histories of the unemployed and the poor. Certainly there is very little reason to be ‘optimistic’ about south Wales in the interwar period but it remains the case that even in the deepest trough of the depression in 1932 55% of the population of Monmouthshire and almost 60% of that of Glamorgan were in employment.4 For the greater part of the period, approximately 2 C. L. Mowat, Britain between the Wars 1918-40 (London, 1955); D. H. Aldcroft, The Inter-war Economy: Britain, 1919-1939 (London, 1970); J. Stevenson and C. Cook, Britain in the Depression: Society and Politics 1929-39 (Essex, 1994 edition). 3 K. O. Morgan, Rebirth of a Nation: Wales 1880-1980 (Oxford, 1981), p.230; see also Gwyn A. Williams, When Was Wales? (London, 1985), p.252. 4 Ministry of Labour, Local Unemployment Index, 1932. Annual average based on 12 monthly averages. A similar observation is made by John Davies, A History of Wales (Harmondsworth, 1994), pp.559-560; for an insight into the relative incomes of employed and unemployed families in interwar south Wales see the diagram in H. Jennings, Brynmawr. A Study of a Distressed Area (London, 1934),

70-80% of the population of south Wales was employed and yet Welsh historians have almost completely overlooked this experience. Perhaps, as Glynn and Oxborrow have observed, historians do not need to arbitrate between ‘optimistic’ and ‘pessimistic’ interpretations but present the interwar period in all its complexity and paradoxical nature.5 The experience of employment was a significant one and an examination of wages and incomes is relevant to a study of the quality of life in interwar south Wales. Many historians have described the broad nature of changes in wages and incomes in interwar Britain.6 Large rises in money wages in the period 1915 to 1920 meant that by the end of 1920 average salaries in Britain were about double what they had been in 1914. Weekly wage rates had risen 2.7 or 2.8 times for a working week that, in 1919, had been reduced to between 44 and 48 hours from its former range of 48 to 60 hours. There followed a collapse in wages with a fall in standard wage rates of 8% in the period December 1920 to December 1921 and of 22% in the following 12 months. The average decreases in weekly wages recorded by the Ministry of Labour in 1921 were 17s.6d. per week for 7.2 million workers and in 1922 a further 11s. for 7.6 million. Reductions were most severe for coalminers, 1.3 million of whose wages declined on average by £2 per week in 1921 and a further 10s. in 1922. Nevertheless, average wage rates in 1924 were still twice those of 1914. Wage rates rested at this point until 1930-31 when a moderate decline occurred until 1934, after which a slow rise occurred so that by 1938 wage rates were 5-6% higher that the 1930 level and about double the pre-war level. These variations in wages and earnings are, of course, meaningless until set alongside trends in prices. The assertion by Smith and Francis that the wages bill in the south Wales coal industry fell from £65million in 1920 to £14million in 1933 is rather disingenuous.7 Generally, prices fell gradually during the interwar period.8 A sharp fall in prices in 1921 meant that while wage rates fell by 23% retail prices fell p.162. 5 S. Glynn, and J. Oxborrow, Interwar Britain: A Social and Economic History (London, 1976), p.13. 6 G. Routh, Occupation and Pay in Great Britain 1906-79 (London, 1980), pp.139-153; Aldcroft, The Interwar Economy, pp.352-354; A. L. Chapman and R. Knight, Wages and Salaries in the United Kingdom 1920-1938 (Cambridge, 1953). 7 D. Smith and H. Francis, The Fed: a history of the South Wales miners in the Twentieth Century (London, 1980), p.33; similarly A. J. Chandler, ‘The Re-making of a Working Class: Migration from the South Wales Coalfield to the New Industry Areas of the Midlands c.1920-1940’, unpublished University of Wales Ph.D. thesis, (1989), p.51. 8 On prices see J. Burnett, A History of the Cost of Living (Harmondsworth, 1969), pp.309-312; Routh, Occupation and Pay in Great Britain, p.141; Aldcroft, The Interwar Economy, pp.362-365.

20 Chapter 1: Income by over 40%, while in the following year wages and prices fell equally by about 19%. Prices continued to fall gradually until about 1933-34 after which time they rose slowly. The fall in prices, when set against the movement in wages, caused a fairly marked rise in real earnings in Britain as a whole. Aggregate wage indices and statistics suffer from being abstract averages that conceal as much as they reveal. They conceal significant variations between occupations, industries and regions, and they cover a workforce that varied according to skill, age and sex. Furthermore, these wage indices refer to fully employed workers and do not take into account irregularities of employment such as short-time working, payments in kind, stoppages and even the wilful under-payment by calculating employers. Finally, they take no account of the nature of the family the wage was intended to support. Its composition in terms of the number and sex of its members, and the proportion of wage earners to dependants, had important implications for the quality of life enjoyed by a family. Therefore, from the broad outline of real wages already given it is important to inquire more fully into the nature of incomes in interwar south Wales.9 In order to examine the nature of incomes in south Wales various rates and indices of wages and earnings have been constructed using the changes in wage rates as set out in the Ministry of Labour Gazette. By these means, wage and earnings quotients have been calculated for miners, agricultural labourers, tradesmen and labourers in the building trades, labourers employed by various county and district councils, and various classes of workers in the ‘engineering, boilermaking and foundry’ trades.10 All the rates constructed for these workers are specific to south Wales and differ slightly from the earnings of the same classes of workers elsewhere in Britain. Unfortunately, the rates are not strictly comparable since they vary from hourly rates to average earnings per shift and to minimum weekly wages. Weekly income is difficult to calculate since no complete series of hours worked exist for the industries and it is impossible to accurately account for factors such as overtime, short-time working and deductions. Nevertheless, the rates serve to show the variations over time in each trade and can, perhaps, be used to outline the general hierarchy of wages among the industries.

9 Aldcroft, The Interwar Economy, pp.351-353. 10 It is also possible from this source to construct similar indices specific to south Wales for workers employed in Siemens steelworks, the tinplate industry, coke ovens, steel sheet rolling mills and the Cardiff-Dowlais and other ironworks, and bricklayers at iron and steel works.

What is immediately clear from Appendix 1.1 is that the pattern of wages followed the general outline for Britain as a whole – a sharp decline in the early 1920s, a period of stagnation until about 1933-34, followed by about five years of gradually rising rates in the mid- to late-1930s.11 Workers in the coal industry, working six shifts a week, could earn between about £2 13s. and £3 4s. a week during the late 1920s and 1930s. Agricultural labourers in Glamorgan could earn a minimum of between £1 12s. and £1 18s. a week while their counterparts in Monmouthshire could earn between £1 11s. and £1 15s. in the same period. Workers in the engineering, boilermaking and foundry trades in Cardiff and Barry earned typically anything between £2 8s. and £3 10s. per week depending on their level of skill.12 But as with the British rates and indices these are next to meaningless unless prices are taken into account. The Ministry of Labour calculated price indices for Britain as a whole, while historians and economists have since refined them using their own calculations.13 Again, these suffer from being averages that take no account of regional differences. Kate Nicholas, in her study of unemployment on Teesside in the interwar years, was able to construct a cost of living index specific to that area by utilising oral history sources, advertisements for shops and grocers in the local newspapers and the records of a local grocer.14 However, this has not proven possible for interwar south Wales and the index for Britain as a whole must be used.15 By factoring the British cost of living index into the wage rates for south Wales the standard of living of wage earners in south Wales can be ascertained.16 In all cases the standard of living of those in employment was considerably higher in the 1930s, the ‘devil’s decade’, than it had been in the 1920s. Despite the decline in wage rates through the interwar period the even greater fall in prices meant that real incomes rose, albeit gradually. Rates and indices of wages often bear little relation to the reality of working conditions. Economic circumstances, industrial and political factors and the nature of the labour market all need to be considered if an accurate account of incomes is to be obtained. Firstly, the level of skill needed to perform a certain job and the amount of

11 This is more easily seen when the rates are transformed into indices; see Appendix 1.2. 12 The flaws inherent in such averages and minimum rates are examined below. 13 Ministry of Labour Gazette, 1920-1939; B. R. Mitchell and P. Deane, Abstract of British Historical Statistics (Cambridge, 1962), p.478; Burnett, A History of the Cost of Living, pp.307-308. 14 K. Nicholas, The Social Effects of Unemployment in Teesside (Manchester, 1986), p.57. 15 See Appendix 1.3. 16 See Appendix 1.4.

22 Chapter 1: Income work carried out determined the wages a worker received at the end of each week. Even within the same industry workers varied greatly in the skills they possessed and the wage they received. Phillip Massey found that in Blaina in the 1930s only one colliery worker in six was a collier. Colliers earned a minimum of 9s. 5d. a shift and were able to earn anything up to 18s. a shift, while repairers earned 9s. 9d. per shift and helpers and labourers, who made up the majority of the workforce, were only getting 8s. 5d. per shift. Colliers were therefore the only workers who could earn more than £2 18s. 6d. a week and those that did, claimed Massey, were perhaps one in ten of all colliery workers. The weekly earnings of all workers, no matter how high or low, were subject to deductions averaging three or four shillings a week.17 The wages of tinplate workers were similarly based on piece-rates and were also determined by the nature of the task of each worker. A new wage agreement in 1920 meant that workers were paid according to the number of boxes of tinplate they worked. Rollermen were paid 4s. 3d. per dozen boxes, doublers 3s. 5d., furnacemen 3s. 2d., behinders 1s. 9d., first helpers 3s. 0d., second helpers 2s. 1d. and shearers 1s. 4d. In addition, a sliding scale bonus was paid that was determined by the price of tinplate bar but because the price of tinplate fell so low workers received no additional payments for many years, although they did receive an ex gratia bonus, paid by the employers for the sake of industrial relations.18 Secondly, the economic conditions of the interwar years and the high levels of unemployment meant that the employers held an advantage in industrial relations. Political activists were victimised and many employers took advantage of their strong position to pay their workers as little as possible.19 In the coal industry young men of 16 or 17 years of age were preferred to older men since they could be paid less.20 Another tactic of employers was not to pay wages for the ‘small coal’ the colliers cut despite selling this for a profit.21 A respondent to James Handley described the nature of the relationship between employers and workers in the coal industry in the interwar period. He told Hanley of a certain mine where the men were being paid short in their

17 P. Massey, Portrait of a Mining Town, Fact, 8, (November, 1937), pp.29-30; these observations help to contextualise the average earnings of coal industry workers examined above. 18 W. Minchinton, The British Tinplate Industry. A History (Oxford, 1957), pp.220-222. 19 Cliff Prothero, Recount (Ormskirk, 1982), p.23. 20 Men Without Work. A Report made for the Pilgrim Trust (Cambridge, 1938), p.69 records an instance of a collier from the Rhondda stopped at the age of 21 because he had become eligible for a man’s wage. 21 Bert Coombes recounted how colliers did not get paid for the three smallest sizes of coal they cut (there were seven sizes in all); B. L. Coombes, Those Clouded Hills (London, 1944), p.10.

23 Chapter 1: Income wages. One miner who mentioned this at the mine office received notice to finish the following week. Pugh explained: Now, of course, the men say nothing about it. The company are continually paying men short, and the Federation can’t do anything at all, though they want to help the men. But some of the men are so afraid they’ll be sacked that they say nothing at all. It goes on week after week. It can’t be otherwise. It’s either that or the workhouse for him.22

More widespread was the practice of short-time working. In some industries this was institutionalised as employers and unions sought to resolve the problems caused by insufficient demand. In the tinplate industry, for example, there existed a surplus of labour throughout the interwar period, which, for various reasons, neither the employers nor the workers did anything to remedy.23 Tinplate companies preferred to share the orders they received amongst their mills than to concentrate production in a smaller number of mills in readiness for periods of increased demand, and so workers commonly faced periods of short-time working or temporary unemployment. In this instance it is evident that not only was short-time working determined by the variation in demand for a product but that industrial factors were also exerting an influence.24 The iron, steel and coal industries were similarly characterised by short-time working as efforts were made to maintain an available labour supply for periods of increased demand. Tinplate companies even exploited the system of unemployment benefit by employing workers for a few days a week thereby allowing them to claim benefit as well as receive wages for the days on which they were employed.25 A Political and Economic Planning (PEP) report criticised the tinplate industry for the way in which it was ‘wastefully and anti-socially organised’ and for ‘drawing continuously a hidden subsidy at the expense of other industries through unemployment insurance.’ It contrasted the industry with the nickel works at Clydach, which it described as a model of ‘socially responsible labour policy.’26

22 J. Hanley, Grey Children. A Study in Humbug and Misery (London, 1937), p.2; see also G. H. Armbruster, ‘The Social Determination of Ideologies: Being a study of a Welsh Mining Community’, unpublished University of London Ph.D. thesis, (1940), pp.273-274. 23 Minchinton, The British Tinplate Industry, pp.210-214. 24 N. Whiteside and J. A. Gillespie, ‘Deconstructing unemployment: developments in Britain in the interwar years’, Economic History Review, XLIV, 4, (1991), pp.665-682. 25 Minchinton, The British Tinplate Industry, p.213. 26 Political and Economic Planning (PEP), The Problem of South Wales, Planning, 94, 9, (March 1937), p.6.

Dock workers in south Wales were similarly able to exploit the unemployment benefit system.27 After 1921 insurance benefit contributors were able to claim benefit if they were employed on no more than three days in a six-day period. Before efforts were made in the 1930s to make dock work a closed trade, labour was very casualised and employees were able to exert some control over their engagements and avoid work that endangered their benefits. ‘Three days on the hook and three days in the book’ became a common working week. In a survey of 1,000 Newport dockworkers in 1930 it was found that the average weekly wage of £2 18s. was supplemented by 8s. 7d. in unemployment benefit.28 A slightly different system prevailed in Cardiff, as Fenner Brockway found, among the 1,500 members of the Coal Trimmers’ Union. The employers at the docks paid the men’s wages, not to the individual workmen, but to the union, which then distributed the total amount equally among the men. The men worked two weeks on and two weeks off, during which time they also received unemployment benefit. By this method each worker was assured 50s. a week whether he worked or not.29 Dockworkers were able to receive the more generous unemployment benefit rather than assistance or transitional payments because the relatively high wages they received for their work and the regular nature of their employment allowed them to make the necessary contributions. If employers were in a position to aid their workers in regard to unemployment benefit eligibility they were also able to frustrate their employees’ efforts to maintain a steady income. A number of miners in the coal industry during the interwar period remembered how underemployment was common and that colliery managers deliberately set out to deny their workers unemployment benefit. Kenneth Maher of Caerphilly believed that the coal-owners and the government were constantly ‘bashing’ the miners: The favourite trick was to work on Monday and Tuesday, off Wednesday, work Thursday, off Friday, work Saturday, or off Monday, work Tuesday, off Wednesday, work Thursday, off Friday, work Saturday. In this way the men

27 G. Phillips and N. Whiteside, Casual Labour: The Unemployment Question in the Port Transport Industry 1880-1970 (Oxford, 1985), pp.184-187, 209; S. Davies, ‘“Three on the hook and three on the book”: dock labourers and unemployment insurance between the wars’, Labour History Review, 59, 3, (1994), pp.34-43 points out that this was not as common as has sometimes been claimed and only served to aid the lowest paid, short-time workers. It did not, Davies claims, make a significant contribution to the incomes of the majority of dock workers. 28 Phillips and Whiteside, Casual Labour, p.187. 29 A. Fenner Brockway, Hungry England (London, 1932), pp.135-136. Whiteside and Gillespie contend that in some cases work-sharing schemes were established in industries where the unions could not afford elaborate systems of unemployment benefit; Whiteside and Gillespie, ‘Deconstructing unemployment’, p.667.

could not claim any dole. They were taking home maybe three days’ pay – about £1 or 25/-.30

Similarly, Bert Coombes, working in a colliery in Resolven in the Neath valley, experienced short-time working and believed that the colliery officials were in league with the labour exchange to deny the colliers their dole. Coombes found that he often had to work a shift for eight shillings instead of claiming 14s. 6d. as three day’s dole. The fate of men with large families was particularly desperate, he maintained, as they lost money every week by working.31 An examination of the record of the Lady Windsor colliery in Ynysybwl in the 1930s demonstrates the fluctuations in work that occurred in the coal industry: Days lost 1930 61 (exclusive of holidays) 1931 69 (including 8 full weeks) 1932 132 (including 16 full weeks) 1933 101 (including 12 full weeks) 1934 117 (including 12 full weeks) 1935 61 (including 10 full weeks) 1936 211 (districts working 31 shared weeks) Total 752 Table 1.1: Number of days lost annually at the Lady Windsor colliery, Ynysybwl, 1930-1936.32

J. E. Morgan, the official historian of the colliery lodge, recorded that during the working week in this period only certain districts of the colliery were working at any one time. Resentment at this grew to a pitch in 1935 after which date a policy was implemented whereby work was shared equally by the different districts of the colliery working in rotation.33 This is an important point. In June and December of each year the Ministry of Labour Gazette published a table showing the average number of days being worked every week by the pits in each coalfield. Throughout the interwar period, the average number of days being worked per week by the mines

30 N. Gray, The Worst of Times. An Oral History of the Great Depression in Britain (Aldershot, 1985), p.36. Grafton Radcliffe of Blaengarw argued that since four days of work each week was invariably the case there must have been some collusion between coal-owners and the employment exchanges; G. Radcliffe, Back to Blaengarw (Blaengarw, 1994), p.54. 31 B. L. Coombes, These Poor Hands (London, 1939), pp.215-216; Jack Wilcox of the Garw valley also remembered the colliery officials deliberately organising the working week so that the men could not claim benefit and stated that he and his work-mates would hide when they saw the colliery truck coming to fetch them for work – ‘Three days dole and three days work was better than working four days’ he commented; Chandler, ‘The re-making of a working class’, p.88. 32 J. E. Morgan, A Village Workers’ Council and what it accomplished, being A Short History of the Lady Windsor Lodge, South Wales Miners’ Federation (Pontypridd, 1956), p.31. 33 Ibid.

26 Chapter 1: Income in south Wales was consistently higher than the average for England and Wales or Great Britain as a whole. While the pits in many coalfields were only working 3 to 3.5 days per week in the early 1930s, those in south Wales were working over 5 days.34 In December 1932, for example, of the 15 coalfields for which statistics are available only four (Northumberland, Kent, North Wales, and Cumberland and Westmoreland) had a higher rate than south Wales.35 Therefore, the evidence from the Lady Windsor colliery suggests that while each working mine in south Wales was working an almost full week the districts in each mine were suffering regular spells of unemployment as the available work was shared between them. This is supported by the observation made in Ferndale in the Rhondda Fach in 1928 that while all the pits were working full-time, coal was not being raised from all the districts and seams, which meant that whereas 3,500 to 4,000 men had worked formerly now only 2,000 to 2,500 were regularly employed.36 This large volume of short-time working meant that large numbers of workers experienced short spells of unemployment at some point during the year. The consequences were predictable. Hilda Jennings found that the average weekly income of the 1,053 Brynmawr miners who had been employed at some time in the year ending March 1929 was only 33s. 8d., a figure, she noted, well below the legal minimum. Jennings concluded that the mining industry was not providing a living for those men still in work and that when stoppages and travelling costs were taken into account miners were in a relatively poor financial situation.37 Similarly, Arthur Lowry, the Ministry of Health investigator appointed to look into allegations of inadequate relief made against the Bedwellty commissioners in 1928, found that a large proportion of the miners in the district were only receiving the subsistence wage of 48s. 4½d. per week despite working full-time. After deductions and the payment of rent, the families of these men were forced to satisfy their food requirements and other needs from about 30s. per week. Workers who only worked three shifts per week and drew partial benefit, Lowry maintained, were even

34 As G. P. Griffin has shown, collieries in the coalfields of the east Midlands were generally working less than four days a week in the period 1928 to 1936 in what has often been described as a relatively prosperous area; C. P. Griffin, ‘“Three days down the pit and three days play”: Underemployment in the East Midland Coalfields between the Wars’, International Review of Social History, 38, (1993), pp.321-323. 35 Ministry of Labour Gazette, (January, 1932), p.16. 36 Labour Party Committee of Inquiry, The Distress in South Wales (London, 1928), p.13. 37 Jennings, Brynmawr, p.159. It must be noted, however, that the year ending March 1929 was a time of excessive short-time working.

27 Chapter 1: Income worse off and were forced to satisfy their dietary and other requirements with much less than 30s. per week.38 Therefore, the weekly wages of miners working short-time were less than the wages received by the dockworkers experiencing underemployment examined above who earned over £2 and even over £3 a week. This was partly due to the economic situation in each industry but also due to the attitudes of employers and the ability of workers to control their own working arrangements.

The difference between ‘temporary’ spells of unemployment and ‘long’ unemployment is an important distinction. Social commentators and propagandists, writing on south Wales and other distressed areas during the interwar period, focused on the long-term unemployment associated with the so-called ‘black-spots’, with the result that the varying conditions of unemployed people were overlooked. The diverse forms and experiences of unemployment need to be examined if the varied consequences are to be understood.39 Broadly, a peak in unemployment in 1921 was followed by a short period of relative prosperity in the early 1920s. But by the mid- to late 1920s between a quarter and a third of the insured population of Glamorgan and Monmouthshire were unemployed.40 Rising steadily through the early 1930s the figure reached a peak of 40.4% in Glamorgan and 42% in Monmouthshire in 1932 after which time it declined, albeit gradually, so that in the first eight months of 1939 about a fifth of the ‘working’ population of south Wales was still unemployed. However, the unemployment rate of 40% that Glamorgan experienced in 1932 is compatible with situations in which all workers were unemployed for 40% of the year, 40% of the labour force was unemployed for the whole year, or anything between these two extremes.41 For this reason aggregate statistics are of only limited value and do not allow an understanding of the consequences that unemployment had for the living standards experienced by the people of south Wales. The experience of unemployment was a diverse one. John Hilton, a civil servant of the Ministry of Labour, found on a tour of employment exchanges in the spring of 1929 that the

38 PRO, MH79/304, Correspondence from Arthur Lowry to Sir Arthur Robinson, 4 December 1928, pp.1-2. 39 Chandler, ‘The re-making of a working class’, pp.48-49. 40 See Appendix 1.5. On the reliability of unemployment statistics see W. Garside, The Measurement of Unemployment: Methods and Sources in Great Britain, 1850-1979 (Oxford, 1980). 41 M. Thomas, ‘Labour Market Structure and the Nature of Unemployment in Interwar Britain’, in B. Eichengreen and T. J. Hatton (eds.), Interwar Unemployment in International Perspective (Dordrecht, 1988), p.98.

28 Chapter 1: Income unemployed were characterised by considerable variation in circumstances, experiences and reactions to unemployment: Every other person is a special case. Even at Merthyr and Bishop Auckland the unemployed colliery workers, who may be thought of as a group if any body of men can, present the most diverse personal characteristics and circumstances . . . Each man is a category to himself.42

Therefore, unemployment varied in nature and extent throughout the interwar period and it is this variation that needs to be examined. Firstly, the incidence of unemployment varied greatly within south Wales.43 Those ‘heads of the valleys’ communities where pits had been worked out and where the closure of iron and steel works both increased the volume of unemployment and caused a diminution in the demand for local coal, suffered the highest levels of unemployment. In an arc of towns stretching from Merthyr in the west to Blaenavon in the east, and including Dowlais, Brynmawr, Nantyglo and Blaina the unemployed often constituted over half the ‘working’ population. The experience of the area served by the Blaina exchange that recorded an unemployment figure of 93% in 1932 is scarcely imaginable. Elsewhere, communities in the valleys of the central and eastern coalfield, while not quite suffering the same high levels of unemployment, similarly experienced considerable levels of unemployment. Communities stretching from Glyncorrwg to Pontypool regularly recorded percentages of between 25% and 50% throughout the 1930s and sometimes higher. To the south and west of these stricken communities relatively lower levels of unemployment were recorded as the more diverse nature of their economies and the relatively more buoyant market for anthracite provided more employment. In the county boroughs of Newport, Cardiff and Swansea levels of unemployment were generally lower than in the inland valleys but even here peaked at about a third of the insured population in the early 1930s. The geographical incidence of unemployment in south Wales has long been recognised by Welsh historians (even if the Local Unemployment Index has been under-utilised as a source)44 and is relatively unproblematic. Much less attention has 42 NLW, Thomas Jones C. H. papers, J. Hilton, ‘Reflections on a Tour of Certain Employment Exchanges’, 15 June 1929, p.2. This tour included a number of exchanges in England and the Cardiff, Merthyr, Dowlais and Aberdare exchanges in Wales and was conducted between 10 April and 12 May 1929. 43 Appendix 1.5 sets out unemployment figures for each labour exchange in the period 1927 to 1939. The story these figures convey is a familiar one; see Morgan, Rebirth of a Nation, pp.210-240; Davies, A History of Wales, pp.509-596. 44 For a notable exception see Andy Chandler, ‘“The Black Death on Wheels”; Unemployment and Migration – the Experience of Inter-War South Wales’, in Tim Williams (ed.), Papers in Modern

been given to the structure of unemployment in south Wales. Firstly, the duration of spells of unemployment needs to be ascertained. A great deal of official attention on the problem of unemployment was focused on ‘long unemployment’45 and it would seem that not only did south Wales experience higher levels of unemployment than other areas of Britain, but that the unemployed in south Wales were out of work for longer periods of time. Systematic measurement of this aspect of unemployment was not carried out until the late 1930s46 but even from this period the relatively worse conditions in Wales are easily apparent: Date District 1 2 3 4 5 6 7 8 9 21 June 1937 7.7 9.6 12.2 19.2 16.4 25.2 40.3 33.0 39.3 13 Sept 1937 7.9 8.6 11.6 18.0 19.8 22.4 38.2 30.6 37.9 13 Dec 1937 6.3 5.5 8.3 13.7 13.4 19.2 30.1 24.1 31.8 14 March 1938 5.6 5.9 8.3 12.3 14.1 17.9 32.3 25.9 27.4 13 June 1938 6.3 7.4 8.9 10.8 13.3 16.0 29.2 26.8 25.1 12 Sept 1938 6.8 7.3 9.3 11.9 13.3 17.3 28.6 25.8 26.5 12 Dec 1938 5.9 5.1 6.5 13.1 13.8 19.6 26.1 24.4 24.7 13 Mar 1939 6.5 6.0 7.1 15.1 15.8 21.0 27.4 25.3 27.0 12 June 1939 8.1 9.1 10.1 16.9 17.9 23.7 33.6 30.5 33.5 14 Aug 1939 8.1 9.6 10.5 14.2 19.8 25.3 35.4 31.0 32.9 Table 1.2: Percentage of applicants for benefit (aged 16-64) in regions of Britain unemployed for a year or more, 1937-1939. (1-London; 2-South Eastern; 3-South Western; 4-Midlands; 5-North Eastern; 6-North Western; 7-Northern; 8-Scotland; 9- Wales).47

In absolute terms, the number of ‘long unemployed’ in the Wales division varied from almost 49,000 in June 1937 to 30,000 in August 1939 and was therefore significant in numerical terms also. The duration of spells of unemployment in Wales can be broken down further: Date Applicants for benefit, aged 16-64, on register (Wales Division) Less 3 to 6 6 to 9 9 to 12 More Total than 3 months months months than 12

Welsh History, 1, (Cardiff, 1982), pp.1-15; Chandler, ‘The Re-making of a Working Class’. 45 In official definitions ‘long unemployment’ referred to a period of unemployment lasting a year or more. 46 Although Crafts found that 21.1% of applicants for benefits in Wales in June 1932 were unemployed for a year or more; Crafts, ‘Long-term unemployment in Britain in the 1930s’, p.422. 47 Ministry of Labour Gazette, (April, 1939), p.123; (July, 1939), p.255; (September, 1939), p.349. These figures apply to Wales as a whole but since the majority of the population of Wales lived in Monmouthshire and Glamorgan they serve as a good indicator of patterns there. See also Men Without Work, A Report made to the Pilgrim Trust (Cambridge, 1938), p.16 for comparable figures. These show that 37% of the unemployed in Wales in the summer of 1936 were ‘long unemployed’; also N. F. R. Crafts, ‘Long-term unemployment in Britain in the 1930s’, Economic History Review, 40, 3, (1987), pp.420-422.

months months 21 June 1937 44,755 14,034 9,350 7,272 48,781 124,192 13 Sept 1937 49,302 10,846 7,630 5,981 44,978 118,737 13 Dec 1937 62,894 15,609 7,383 5,670 42,716 134,272 14 March 1938 73,618 18,387 10,357 5,410 40,765 148,537 13 June 1938 82,750 17,390 10,336 7,372 39,472 157,320 12 Sept 1938 77,794 15,019 10,747 6,895 39,798 150,253 12 Dec 1938 84,873 18,157 9,447 7,020 39,114 158,611 13 Mar 1939 65,098 21,267 11,496 6,776 38,765 143,402 12 June 1939 43,348 11,149 9,057 6,005 35,031 104,590 14 Aug 1939 43,489 8,760 5,180 5,214 30,675 93,318 Table 1.3: Duration of unemployment in Wales, June 1937 to August 1939.48

The persistence of high levels of unemployment for spells of less than three months duration at these various dates demonstrates how, in the late 1930s at least, most people experienced unemployment for only short periods. Detailed figures on the duration of unemployment such as those shown in Table 1.3 are only available for the mid- to late 1930s and any understanding of this issue for an earlier period is based upon less complete information. Various sources suggest that the number of ‘temporarily stopped’ workers was highest in the aftermath of the coal lockout of 1926 but declined with the small improvement in the coal trade in 1929.49 This is supported by figures in a Treasury memorandum on the Coalfield Distress Fund.

Wholly unemployed Temporarily stopped November 1928 71,572 26,818 March 1929 53,285 13,171 September 1929 49,930 15,298 October 1929 52,197 14,698 November 1929 53,472 18,799 Table 1.4: ‘Wholly unemployed’ and ‘temporarily stopped’ unemployed in ‘South Wales’, 1928-1929.50

48 Ministry of Labour Gazette, (April, 1939), p.123; (July, 1939), p.255; (September, 1939), p.349. 49 Chandler, ‘The re-making of a working class’, pp.24-25. Chandler’s assertion is based on PRO, LAB 2/1293, comments of the Divisional Controller. See also The Times, 29 March 1928, pp.17-18 which shows that south Wales, although suffering comparable levels of ‘permanently unemployed’ as the Durham coalfield, was also experiencing higher levels of ‘temporarily stopped’. 50 NLW, Thomas Jones C.H. papers, ‘The Coalfield Distress Fund’, Treasury Memorandum for the Cabinet, (November, 1929), p.8.

Mark Thomas, in his analysis of unemployment in Britain as a whole, has found that the turnover of the unemployment register was very rapid during the 1920s and that the depression of the early 1930s was marked by a decrease in the rate of absorption of the unemployed by the labour market and a resultant increase in the expected duration of unemployment. After 1932 the labour market gradually returned to patterns of more rapid turnover so that the average duration of unemployment followed an almost identical pattern to the unemployment rate across the interwar period: in the trough of the depression in the early 1930s unemployment was not only of greater magnitude but was also of longer duration.51 Although it is difficult to arrive at any strong conclusions about south Wales it seems that the peak in unemployment in the early 1930s was characterised by an increase in short-term unemployment: Numbers of men applying for benefit who had been continuously on the registers less than 3 months: 27 June 1932 – 114,225 19 December 1932 – 98,024 22 June 1936 – 69,684 14 December 1936 – 57,065 Numbers of men applying for benefit who had been continuously unemployed on the registers for 12 months or more: 26 June 1933 – 62,399 18 December 1933 – 64,822 22 June 1936 – 63,744 14 December 1936 – 57,015.52

The increase in unemployment in south Wales in the early 1930s was due to an increase in the numbers becoming unemployed and a decrease in the labour market’s ability to reabsorb unemployed individuals, rather than a decline in the labour market rate of absorption alone as Thomas found in regard to Britain as a whole. Unemployment in south Wales during the early 1930s was primarily characterised by short, rather than long, spells of unemployment. As the figures suggest, while the numbers of short-term unemployed declined greatly from their peak in 193253 the numbers of ‘long unemployed’ fell only gradually and, as seen in

51 Thomas, ‘Labour Market Structure and the Nature of Unemployment in Interwar Britain’, p.105. 52 Ministry of Labour Gazette, (January 1937), p.8. These figures again refer to the ‘Wales Division’. The comparison of short-term unemployment to the situation in 1932 and long-term unemployment to that of 1933 was because these were the periods in which each was at its highest. 53 The evidence in Appendix 1.6 supports this contention that the peak in unemployment in the early 1930s was occasioned by an increase in the short-term unemployed. This interpretation is also supported by anecdotal evidence provided in PEP, The Problem of South Wales, p.3.

Table 1.3, still amounted to about 40,000 by the late 1930s. Although it is possible to obtain statistics for the duration of unemployment in Wales, figures for the number of spells of unemployment experienced by a worker during a twelve month period, or indeed any period, do not exist and so an unknown proportion of those in work or registered as short-term unemployed at any point in time might have experienced numerous spells of unemployment but still be registered as in employment or as short- term unemployed. Numerous spells of unemployment brought with them their own distinct consequences for standards of living. A further insight that can be gained from the evidence presented above is that as the duration of unemployment increased so the likelihood of gaining employment decreased. This seems rather obvious since the short-term unemployed were unemployed for short periods precisely because they re-entered the labour market rapidly after becoming unemployed, while the long-term unemployed were without work for so long precisely because they could not obtain work. This seems a tautology but is not. At any point in time a short-term unemployed person was more likely to gain employment than somebody experiencing long-term unemployment. The Rhondda collier in Beales and Lambert’s Memoirs of the Unemployed, who found himself refused a job because he had been without work for too long, was probably typical of a large number of individuals who were victims of the belief that without work a man became ‘soft’ and ‘out of shape’.54 This is supported by evidence on the incidence of unemployment according to age. Anecdotal evidence suggested that it was older men who suffered a disproportionate amount of unemployment and for longer periods of time.55 A survey carried out in Merthyr in 1929 estimated that 60% of the unemployed were over 40 years of age.56 This report also noted that many of these men were physically unfit

54 H. L. Beales and R. S. Lambert (eds.), Memoirs of the Unemployed (Yorkshire, 1973 edition), p.65; Massey, Portrait of a Mining Town, p.70, commented that there was much contradictory evidence as to whether this belief was true or not but that it was a belief that was ‘unofficially recognised’ by colliery officials when they came to take on new workers. 55 For statistics on the duration of unemployment according to age see Ministry of Labour Gazette, (June, 1938), p.233; (July, 1939), p.263; For quarterly statements on the age incidence of unemployment see Gazette 1935-1939. 56 NLW, Thomas Jones C.H. papers, John Davies and David E. Evans, ‘Report on Merthyr Tydfil’, 11 July 1929, p.5; see also NLW, T. Alban Davies papers, ‘The Memorandum on the Human, Cultural and Spiritual Reactions of the Depression in the Special Area of South Wales’, submitted to the Prime Minister by the United Committee of the Churches in Wales, 22 July 1937, p.3, which comments on the impoverishment of the 45-65 age group who suffered a greater proportion of unemployment than other age groups; this incidence of unemployment according to age was probably also influenced by the greater likelihood of young men to emigrate from south Wales than older men.

33 Chapter 1: Income and were unable to do the work found for them. It is impossible to know whether ill- health was a cause or a symptom of unemployment in such cases, whether it was caused by unemployment or merely revealed by it.57 Mark Thomas found that unemployment rates for older workers were high because older men were as likely to become unemployed as younger men but were less likely to subsequently obtain another job.58 Long-term unemployment was most often caused by the closure of pits or iron and steel works whereas temporary unemployment was caused by places of employment closing for short periods due to insufficient demand. Francis and Smith maintain that 241 mines closed in south Wales between 1921 and 1936 and demonstrate how closures were concentrated in certain periods. Between January 1927 and April 1928 56 collieries employing 23,370 men closed, while in the 18 months following January 1929 a further 138 pits employing 18,300 men closed.59 Pits in the north of the coalfield, producing coal for the local iron industry, closed as the ironworks became redundant or relocated nearer the coast, or as their coal seams became exhausted. In communities such as Blaina, Brynmawr, Merthyr, Dowlais and Ebbw Vale the abandonment of a mine or steel works meant the loss of virtually the sole place of employment in a community. The report made by Wyndham Portal highlighted the long unemployment found in the ‘Eastern Section’ of the coalfield and demonstrated that of the 80,661 wholly unemployed workers registered at employment exchanges in May 1934, 74% were unemployed for over a year.60 In the iron and steel industry this figure stood at 90% due to the large-scale closure of plants such as those at Blaenavon, Ebbw Vale and Dowlais. The closure of iron and steel works at Ebbw Vale in 1929 and Dowlais in 1930 meant that 2,000 men had had no regular work in the intervening three or four years.61 Thus in Ebbw Vale, 87% of the unemployed had not worked for over a year by May 1934 and 83% of the unemployed at Merthyr were similarly disadvantaged.62

57 ‘Physically unfit’ men were able to find work as labourers or unskilled workers in times of prosperity but were the first to lose their jobs in slack times. 58 Thomas, ‘Labour Market Structure and the Nature of Unemployment in Interwar Britain’, pp.117- 119; Thomas also noted that the higher mobility of younger men increased their chances of finding another job. 59 Francis and Smith, The Fed, pp.33, 97, 176. 60 Ministry of Labour, Reports of the Investigation into the Industrial Conditions in certain Depressed Areas, III: South Wales and Monmouthshire, [Cmd. 4728], 1933-34, xiii, p.135. 61 Ibid. 62 Ibid., p.136.

In the Rhondda, on the other hand, collieries did not close but were often temporarily stopped,63 while in the Garw valley, which produced coal for export, the level of employment and unemployment fluctuated with variations in international demand and the size of stockpiles at the ports. Furthermore, the collieries of the Garw valley were all owned by the same company, which meant that they were all working or shut as demand fluctuated. Also, the colliers of the Garw valley all lived in the immediate vicinity and did not travel outside the area to work, as did colliers in other parts of the coalfield. For these reasons fluctuations in the demand for coal registered itself more directly on the unemployment figures for that area.64 Clearly, the local circumstances of the coal industry determined the nature of the labour market in an area and the experiences of employment and unemployment in the interwar period. A further source that sheds light on the variations in unemployment is the Local Unemployment Index published by the Ministry of Labour. This index, published from 1927 until 1939, recorded monthly unemployment percentages for each employment exchange in Britain. This index has already been used to demonstrate the varying levels unemployment throughout south Wales65 but can also be used to indicate the nature of unemployment in each area. By calculating the unemployment range during the course of a year it is possible to determine whether an area contained a great deal of long unemployment or, alternatively, temporary unemployment. For example, 15.9% of the insured workers registered at the employment exchange in Ogmore Vale were unemployed in February 1930. By March this proportion had risen to 43.7% but had then fallen to 16.1% the following month. From May to August figures of 42.5%, 20.1%, 46.4% and 18.1% respectively were recorded.66 Quite clearly, the unemployment average for Ogmore Vale in 1930 concealed a great deal of short-term unemployment. Conversely, the unemployment figure of the Merthyr exchange did not vary a great deal during the course of a year despite the high level at which it remained throughout the period. In 1932, for example, unemployment varied between 51% and 66% of the insured population. The low range in percentage undoubtedly indicated

63 Ibid., p.130; see also Labour Party Committee of Inquiry, The Distress in South Wales, p.12 which states ‘We were told that the mining situation in the [Rhondda] . . . could be summed up in the words, intermittent employment. With one or two exceptions no pits were fully working, and several were completely closed.’ 64 Chandler, ‘The Re-making of a Working Class’, pp.81-84. 65 See Appendix 1.5. 66 Local Unemployment Index, 1930.

35 Chapter 1: Income the large numbers of long-term unemployed and the absence of short spells of unemployment. Appendix 1.6 sets out the range in unemployment percentages for each local exchange in south Wales. Exchanges such as Ogmore Vale, Ferndale, Pontycymmer and Maesteg display high ranges, indicating a great deal of temporary unemployment, whereas exchanges such as Merthyr, Blaenavon, Ebbw Vale and Tredegar display low ranges despite high levels of unemployment. Figure 1.1 represents these ranges and plots the average annual range in unemployment percentage points for the period 1927 to 1939.

Figure 1.1: Average annual range in unemployment percentage, 1927-1939.67

67 Ibid., 1927-1939.