2 Expansion and decline of iron mining in BiscayFootnote 5

In 1876, Biscay produced 400,000 tons of phosphorus-free iron ore. The average production during the 1880s was 3.5 million tons. The figure of 5 million tons was reached at the end of the century and 4 million tons were extracted in 1913. However, production dropped to an average 2.8 million tons during the World War I and fell even more between 1919 and 1936, with an average 1.7 million tons of iron ore being mined.

In all, 75 per cent of the iron ore was shipped to Great Britain between 1876 and 1913. Therefore, the sector’s expansion cycle must be explained by the huge British demand. 78 per cent of the steel manufactured in Great Britain during those years was Bessemer or acid Martin–Siemens steel. This preference for acid systems made British steel technologically dependent on phosphorus-free iron ores. Pure ores like those from Biscay were then exploited in Cumberland–Lancashire, in some Swedish ore bodies, in Spain (Santander and the Penibetic Range) and in Greece. Nevertheless, the production of these mining areas was insufficient to cater for the whole British demand, as a result of which and due to the shortage in the supply of alternative minerals, the demand for Biscayan ore became inelastic between 1876 and 1913. This is the key aspect that explains the expansion within this sector. The increase in Britain’s production of acid steels pulled with it both the demand for Biscayan ore and its prices. Profits grew and so did ore production and the amounts of factors used — in the case of the labour factor, the number of workers rose from 950 in 1876 to over 8,000 during the 1890s and then to 12,000 in 1913.

The decline of the Biscayan mining industry stemmed from the technological change introduced by the British iron and steel industry during the World War I and was aggravated during the post-War period as a result of the modifications experienced in the demand for steel inputs. Between 1914 and 1918, Great Britain had to replace phosphorus-free iron ore imports due to the rise in freight prices. To this end, the iron and steel industry enterprises received help from the government to build basic Martin–Siemens furnaces, which extracted domestic phosphorous minerals as well as phosphorous scrap metal. Once the factories had been equipped with the new technology, 70 per cent of the steel they produced between 1919 and 1936 was basic Martin–Siemens steel because its costs were lower than those associated with acid steel. The reason for this difference lies in the prices of their respective inputs. At that time, basic Martin–Siemens furnaces consumed 70 per cent of phosphorous scrap metal — cheaper than pig iron after the war, as a result of its oversupply — and 30 per cent of pig iron, which was cheaper than acid steel because it was obtained from phosphorous minerals extracted in relatively young mining areas (Lincolnshire, the Midlands and Nordic and French imports). Instead, acid Martin–Siemens furnaces consumed phosphorus-free scrap metal — the price of which was higher than that of the phosphorous one due to its relative shortage — and acid pig iron, which was more expensive than the basic one because it was obtained from pure minerals found in old mining areas subject to decreasing performances (Cumberland–Lancashire and Spanish deposits). In turn, Bessemer converters required 100 per cent of acid pig iron, which cost more than basic pig iron and scrap metal. The construction of Martin–Siemens furnaces was consequently the key factor triggering the decline of the Biscayan mining industry, «protected» until 1913 by British loyalty to acid systems. These furnaces provided the British demand for steel inputs with elasticity and made acid steel production plummet, which had a very strong impact on the demand and prices of ore from Biscay. Profits dropped and so did ore production and the amounts of factors used — in the case of the labour factor, the number of workers fell from 12,000 in 1913 to an average 6,000 between 1914 and 1936.

3 Three indicators of the living standard of miners in Biscay (1876-1936)

Building a rigorous series of miners’ real wages requires the availability of nominal wages obtained from reliable sources and a cost of living index that takes into account the following aspects: (i) retail prices and those in the mining area corresponding to the goods and services consumed by miners; (ii) the relative importance that each one of those goods and services had in their consumption; and (iii) the variations in the demand derived from the Engel lawFootnote 6.

Table A1 and Figure 1 show the annual series of daily wages for workers in Biscayan mines (this unskilled labour represented 80 per cent of workers in this sector)Footnote 7. Although these wages did not always come from payroll books, the sources used are reliable. Those corresponding to the 1876-1899 period were found in two reports about mines in Bilbao written by engineers working for the Financial Studies of Crédit LyonnaisFootnote 8. Two tables prepared by them show the highest and lowest wages paid each year to the workers of the two largest companies in the sector, Orconera Iron Ore and Franco Belga des Mines de Somorrostro, according to the data provided by the respective offices. The nominal wages corresponding to the 1876-1899 period are the arithmetic mean of the average daily wages paid by these two firms. Those corresponding to the years between 1900 and 1921 are the arithmetic mean of the daily wages obtained from another three sources: the Council of the mining town of San Salvador del Valle, two employers’ organisations (the Asociación de Patronos Mineros and the Círculo Minero) and the payroll books of the mines exploited by the iron and steel firm Altos Hornos de Vizcaya Footnote 9. Finally, the wages for the 1922-1936 period are the arithmetic mean of those found in the payroll books of the Ollargan mine and in the Memorias de la Cámara Minera de Vizcaya Footnote 10.

Figure 1 NOMINAL WAGES OF MINE WORKERS IN BISCAY (1876-1936)

Source: See text and Table A1.

The cost of living index used to transform nominal wages into real ones appears in Table A1 as well as in Figure 2Footnote 11. The index corresponds to the Ría de Bilbao (Bilbao estuary) area, where the mines and the iron and steel industry, the metal–mechanical industry, the chemical industry and the shipyards were located. The prices are retail ones and they come from municipal sources, hospitals and dwelling rent registersFootnote 12. The goods and services consumed by the Biscayan working class and their weighting appear in Table 1, which was constructed using more than twenty family budgets, information from official institutions and documents obtained from the Hospital de Basurto and the Santa Casa de Misericordia hospice in BilbaoFootnote 13. Once the prices and the composition of the expenditure were known, the cost of living index was obtained using a Laspeyres-type formula (Pérez Castroviejo Reference Pérez Castroviejo2006).

Figure 2 COST OF LIVING IN THE RIA DE BILBAO INDUSTRIAL AREA (1876-1936)

Sources: See Table A1. Pérez Castroviejo (Reference Pérez Castroviejo2006).

Table 1 COMPOSITION OF BISCAYAN WORKING-CLASS EXPENDITURE (1876-1936)

Source: Pérez Castroviejo (Reference Pérez Castroviejo2006).

The evolution of the cost of living index for the Ría de Bilbao (Figure 2) resembles the situation for Spain and other regions: a modest price growth between 1876 and 1913 — despite which prices in the Ría de Bilbao grew more than in Spain as a whole during the 1880s — an inflationist «gap» during the World War I and the post-war period and finally, a drop, though remaining at a much higher level than in 1913Footnote 14. In turn, the composition of the working-class family expenditure shown in Table 1 is similar to that identified in Spain and other regions by some colleagues and does not differ much from the situation in Great Britain during the Industrial RevolutionFootnote 15.

Nominal wages and the cost of living index have helped us to estimate the first indicator of miners’ living standard: their real wages (Table A1 and Figure 3).

Figure 3 REAL WAGES OF MINE WORKERS IN BISCAY (1876-1936)

Source: See Table A1.

The second indicator — their HDI — appears in Table 2.

Table 2 HDI OF BISCAYAN MINERS (1877-1930)

Notes: HDI: Human Development Index.

*We have transformed the constant pesetas of 1876 into constant pesetas of 1990 with the series of Prados de la Escosura (Reference Prados de la Escosura2003) and later applied the exchange rate with purchasing power parity in 1990 for the peseta and the dollar estimated by Maddison (Reference Maddison1995).

Source: See text.

We will first mention the sources of the figures, after which we will describe how the HDI was calculated. In the absence of data for life expectancy in the eminently mining towns, we decided to use a «proxy»: the life expectancy at birth in Sestao and Baracaldo calculated by María Eugenia González Ugarte and Mercedes ArbaizaFootnote 16. Since these two towns were the most densely populated ones in the Bilbao industrial area, where the mining villages were also located, taking additionally into account that there were also mining quarters in Baracaldo, it seems reasonable to extrapolate that life expectancy to the case examined in the present paper. Literacy rates come from the Censos de la Población Española and are the weighted average of the seven eminently mining towns (Abanto and Ciérvana, Galdames, Güeñes, Musques, San Salvador del Valle and Sopuerta) and of another three with mining quarters (Baracaldo, Basauri and Santurce)Footnote 17.

Estimating the income per person component is complicated because there is no information available for family income that can be divided by a figure that we do know, that corresponding to the average family size (4.7 individuals)Footnote 18. However, the Male Breadwinner Family model was the predominant one because there was very little child and female work in mines — only 6 per cent of the workers were boys between 14 and 16 years of age and only 2 per cent womenFootnote 19. It thus seems reasonable to use the annual income obtained by the head of the family divided by 4.7 as a «proxy»Footnote 20. The income component estimated in this way is biased downward, since an undetermined number of families obtained more money because the women exercised wardship, but, as we have explained, it seems sensible to use the figures appearing in column 3 of Table 2Footnote 21.

The HDI was estimated according to the rules recommended by United Nations Development Programme (UNDP) for the calculation of the HDI but introducing two modifications: (i) no compound indicator of the education variable was used; the literacy rate was introduced insteadFootnote 22; (ii) we operated with minimum and maximum life expectancy values of 19.9 and 85 years, instead of 25 and 85, because the life expectancy in the mining area in 1890 was 19.9 years.

Military service was compulsory in Spain until very recently. All conscripts were measured at the town halls, including those who later managed to avoid military service by paying their «redemption» to the StateFootnote 23. In many cases, the local authorities have kept the so-called Actas de Clasificación y Alistamiento, in which the name of the conscript as well as his birthplace, profession and height are collected. Nevertheless, this rich documentation presents a problem: the changes in the recruitment age. While youngsters between 19 and 20 years of age were conscripted between 1856 and 1907, after 1907, young men were conscripted at the age of 21 yearsFootnote 24. This makes it necessary to standardise the height at the age of 21 years, because boys still grow between the ages of 19 and 21 years. A number of Spanish anthropometrists have standardised series adding to the average height of 19- and 20-year-olds, the average growth experienced by short youngsters, a growth that we know because these youngsters had to be measured for 3 years and their height at the age of 20 or 21 years was noted in the so-called Actas de Revisión (Martínez Carrión and Moreno Lázaro Reference Martínez Carrión and Moreno Lázaro2007; Cámara Reference Cámara2009; García Montero Reference García Montero2009; Martínez Carrión and Puche-Gil Reference Martínez Carrión and Puche-Gil2009; Ramon-Muñoz Reference Ramon-Muñoz2009; Hernández and Moreno Lázaro Reference Hernández and Moreno Lázaro2009).

For the purpose of determining the evolution of the average height of miners, we chose the height of the conscripts measured in the town of San Salvador del Valle as a «proxy» for two reasons:. first, because it was the largest mining town (6,500 inhabitants) and nearly 93 per cent of the conscripts were registered as jornaleros (mine workers); and second, because the documents corresponding to all the years between 1878 and 1935 have been preserved to the present day. The series was standardised using the recruitment certificate of 1885 because, due to special military needs, youngsters of different ages were measured that year. This information led us to increase by 0.55 centimetre the average heights of those measured at the age of 20 years and by 1 centimetre that of those measured at the age of 19 years, an increment similar to that estimated by the aforementioned anthropometrists. Table A2 and Figure 4 show the average height of the conscripts measured in San Salvador del Valle before and after standardisation at 21 years of age.

Figure 4 AVERAGE HEIGHT OF THE CONSCRIPTS MEASURED IN SAN SALVADOR DEL VALLE (1877-1935; 5-yearly averages in centimetres)

Source: See text and Table A2.

4 The living standard of Biscayan miners (1876-1913)

During the expansion cycle of the mining industry, workers’ nominal wages increased at an average cumulative rate of 0.75 per cent (Table A1 and Figure 1). This weak growth can be justified by the existence of a labour market that was close to the perfect competition model, mainly from the point of view of atomicity (many demanders and suppliers of labour, none of whom could alter the balance wage). This atomicity was due to the weak power of unions — none of the workers’ strikes for pay rises between 1876 and 1913 was successful — and to the non-existence of a monopolistic labour demand — mines were exploited by over 110 firmsFootnote 25. Under these circumstances, the productivity increase made both labour demand and wages grow, but the latter only increased modestly as a result of the simultaneous increase in labour supply caused by the mass immigration into the mining districts from nearby farming areas with a zero marginal productivityFootnote 26. It is worth mentioning that — within the general context of sluggishness — the wages of Biscayan miners grew more than those of other unskilled Spanish workers (Table 3).

Table 3 SPAIN. NOMINAL DAILY WAGES OF UNSKILLED WORKERS (1876-1913; PSESETAS)

Sources: (1): Table A1; (2): Martínez Soto (Reference Martínez Soto2005); (3): Arenas Posadas (Reference Arenas Posadas2000); (4): Bringas (Reference Bringas2000); (5): Lana Berasaín (Reference Lana Berasaín2005); (6): Garrabou and Tello (Reference Garrabou and Tello2002); (7): Moreno Lázaro (Reference Moreno Lázaro2006).

In Figure 3, it can be seen that the real wages of Biscayan miners decreased between 1883 and 1890 because the cost of living (Figure 2) increased by 20 per cent as a result of strong immigration flows and inelastic housing and food supplyFootnote 27. Real wages increased outside that period, but their growth was smaller than that of nominal wages: an average cumulative rate of 0.18 per cent between 1876 and 1913 and 0.35 per cent if one makes the 1876 wage per average 1909-1913 wage estimate to offset the disturbing effect of the last yearFootnote 28. This very slight increase is similar to that experienced by the real daily wage paid to other non-skilled Spanish workersFootnote 29. Therefore, taking salary as an indicator, the welfare of Biscayan miners improved modestly between 1876 and 1913. However, it deteriorated during the 1883-1890 period.

Following the line of research mentioned in the introduction to the paper, we will first compare the real wages of miners with their HDI and then with their height. The HDI has the disadvantage that it weights each one of its three components at 33 per cent — which implies a value judgement — and the advantage that it collects two «access rights», which do not necessarily depend on income (health and education)Footnote 30. Table 2 shows the HDI of miners. Let us look first at the evolution of each of its components until 1910, after which we will study the evolution of the indicator.

The evolution of the income component has already been analysed in the discussion of real wages: modest growth between 1877 and 1910 and a drop during the 1880s. The literacy rate rose from 40.9 to 57.4 per cent. This is a small growth, which, however, exceeded that of Spain as a whole (36.2-49.7 per cent) for reasons of supply and demand. The demand for education is related to family income, child labour and parents’ cultural level. In fact, the wages of Biscayan miners were among the highest paid to unskilled workers in Spain. There was little child labour in these mining areas, unlike the situation in farming and in the rest of the Spanish mining industry, where it reached 20 per centFootnote 31. Finally, and also contrary to the situation in other regions where illiteracy became a vicious circle, the higher literacy rate achieved in the Biscayan mining areas favoured an improvement in schooling levelsFootnote 32. In addition, the Biscayan municipalities created more school places than those located in other Spanish regions (Núñez Reference Núñez1992).

The evolution of life expectancy shows that, until the early 20^th century, mortality was higher than before industrialisation. Surprising as it might seem, the life expectancy figure for 1890 (19.9 years) is not a misprintFootnote 33. The population in the industrial and mining area trebled between 1876 and 1900; with one case, the mining town of San Salvador del Valle, the number of inhabitants are actually multiplied by sixFootnote 34. This human landslide caused by immigration had negative consequences (contaminated water and overcrowding), which, together with the sale of food in bad condition, considerably increased the risk of disease and, consequently, the mortality rate. As we are trying to explain below, the three problems derived from market faults, which would have required rapid state intervention, which unfortunately came too late.

Between 1876 and the early19^th century, the Town Councils in the mining areas permitted the construction of buildings without simultaneously building a water supply infrastructure and sewer system, a situation criticised by engineers and hygienist doctors from Bilbao, who warned that it would bring water contamination and an increased morbi-mortality as consequencesFootnote 35. Therefore, knowing these consequences — a market failure that we call «negative externality» nowadays — the public powers should have built the infrastructure of water supply and the sewer system or, failing that, should have encouraged construction by private firmsFootnote 36. In the case of food in bad condition, the market fault derived from an imperfect information flow. The hygienist doctors of the time explained how manipulation, poor hygienic conditions in warehouses and shops and adulteration resulted in the sale of food in bad condition that consumers could not recognise (see Arbaiza Reference Arbaiza1995; Pérez Castroviejo Reference Pérez Castroviejo2006). Therefore, the public authorities should have acted as regulators of economic activity. However, sanitary inspection in markets and bromatological control in laboratories did not become widespread until the end of the 19^th century or at the beginning of the 20^th centuryFootnote 37.

The problem of overcrowding can be attributed to a housing supply that acquired elasticity only in the long term. The literature regarding the construction economy shows that the housing supply was inelastic in the short termFootnote 38. If we add to this structural problem the fact that the demand for flats «soared» due to the rapid population growth, it can be understood why rent prices increased fourfold between 1876 and 1900. This forced working-class families to turn to board and lodging, thus generating serious overcrowding problemsFootnote 39. We are thus faced by two market failures: the slowness of the market under certain circumstances when it comes to offering a preferential good like housing; and the negative externalities that overcrowding causes on health. In these conditions — and following the hygienists’ recommendations — the public powers should have built cheap dwellings, which they did not do until the early 20^th century (Pérez Castroviejo Reference Pérez Castroviejo1994; Domingo Herández Reference Domingo Herández2008).

The aetiological studies carried out by González Ugarte (1994) and Arbaiza (1995) show that the three above-mentioned factors account for the extremely high mortality rate during the last third of the 19th century, as 84 per cent of deaths were caused by infectious diseases propagated through the air or via consumption of water or food in bad condition (González Ugarte Reference González Ugarte1994; Arbaiza Reference Arbaiza1995). In contrast, the gradual disappearance of these three problems in the early 20^th century justifies the reduction of mortality and the start of the demographic transition in the Biscayan industrial and mining area.

Following this analysis of the evolution of its components, let us now look at the evolution of the HDI between 1877 and 1910 (see Table 2):

1. (1) According to the UN conventional values, an HDI above 0.8 indicates a high welfare level; one situated between 0.5 and 0.8, an average welfare level and, finally, an HDI below 0.5 would indicate a low welfare level. In other words, Biscayan miners did not reach what is currently considered an intermediate living standard until the 1930s.

2. (2) The HDI declined from 0.316 to 0.232 between 1877 and 1890, a remarkable decrease mainly caused by the decrease in life expectancy and, secondarily, by the drop in income. Indeed, life expectancy evolved from a 0.277 index (36.5 years) to a 0 index (19.9 years), while the income component suffered a slight decrease.

3. (3) The HDI increased from 0.232 to 0.302 between 1890 and 1895 — a figure that was still below that of 1877 — because, although the income and literacy components in 1895 exceeded the 1877 indices, the figure for life expectancy in 1895 (0.155-29.2 years) was still well below that of 1877 (0.277-36.5 years).

4. (4) In 1900, the indicator reached the same figure as in 1877 (0.316), since the combined increase in income and literacy placed it at the initial level despite the fact that life expectancy (0.175-30.4 years) was still lower than in 1877 (0.277-36.5 years).

5. (5) Finally, the 1910 index (0.395) was above that of 1877 (0.316) because income and literacy continued to improve and life expectancy (0.320-39.1 years) was already higher than that of 1877 (0.277-36.5 years).

Thus, taking the HDI as an indicator, miners’ welfare deteriorated between 1876 and 1900, but improved at the beginning of the 20^th century. This conclusion clashes with the one for real wages, since using the latter, miners’ living standard became worse only between 1883 and 1890. We are faced with a dilemma that is as old as the debate between «pessimists» and «optimists»: the relative weighting of welfare components, income and life expectancy in this caseFootnote 40. During the 1980s, Williamson (1984) tried to solve the problem by estimating what he called the «real living standard», an adjustment between the increased life expectancy and the extra future consumption (Williamson Reference Williamson1984). The calculation turned out to be optimistic because it was based on the life expectancy of England as a whole provided by Wrigley and Schofield and on the real wages calculated by Williamson himself, which grew by 100 per cent during the Industrial Revolution. The outlook changed in the 1990s, as the data offered by Szreter and Mooney for life expectancy in English and Welsh industrial cities are pessimistic and the new series of real wages offered by Feinstein only rose by 30 per cent, so that, as this author argued, the increased income does not seem to have compensated for the decrease in life expectancy (Szreter and Mooney Reference Szreter and Mooney1998; Feinstein Reference Feinstein1998). This point of view, which seems reasonable in the British case, also makes sense in the case studied here, since the real wages paid to Biscayan miners between 1876 and the beginning of the 20th century grew by 12 per cent, while life expectancy in 1877 fell by 16 years during the 1880s and 6-7 years in the 1890s, the level of 1877 only being recovered during the first 5 years of the 20^th century.

It would have been desirable to construct HDIs for miners in other Spanish regions in order to compare them with that of Biscayan miners, but this was impossible because, in the case of the Alquifes or Río Tinto mines, we do not know the real wage, the average family size or the life expectancyFootnote 41 and, in the Cartagena mining industry, we know the life expectancy in El Beal, which reached a lower level than that of Biscayan mines (18.2 years in 1890 and 20.5 in 1895), but not the real wage or the average family size (Navarro Ortiz et al. Reference Navarro Ortiz, Martínez Soto and Pérez de Perceval2004). However, it is possible to compare the HDI of Biscayan miners with the figures Escudero and Simón have estimated for Spain, though this would require a reconstruction of the latter with minimum and maximum life expectancy values of 19.9 and 85 years — instead of 25 and 85 — for the purpose of homogenising it with that of the miners (Escudero & Simón Reference Escudero and Simón2003). After making this adjustment, the HDI of miners is situated below the Spanish index during the 1880s (0.232 as opposed to 0.310), the 1890s (0.302 as opposed to 0.314) and also in 1900 (0.316 as opposed to 0.399). The main reason lies in life expectancy although, as previously noted, the fact that miners’ HDI is biased downward on the income side surely has an influence too.

The third indicator of miners’ welfare, average height, has yet to be analysed. Some of the anthropometric series for the industrialisation period have been criticised because they come from biased sources, as military service was not compulsoryFootnote 42. However, after 30 years of anthropometric research, it hardly seems arguable that height is subject to cycles and that, although it contains a genetic load, that load is potential and can be altered by three factors, which, along with the genetic component, model it until the individual is 20-21 years old. These factors are: nutrition, morbidity and physical debilitation, as height results from the «net nutritional input», the difference between the «gross nutritional input» — the nutrients ingested — and the energy consumed via basal metabolism, physical activity and diseasesFootnote 43. Height thus becomes a synthetic indicator of what Komlos calls the «biological living standard», since nutrition is associated with real wages, morbidity and physical debilitation reflecting other welfare components (epidemiological environment, public health level and working conditions). On the other hand, the factors modelling height are especially sensitive during the early years of life and above all during the jump towards adolescence because, whereas other unfavourable conditions present during the childhood period can be compensated later, they become irreversible when they appear between the age of 13 and 17 years (see Eveleth Reference Eveleth1985; Tanner Reference Tanner1990; Komlos and Baten Reference Komlos and Baten2004).

The height series for the mining town of San Salvador del Valle (Figure 4) does not show biases because, as we have already said, all the conscripts were measured — including those who paid their «redemption» later — the series is standardised at 21 years of age and, in addition, nearly 93 per cent of those measured are registered as jornaleros (mine workers). There is, however, a problem. Between 1876 and 1911, half of the conscripts appear as born in San Salvador or in towns located in the mining and industrial area, while the other half are registered as born outside Biscay. The studies carried out by Pérez Castroviejo and Arbaiza show that most of the immigrants were families who found themselves in the most problematic period of their life cycle, that is, they had very young children (Pérez Castroviejo Reference Pérez Castroviejo1986; Arbaiza Reference Arbaiza1994). Therefore — and although it is impossible to know the exact proportion of conscripts born outside Biscay who spent all or part of their childhood and/or adolescence in the mining area — it seems logical to assume that most of them did, which is why we decided to use the average height of all the conscriptsFootnote 44.

As can be observed in Table A2 and in Figure 4 (standardised series at 21 years of age), the youngsters in the first cohort (measured between 1876 and 1880) had an average height of 163.65 centimetres and the following five cohorts (measured between 1881 and 1905) were shorter than the first one. The youngsters measured between 1876 and 1880 had been born between 1856 and 1860, and therefore their childhood and adolescence took place before the start of the mining «boom» in 1876. In contrast, the following five cohorts (measured between 1881 and 1905) spent their childhood and adolescence between 1876 and 1901, a period characterised by a diet poor in protein, high morbidity and worse labour conditions than those that existed since the beginning of the century. All of these aspects would explain the decrease in height with respect to the first cohortFootnote 45.

Table A2 and Figure 4 show that the seventh cohort (measured between 1907 and 1911) exceeded the height of the first one (164.07 centimetres as opposed to 163.65 centimetres). The youngsters spent nearly all their childhood during the 1890s and their adolescence between 1902 and 1906, a period characterised by an increased real wage, better nutrition, lower morbidity and better labour conditionsFootnote 46. Thus, using height as an indicator, miners’ welfare deteriorated between 1876 and the end of the 19^th century, but improved at the beginning of the 20^th century. This conclusion is in keeping with the one produced by the HDI, but differs from the conclusion drawn from the evolution of real wages (fall in the standard of living only between 1883 and 1890). Comparing all three indicators therefore proves to be more fruitful than using wages alone and reinforces the idea that wages can hide the behaviour of other welfare components.

We can only compare the evolution in the height of Biscayan miners with that of miners in Murcia since no anthropometric studies have been carried out in other mining regions. The height of miners in Cartagena and Mazarrón fell much more than that of Biscayan miners (3 vs. 1.62 centimetres) between the 1860s and the end of the 19^th century, the height of miners in Cartagena having gone down to 157-158 centimetres during the 1880s and 1890s (Martínez Carrión Reference Martínez Carrión2005). This greater deterioration in the «biological living standard» was probably a consequence of the extremely high percentage of child work in a mining industry that was additionally underground — 33 per cent of the workers in the lead mines of Cartagena were children between 8 and 16 years of age (Pérez de Perceval and Sánchez Picón Reference Pérez de Perceval and Sánchez Picón2005). It is equally worth highlighting that the as yet unpublished Spanish height series calculated by Martínez Carrión also shows a decrease in the height of those born between 1850 and 1875 (measured between 1870 and 1895) and this is also the case with the regional series of the South-East, Castile-León, Andalusia, Catalonia, Madrid and the Valencian Autonomous CommunityFootnote 47.

5 The living standard of Biscayan miners (1914-1936)

It can be seen in Figure 3 that the real wages of miners decreased between 1916 and 1920 because, although the daily rate improved following several strikes, the inflationist «gap» caused by the Great War (Figure 2) reduced their purchasing power. What is more, a flu pandemic affected the whole of Spain between 1918 and 1919, and therefore it makes sense to argue that miners’ welfare deteriorated between 1916 and 1920. The HDI (Table 2) does not reflect this because the index for 1910 is followed by the one corresponding to 1925, as it turned out to be impossible for us to calculate the 1920 index due to lack of data for life expectancy. Nevertheless, the San Salvador del Valle conscripts measured between 1917 and 1920 lost 0.51 centimetres with respect to the preceding cohort (Table A2 and Figure 4). Since their teenage years, or some of them, coincided with World War I, this drop would be the reflection of poorer nutrition and a higher morbidity rate.

The three indicators show that the living standard of miners improved after 1920. From that year to 1936, nominal wages grew at an average cumulative rate of 1.54 per cent, precisely when the demand derived from the work factor plummeted following the fall in the demand for Biscayan iron oreFootnote 48. This evolution of wages must be explained by the existence of a labour market that was now characterised by a bilateral monopoly, in the context of which wages were determined in advance as a result of the negotiation between employers and unions. In 1919, the Sindicato Minero (Miners’ Union) persuaded the Cámara Minera (Chamber of Mining) to accept the establishment of a minimum wage below which enterprises were not allowed to hire workers. The agreement also stipulated that wages would increase or decrease according to the price of sustenance. In 1922, the Sindicato Minero — controlled by a socialist majority — accepted a reduction in the wage agreed upon in 1919. This decision led to clashes with the communists as a result of which several people died. Nevertheless, the union managed to increase the minimum wage in 1924 and 1930Footnote 49. It is worth highlighting that the growth of wages modified the distribution of added value in a substantial way (Table 4).

Table 4 ADDED VALUE DISTRIBUTION IN THE BISCAYAN IRON MINING INDUSTRY (1876-1936)

Source: Escudero (Reference Escudero1998, pp. 282-285).

It must be remembered that real wages increased modestly (at an average cumulative annual rate of 0.35 per cent) during the expansion cycle of the mining industry. Between 1920 and 1936, however, the growth was considerable (2.65 per cent) and similar to the one calculated by Maluquer de Motes for average wages in Spain as a whole (2 per cent), wages that, as Rosés & Sánchez Alonso have demonstrated, experienced a strong convergence process after World War I (Rosés and Sánchez Alonso Reference Rosés and Sánchez Alonso2004; Maluquer de Motes Reference Maluquer de Motes2005).

In addition, the HDI (Table 2) shows a higher welfare index, which increased from 0.305 in 1910 to 0.501 in 1930 as a result of the improvement in its three components, particularly life expectancy, which «jumped» from the 0.320 index (39.1 years) to a 0.507 index (50.3 years). This is closely linked to better nutrition, bigger investments to improve sanitary conditions, the disappearance of overcrowding in private homes and increased public expenditure on preventive measures, especially during Primo de Rivera’s dictatorshipFootnote 50. Finally, height (Table A2 and Figure 4) shows an increase of 1.76 centimetres among the cohorts of conscripts measured between 1907 and 1910 and between 1932 and 1936, which is a reflection of the improvements achieved in the areas of nutrition, health and labour — the eight-hour working day was introduced in 1919 and piecework practically disappeared during the 1920s as a consequence of the increase in real wages.