INTERNATIONAL LABOUR OFFIGE STUDIES AND REPORTS Series N (Statistics) No. 14 METHODS OF COMPILING STATISTICS OF COAL-MINING ACCIDENTS GENEVA 1929 Published in the United Kingdom For the INTERNATIONAL LABOUR OFFICE (LEAGUE OF NATIONS) By P. S. KING & SON, Ltd. Orchard House, 14 Great Smith Street, Westminster, London, S. W. 1 CONTENTS Pagre INTRODUCTION 1 CHAPTER I. — Casualties and Accidents Casualties Accidents CHAPTER II. — The Demarcation of Industrial Risk in Coal Mining 10 Coal-Mining Risk 10 The Risk Groups 13 CHAPTER III. — The Measurement of Accident Risk Frequency Rates Severity Rates The Exposure to Risk Rates per Average Number of Workers Rates per Number of Manshifts (Days) or Full-Time Workers Rates per Number of Hours Worked Tonnage Rates and Rates per Aggregate Wages Paid. CHAPTER IV. — Classification by Causes V. — An International Comparison of Coal-Mining Accident General Rates Comparison by Causes Rates by Causes Percentages by Causes 5 5 8 17 17 21 23 24 26 28 34 37 CHAPTER 49 49 57 58 61 Notes on the Statistics of Coal-Mining Accidents in Different Countries 65 APPENDIX. INTRODUCTION The methods of compiling statistics of industrial accidents were discussed in a report of the International Labour Office submitted in 1923 to the International Conference of Labour Statisticians x, which adopted a series of resolutions thereon. The subject, however, is so vast that a general study could not treat all aspects exhaustively. The Office, therefore, has continued its investigations in this direction and has studied separately the industrial accident risk of certain branches of industry. In this Report statistics of coal-mining accidents are analysed from the point of view of the methods of compilation. In most coal-mining countries special mining statistics exist in addition to the general insurance or compensation statistics, and these only will be considered here. The main reason for the compilation of special statistics is probably the particularly high risk of accident in coal mining, which is closely connected with the fact that unforeseen events which may involve the lives of several persons are of special importance and frequency. Such events, which constitute disturbances in the regular operation of the mine, are to a large extent attributable to the fact that the work is carried on underground. Falls of ground, escapes of gas, and explosions of coal dust are events not entirely within human control. Their injurious consequences, however, frequently might be prevented. On the other hand, disturbances of operation caused by improper working, carelessness of the workers or fault of management are apt to be particularly frequent and easily result in serious injury owing to the fact that work is often carried on in the dark and on narrow underground roads. Access to the mine by the shaft also involves particular dangers. 1 Methods of Compiling Statistics of Industrial Accidents. and Reports, Series N, No. 3. Geneva, 1923. Studies — 2 — Finally, apart from these disturbing events, the risk of injury in any work carried on underground may be enhanced by the special conditions of the :nine and the dampness of the atmosphere. In compensation statistics, the case of injury is the only practical unit of the statistics, as the compensation paid for the injury is the point of interest for the insurance institution. Special statistics, however, attach interest mainly to the safety point of view, and therefore also to the disturbing events causing injury to persons. Actually a new unit is thus introduced which does not appear in insurance statistics : the " accident " as distinct from the case of injury, the term " accident " referring to the event causing the injury, whilst the case of injury refers to the person injured. The term " accident ", which in insurance statistics can only convey one meaning, viz. the case of injury, thus assumes a different and often ambiguous meaning in special statistics. Accident, on the one hand, and case of injury, which is termed here "casualty" in accordance with the terminology employed in British railway statistics, on the other, are easily distinguishable when a disturbing event interfering with the regular working of the mine occurs, as, for instance, an explosion of coal dust or a fall of ground. In many cases, however, it is impossible to distinguish the material event from the casualty. A haulage man, for instance, is crushed by falling from a tub. In this case " accident " and " casualty " constitute in fact one single case in a statistical sense. Nevertheless, even in these cases the event is frequently recorded separately, i.e. the case is recorded twice, once as an event usually termed " accident ", and once under the heading " person injured ". It is probable that this recording of the event resulting in injury, irrespective of whether it is a disturbance of operation or only a personal event, has led to ambiguity in the terminology of coal-mining accident statistics. The inconsistency in the terminology is still greater if special accident statistics for coal mining, shipping and railways are considered, in all of which the event is registered as a separate statistical unit. In Great Britain, the term " accident " in coal-mining statistics is applied vaguely in a general sense, while the term " separate accident " refers to the disturbance of operation or to any other unforeseen event causing injury to persons; the cases of injury are classified under the headings " deaths — 3 — and injuries " or " persons killed and injured ". The term " accident rates " is nevertheless applied to rates of cases of injury x. In the United States of America, " accident " generally indicates either the disturbing or the personal event. In coal-mining statistics, an accident is an event causing or resulting in personal injury, the consequences being designated as " fatality " and " non-fatal injury " 2. In Belgium and France, " accident " is an event involving injury to persons. The cases of injury are found under the headings " blessés " or " tués " or " victimes ". The term " Unfall " in German coal-mining statistics relates to the personal case of injury, while " Unglücksfall " is reserved for the injurious event 3. It would appear that, according to the predominant terminology, the term "accident" (Unglücksfall, accident) might best be confined to the event resulting in injury, while the term " casualty " (found in British and American railway statistics) could with advantage be accepted as the term designating the case of injury. In the course of this Report " accident " and " casualty " are taken in this sense. 1 In British railway statistics, an accident is an event causing either in jury or material damage, sometimes even an event likely to have caused loss of life or injury ; the case of injury is termed " casualty ", whereas in shipping statistics the term " casualty " designates, on2 the contrary, the disturbing event involving the vessel. In railway statistics, an accident is an event resulting either in injury to persons or damage to railway property, while the " casualty " is 3the case of injury. In railway statistics, however, " Unfall " is the event resulting in injury or damage, while the consequences are indicated under the headings " Tötungen und Verletzungen ". CHAPTER I CASUALTIES AND ACCIDENTS Casualties are reported under workmen's compensation legislation primarily for financial purposes. In order, however, to study safety in different industries and the prevention of accidents, special information must be obtained additional to that on casualties compensated. For coal-mining statistics, such information is largely derived from the reports under compensation laws and completed in some countries by reports under special legislation or special provisions in the compensation laws. CASUALTIES In compiling special statistics the data derived from reports under compensation legislation are usually adapted to the purposes of such statistics. These may include casualties notified but not compensated, their aim being the measurement of the risk during the period of reporting. German insurance statistics, for instance, include those casualties for which compensation was paid for the first time in the year of reporting 1. Coal-mining statistics, however, include all casualties notifiable under the insurance law, that is, all casualties disabling for more than three days. In Great Britain, the compensation statistics include all cases compensated for the first time in the year under review, but also, and separately, cases in which compensation was continued from the previous year. In the special coal-mining statistics furnished according to the provisions of the Mining Acts, only cases occurring in the current year are included. A further difference may arise in regard to the number of fatal casualties reported in the two sets of statistics. The compensation statistics generally include fatal cases in which compensation was paid to the dependants of the deceased or 1 Fatal casualties and casualties lasting more than thirteen weeks. Casualties lasting less than thirteen weeks are compensated under sickness insurance. — 6 — in which funeral and medical benefits were granted. As " fatal ", are considered those cases in which the death is attributed to the accident anc usually include cases from the previous year in which they were non-fatal ; on the other hand, particular rules on the definition of fatal casualties are usually laid down in the compilation of special statistics K In addition to these discrepancies between compensation statistics and special statistics, the industrial classification of the former does not always agree with the demarcation of the industry in the latter. German insurance statistics, for instance, include all casualties reported by the mining trade unions (Knappschaftsberufsgencssenschaften) in the Reich indiscriminately, while special statistics exist for coal mines in Prussia. In Great Britain, Belgium, and various other countries, compensation statistics cover accidents in all mines, while special statistics distinguish coal mines from other mines. Though the objects of the special statistics, however, are evidently much the same in the different countries, the definition of reportable casualties in these statistics differs widely from country to country. With regard to fatal casualties, an accident may not result in instantaneous death but cause an injury leading to death after a shorter or longer period of incapacity. Each country decides how far death occurring after the accident shall still be attributable to the accident. In British coal-mining statistics, for instance, a fatal casualty is one which leads to death up to the date of the annual report, with the restriction that a casualty is no longer classified as fatal if death occurs later than one year and one day after the accident. In German and Dutch coal-mining statistics a casualty ending in death before the completion of the annual report is recorded as fatal. The definition is rather vague in French statistics (death must occur immediately after the accident or at most some months later). In the United States no period between the accident and the death is fixed, but in practice statistics include those casualties that had resulted in death when the State Mine Inspector of each State furnished his summary to the Bureau of Mines. In Belgium, death 1 For further discussion and illustration oí this point, ci. NIXON : " On Some Principles of Accident Statistics ", International Labour Review, Dec. 1928. _ 7 — must occur within thirty days from the occurrence of the. event. In Czechoslovakia a fatal casualty is one which causes instantaneous death or leads to death, no limit being fixed for the period which may elapse between the accident and death. It will thus be seen that fatal casualties are not, strictly speaking, comparable from country to country or even as between different statistics of the same country. Three methods are applied for compiling fatal casualty statistics. The first method reports as fatal all casualties which lead to death up to the issue of the annual report. According to another method, the period which may elapse between the accident and death in order to render death accidental is limited. A third method adds to the limitation of this period the indication of a final date after which the fatal issue is no longer considered for the special statistics. While the limiting of the period elapsing between accident and death is based on the consideration that the causal relation between accident and death becomes doubtful if too long a period elapses between the two, the other limit, the final date, is necessitated by the fact that the issue of the casualties occurring towards the end of the period of observation cannot be awaited as the publication of the report would be greatly delayed. It is evident that casualties occurring in the year of observation may lead to death after months or even years. For this reason insurance statistics in some countries are revised after some years or published for longer periods. Such a revision can only be carried out by the insurance institution which has control over the consequences of the casualties. It is a problem concerning accident statistics generally, and therefore cannot be treated here. For special statistics, however, agreement should be reached between the different countries on the following points : (1) Should the period which may elapse between the occurrence of the accident and death be fixed ? (2) What date is to be taken as the final date after which a fatal issue is no longer to be taken into consideration ? It has been seen that the term " fatal casualty " is not a uniform conception. The term " non-fatal casualty " evidently is still less uniform. In Prussian statistics all casualties disabling — 8 — totally or partially for more than three days are reported since 1923, while prior to that date only fatal casualties were reported. In Great Britain the limitation was seven days of incapacity up to 1923 and three days since. In France all casualties disabling for more than four days are published. In Belgian statistics, casualties disabling permanently only are included. In Holland all casualties disabling for more than three weeks and in India all casualties enforcing absence from work for more than twenty days or leading to serious injury as defined by law are published. In Japan all non-fatal casualties are included but classified as serious and slight, the latter including all injuries disabling up to one week. In the United States of America (federal statistics) non-fatal casualties are excluded altogether, as the non-fatal casualties reported in the different States are incomparable. Some States report serious casualties only (the definition of serious being even left to the discretion of the reporting inspectors in some cases) and the limits with regard to the severity of the reportable casualties vary. In Czechoslovakia, serious casualties are, on the one hand, those which are serious in respect to the injury, that is, those which lead to mutilation or permanent ill-health, and, on the other hand, those serious only in respect to the duration of incapacity, the latter including casualties disabling for more than nineteen days. The effect of these differing limits of reportable casualties on the comparability of frequency rates of non-fatal casualties will be seen in the international comparison of coal-mining casualties. It is evident that comparison of non-fatal casualties is almost impossible so long as no agreement regarding their classification by duration is reached. Classification by duration is not often made in special statistics except in the case of classification according to temporary and permanent incapacity. This classification can only be based on insurance data and is a general problem of accident statistics. It cannot therefore be treated in this Report. ACCIDENTS Data on accidents as separate statistical units are usually furnished along with the reports notifying the casualties. Where the material event cannot be clearly distinguished from the case of injury, accident and casualty are identical statistical units. Some countries provide for a separate and special — 9 — notification of events involving injury to several persons or of a nature which has or might have had serious consequences. In Great Britain, for instance, certain accidents considered serious have to be reported to the divisional inspector of mines under the Mines Act, 1911. Accidents are thus reportable if they result either in some specific kinds of injuries or are caused by explosion of gas or dust, or by any explosive, by electricity or by overwinding, and cause any personal injury whatever. These accidents are published separately. Generally, all events resulting in injury, disturbances of operation as well as other injurious events, are registered separately as accidents. This practice is adopted also in Indian, French and Belgian statistics. An accident involving both fatal and non-fatal casualties would be classified as fatal. In the German statistics, accidents are altogether omitted from the tables, notifiable mine disasters only being separately mentioned in a descriptive way. Dutch statistics are also confined to casualties. The United States statistics report the principal mine disasters ; first, those are reported which resulted in a hundred or more deaths and secondly all accidents in which five or more persons are killed. Special accident tables have been published from time to time in which accidents are classified according to the number of persons killed. The number of accidents approximates very closely to that of casualties where all the injurious events are registered as " accidents ". It would appear, however, that interest attaches only to those events which constitute disturbances of operation and involve danger to one or several persons. If these events only were registered as statistical units, distinct from the cases of injury, the prevalence of disturbing events as well as their importance with regard to casualties could be determined. CHAPTER II THE DEMARCATION OF INDUSTRIAL IN COAL MINING COAL-MINING RISK RISK By the " coal-mining industry " is generally understood the industry engaged in the extraction of coal. Accident statistics aim at showing the industrial risk of coal miners, i.e. the risk incurred during and in consequence of employment. Accidents, therefore, sustained by the miner outside his employment are not taken into account. In order to show this risk, it is necessary to define it in such a way as to demarcate it from other industrial risks on the one hand and from non-industrial risks of the workers on the other hand. The demarcation can either be based on the " enterprise ", or on the " premises ", or on the " worker ", viz. the operation in which the worker is engaged. (1) A " coal-mining enterprise " is a commercial or financial unit and may cover processes of different technical character combined for financial purposes, and therefore includes premises annexed to the mines, as coke ovens, briquette works, etc., where the coal raised is further treated and transformed, and even iron foundries, power stations, and other works attached to the mine in order to render the enterprise independent of raw materials and other supplies. The risk in these processes and premises is evidently very different, as it depends largely on the technical process in which the worker is engaged. The enterprise does not therefore seem a suitable criterion of demarcation. For financial reasons, a colliery company may possess many kinds of industrial works. The criterion " enterprise " is, in fact, not applied in coalmining accident statistics. — 11 — (2) The operations, i.e. the processes in which the workers who are victims of the accidents are engaged, is applied as criterion in the Prussian and Indian statistics. In the former, all casualties to mine workers (Belegschaft) and technical employees in consequence of the mining process (infolge des Bergwerkbetriebs) during the shift are included. Thus the operation of the worker and the causal relation between the casualty and the process determine the inclusion of the casualty, while nonindustrial risks of the workers are excluded by the phrase "during the shift " 1. Coke ovens, briquette works and other processes for treating coal are considered as belonging to the mining process. Similarly, coke making and the dressing of minerals are covered by the Indian statistics. The statistics are then grouped according to the minerals raised and prepared, viz. coal, lignite, iron-ore, etc. This method of demarcation, however, allows of different applications ; and the inclusion of the ancillary processes of dressing and treating coal extends the statistics to risks differing essentially from the risks peculiar to the mine, i.e. the raising of minerals underground. (3) A third criterion of demarcation is the premises on which the casualties happen. It is applied in most countries, e.g. Great Britain, Belgium, France, and the Netherlands. According to this criterion, all casualties happening to workers in and about the mine are covered by the statistics. By this method casualties to the workers outside their employment are automatically excluded. The exclusion of certain non-industrial " accidents " which might happen in or about the mine, such as suicide, homicide, and accidents due to ill-health, etc., is specially provided for. The results of the second and third methods are only slightly different with regard to the mine workers properly so called. By a demarcation on the basis of the premises, however, ancillary works, such as those above cited, are excluded from the statistics — as in Great Britain, France, Belgium, etc. On the other hand, not only technical employees. but also clerical staff are covered, if not explicitly excluded as 1 It follows that casualties not connected with the mining process, as those caused by lightning, suicide and homicide, casualties in consequence of unhealthy constitution and casualties on the way to and from work and to other than mine workers, are excluded. — 12 — in Belgium, as they also work on the premises. Though clerical staff have risks very different from those of the miners, they form only a very small percentage of all mining workers, and, if excluded, must be considered in another department of accident statistics. As coal-mining accidents are under consideration, it might be argued that miners hewing or hauling other minerals than coal in the same mine where mainly coal is raised are not " coal miners ". The mining of such minerals, however, being carried on under exactly the same conditions as that of the coal, possibly by the same workers, will involve much the same risks with the same causes. To conclude, the third method of demarcation, by the premises, would appear to be the most suitable one for coal-mining accident statistics. The term " coal " however, may refer to several kinds of coal, such as bituminous, anthracite and lignite, for which the technical process differs. The extraction of lignite does not necessitate the sinking of shafts, and is therefore not liable to the same risks as pit-coal mining. Bituminous and anthracite coal mining may also entail different risks owing to the, natural conditions under which they are raised. The following tables show the risks attaching to anthracite, bituminous and lignite mining. PENNSYLVANIA : FATALITIES P E R 1 , 0 0 0 3 0 0 - D A Y Bituminous Year 1920 1921 1922 1923 1924 1925 1 WORKERS1 Anthracite Underground and surface Underground Underground and surface Underground 3.01 2.98 4.51 2.91 3.43 2.95 3.35 3.18 4.87 3.14 3.71 3.26 3.74 3.80 3.81 3.62 3.39 4.12 4.55 4.84 4.60 4.38 4.05 4.80 U.S. DEPABTMEKT OF COMMERCE, BUREAU OP MIKES : Coal Mine Fatalities in the: United States, 192«. Bulletin 283, pp. 90 et aeq. — 13 — FATAL CASUALTY RATE I N COAL MINING AND LIGNITE MINING IN PRUSSIA Year 1911 1912 1913 1920 1921 1922 1923 1924. 1 1 Pit-coal raining: fatal casualties per 1,000 workers Lignite mining : fatal casualties per 1,000 workers 2.01 2.54 2.48 2.31 2.00 2.05 1.60 2.21 1.69 1.72 1.48 1.59 1.36 1.35 1.37 1.46 MINISTEBIUM FÜR HANDEL UND GEWEBBE, PBEÜSSEN : Zeitschrift für Berg-, Hütten- und Salinenmeten im preussischen Staate. Berlin. Lignite mining shows a considerably lower risk than pit-coal mining, and bituminous coal mining a lower one than anthracite mining. The latter difference, however, might be partly due to the different classes of labour employed in the two coalfields : more immigrant and non-union labour being engaged in anthracite, and more national and union labour in bituminous, coalfields. Figures, therefore, should be published separately for the different branches of coal mining. T H E RISK GROUPS A coal mine contains two distinct parts : underground workings and surface workings. Underground work has risks peculiar to mining both with regard to the high incidence and the causes of the casualties, which are largely disturbing events easily involving several injuries. A distinction between underground and surface risks is therefore fundamental. There is another feature peculiar to the mine premises which calls for attention : the shaft. The shaft is the way between underground and surface through which all underground workers and the coal must pass. As the shaft is usually, in European mines, a vertical passage requiring special mechanical haulage, accidents may happen in the shaft though the miners do not work there ; and these accidents are apt to be serious, since the breaking of ropes or chains, overwinding and even personal events, as falls, are more dangerous in the shaft than elsewhere below ground. — 14 — For these reasons the shafts are in some statistics considered as separate premises of the mine, like underground and surface. In other statistics, however, shaft accidents are considered as a separate group of causes only. This is due to the fact that for the calculation of rates there is no " shaft exposure " which could be separately calculated, there being no " shaft workers " as there are surface and underground workers. A separate assessment of underground and surface risks, however, is not sufficient, as these risks are unevenly distributed between the different occupations, underground and surface respectively. Underground hewers may be most exposed to falls of ground, while haulage men are most exposed to risk of haulage. The following table for anthracite coal mining in Pennsylvania illustrates the risk distribution of underground workers in those mines. However, it is not necessarily representative of risk distribution in other mines. NUMBER KILLED P E R 1 , 0 0 0 WORKERS UNDERGROUND Tears 1900-1913 1 Mine Fire foremen bosses and as- and assistants sistants 2.98 3.86 Miners 6..00 Miners' Drivers labourand ers runners 4.29 3.35 1 Door boys, etc. AU others 3.86 2.06 XL S. DEPARTMENT OF THE INTEBIOB, BUBKAC OP MINES : Coal Mine Fatalities in the United States, 1870-1914, p. 95. Washington, 1916. By far the greatest risk is found for miners properly speaking (hewers and hewers' labourers), while the risks are low in the cases of foremen and assistants and other underground workers not directly employed in the getting of coal. The occupations of surface workers are also very varied, as besides the workers hauling, sorting and washing coal, skilled workers are employed, such as enginemen, plumbers, and sawyers, whose work and risks are similar to those in other industries. Technical employees, mainly entrusted with management and surveying, may be working underground or on the surface. Their accidents should be classified as underground or surface accordingly. — 15 — A classification of accidents into underground and surface is made in all coal-mining statistics. Classifications by occupations, however, are not, to our knowledge, given except in the Pennsylvanian statistics, in the reports of the miners' insurance funds, and in the English occupational mortality statistics \ In speaking of "underground" and "surface", however, one may mean either the premises on which the accidents or casualties happen or the workers to which they happen. Underground workers may have accidents whilst on the surface and surface workers may be injured whilst temporarily assisting in some work underground. In the British report for 1925, we find, for instance, that the number of " persons injured underground " and " persons injured on the surface " in the table of causes was 162,687 and 14,660 respectively, whilst in the classification by nature and duration of injury the number of " underground workers " injured was 163,253 and of " surface workers " injured 14,094 ; t h e difference is due to a certain number of underground workers being injured on the surface and vice versa. In the first case, the classification of casualties was made according to the premises, because the causes depend on the premises ; in the second case, the classification was made by the workers' occupations, probably because the consequences of the injuries were known for each person only. It is evident that statistics on the risk of underground work and its causes should contain all accidents happening underground. The premises appear therefore to be the most suitable criterion of classification. It is otherwise, however, with regard to occupations. Most workers have their regular occupation, no matter where they happen to perform their job. A classification by occupations only aims at showing the occupational risk of each occupation, as distinct from the industrial risk of the whole industry, i.e. of the premises. The classification by occupations must nevertheless group occupations according to underground and surface, as a technical worker and a repair or haulage man usually employed in underground work cannot be grouped together with the same kind of worker mainly employed in surface work, their risks being too divergent. The occupational 1 Accidents in the new occupational mortality report for 1921-1923 comprise non-industrial casualties. Moreover, the " exposed to risk " contains also retired workers. — 16 — classification of casualties by underground and surface will, however, not entirely coincide with the classification by underground and surface premises. As the terminology for coal-mining workers varies greatly in different coalfields and countries, a uniform detailed international classification by occupations would be difficult. The following summary classification, however, might be attempted : Underground workers : (a) hewers and assistants ; (b) haulage men ; (c) repair workers ; (d) foremen, deputies, overmen, etc. ; (e) other workers* Surface workers: (a) colliery weighmen ; (b) coal washers ; (c) coal pickers (coal cleaners, etc.) ; (d) banksmen ; (e) other haulage men, dirt tip men, etc. ; (/) craftsmen (plumbers, enginemen, etc.) ; (g) technical employees ; (h) other surface workers. Clerical worke.rs. The tabulation and classification of casualties and accidents, however, even if these are classified according to duration, does not in itself furnish a comparison of the risks. By the " risk " ismeant the incidence of casualties as compared with the exposure, and the measurement of this risk and of exposure offers special problems in coal mining. CHAPTER III THE MEASUREMENT OF ACCIDENT RISK The number of casualties alone does not give any information as to the comparative risk incurred by the workers in different countries or at different dates, as it necessarily depends to a large extent on the number of workers engaged in the industry and the amount of labour performed. The risk can only be measured by relating the number of casualties and their consequences to the units of exposure to risk. The calculation of rates, however, is a general problem of accident statistics and we can only touch on some special aspects in this Report. As both the number and the importance of casualties differ, it is essential to measure both the frequency of the cases and their severity. The severity rate is calculated by relating the time lost to the exposure. The calculation of frequency and severity rates offers certain difficulties as the rates are affected by non-occupational factors, whose effects should be eliminated if we desire to compare the industrial risks. FREQUENCY RATES Frequency rates should be independent of factors which may be unequally distributed in the areas or industries concerned. The main non-industrial factor affecting frequency rates is age. There are circumstances which suggest that workers may not be equally liable to accidents at every age ; new workers entering mining, not being fully acquainted with the risks, are likely to be less liable after they have adapted themselves to their surroundings and learnt how to avoid dangers. The reason is, of course, not their age but their inexperience ; it is related to the age factor as the greatest number of new workers are young workers. As national coal-mining statistics do not generally publish age-rates x, an instance is given from Heymann and Freudenberg's Morbidität und Mortalität der Bergleute im Ruhrgebiet (Morbidity and Mortality of Ruhr Miners), and some data 1 The Report was already in the press when the British report for 1927 appeared, which contains a classification of fatal and non-fatal casualties by age-groups and total casualty rates by age-groups. The groups are as follows : under 16 years, 16 and under 18 years, 18 and under 20 years, 20 years and over. _ 18 — F R E Q U E N C Y O F CASUALTIES BY AGE GROUPS I N T H E RUHR COAL DISTRICT 1 Fatal and Non-Fatal Casualties per 1,000 Miners Age Up to 20 20-30 30-40 40-50 Over 50 Total 1907 1908 1909 1910 1911 1912 1913 19U 237.1 201.7 162.5 168.9 75.6 236.8 188.1 149.8 156.3 65.7 237.5 187.8 145.8 152.6 68.7 236.9 190.0 155.9 164.8 75.1 236.5 190.1 162.4 172.9 85.7 243.5 186.8 173.4 183.3 89.3 266.4 188.8 164.9 192.1 90.3 249.8 201.5 177.9 215.2 99.9 177.5 166.5 165.5 171.5 178.8 178.7 183.7 194.0 Age 1915 1916 1917 1918 Up to 20 20-30 30-40 40-50 Over 50 229.0 179.1 180.3 201.8 103.4 230.4 173.5 179.6 211.7 116.4 206.6 197.0 213.9 192.8 96.6 200.5 239.7 261.8 220.9 113.5 Total 1919 1920 1921 1922 183.9 180.7 190.7 112.6 116.3 100.8 114.5 86.7 81.8 144.9 111.8 104.8 107.0 95.1 89.7 — 185.7 188.5 187.7 211.8 133.4 120.2 113.7 102.0 Deaths by Accidents per 1,000 Miners Age 1907 1908 1909 1910 1911 1912 1913 1914 i Up to 20 20-30 30-40 40-50 Over 50 1.8 2.0 1.5 2.2 1.0 2.3 2.9 2.8 3.1 1.4 1.7 2.8 2.7 3.0 1.6 1.6 2.1 2.5 2.5 1.2 1.6 2.0 2.3 2.5 1.6 2.4 2.3 2.9 3.5 2.0 2.1 2.3 2.6 3.0 2.0 2.0 2.3 2.3 3.2 2.0 Total 1.8 2.7 2.5 2.1 2.1 2.7 2.4 2.4 Age 1915 1916 1917 1918 1919 1920 1921 1922 Up to 20 20-30 30-40 40-50 Over 50 2.5 2.4 2.8 4.3 2.5 2.3 2.2 3.6 5.4 3.1 2.4 2.7 4.6 5.3 3.8 1.6 2.3 4.4 4.2 2.7 1.7 1.5 2.2 3.7 3.0 1.5 1.7 1.8 2.6 2.8 1.5 1.4 1.5 2.4 2.3 1.4 1.5 1.7 2.5 2.2 Total 2.9 3.3 3.7 3.0 2.2 1.9 1.7 1.8 1 HEYMANN and FREUDENBERG : Morbidität und Mortalität der Bergleute im Ruhrgebiet, pp. 170 et seq. Essen, 1925. — 19 — ENGLAND AND WALES : FATAL CASUALTIES IN COAL MINING BY AGES, 1910, 1911, AND 1912 » Age-group 15-19 20-24 25-34 35-44 45-54 55-64 65-74 75 and upwards Workers in coal raining * Fatal casualties Fatal casualties per 1,000 of population 429,486 396,222 677,898 512,451 316,455 170,814 65,982 13,185 466 449 797 663 457 279 56 2 1.08 1.13 1.17 1.29 1.44 1.63 0.85 0.15 2,582,493 3,169 1.23 ' Mortality of Men in Certain Occupations. Supplement to the 75th Annual Report ol the aRegistrar-General, Part IV, p. 25. Census population of 1911 multiplied by three. Retired -workers are included. from the English occupational mortality statistics of miners. The British figures are taken from the occupational mortality report for 1910-1912, as the new report for 1921-1923 does not give the mining accidents separately, but classifies industrial and non-industrial accidents of coal miners together. According to the Ruhr statistics, which refer to fatal and non-fatal industrial casualties of the occupied population, the frequency of casualties generally (fatal and non-fatal) is highest at youngest ages, that is, up to 20, and then drops down to 40, rising again between 40 and 50, and then decreasing very considerably. The high frequency rate at youngest ages and the subsequent decrease are probably due to the fact already mentioned that young miners are not yet accustomed to their work, nor acquainted with its dangers. It may be added that, according to the English census statistics for 1921, a large proportion of young men are engaged in haulage work, the accident risk of which is very great. It is of course difficult to decide how far high frequency rates at younger ages may be due to the dangers of the haulage work or how far the employment of young persons entering on mining in haulage work raises the risk of this occupation. The decrease of the total frequency rates towards higher ages will, moreover, be accentuated by the elimination of unfit or weaker workers who leave the occupation or die in consequence — 20 — of sickness or accidents. The rise between 40 and 50 in the Ruhr is difficult to explain. The considerable decrease after 50 might be due to the fact that older men do not take on the more dangerous jobs, or if so, only the specially experienced and healthy workers. The age curve of the fatal casualty rate does not show the same trend as the general casualty rate. The rate on the whole rises with age up to 40 to 50 but drops rather after that age, though the drop is much less marked in post-war years. The rise may be due to the fact that older men are more liable to disease and death generally and their injuries therefore lead more frequently to death. The drop after 50 can probably be explained in the same way as that of the total frequency rate. Occupational mortality statistics for England and Wales show that fatal casualty rates of coal miners rise with age, though much less so than the general death rate. The decline in the case of the highest age groups is probably due to the inclusion of retired workers in the statistics of exposure. It may be concluded that age has an influence on accident frequency. A classification of rates by ages is therefore desirable in order to compare the risks at different ages. Moreover, the different age incidence of casualties must be taken into account in the calculation of total casualty rates (all ages) as the age distribution of coal miners differs in the countries or areas compared. Total frequency rates should therefore for international comparisons be standardised in regard to the age distribution of the population under observation. A problem similar to that of age would arise for sex were it not that the proportion of women employed in coal mining is negligible in most countries. Those employed, moreover, are mainly engaged on the less dangerous jobs. In coal-mining accident statistics, casualties, but not rates, are occasionally published for the sexes separately. There are other non-industrial factors besides age and sex which may be differently distributed and affect casualty rates, such as race and nationality of the workers. Certain nationalities may be less careful and less skilled. It is evident that fatal casualty rates will tend to be higher for those " races " which are generally more liable to disease or have higher death rates. Apart from these " natural " factors, there are numerous social and technical factors affecting casualty rates : the — 21 — standard of living of the industrial population, for instance, which by influencing the health standard may weaken resistance to the effects of casualties, and tend indirectly to affect rates. The influence on accident frequency of labour conditions and technical organisation, however, should not be eliminated in the rates, as it is the aim of accident statistics to show the differences in the risks due to different labour conditions and organisation of production as well as those due to natural conditions. Enquiries into these conditions are then no longer a matter of statistics. The industrial risk of accident, however, is not only or even mainly expressed in the frequency of casualties. If frequency rates were calculated for casualties classified by duration of casualty, the comparison would be more fruitful, as the time lost per case could be calculated, viz. the average duration of the case. The measurement of risk by frequency rates, however, can also be completed by the measurement of the risk by severity rates relating the days lost directly to the exposure to risk. SEVERITY RATES Severity rates, if correctly calculated, would in fact be a more comparable measure of accident risk than frequency rates. Their calculation, however, requires not only complete data, but gives rise to certain difficulties owing to the fact that the " time lost " by casualties is not always known. Cases of injury may disable temporarily, permanently, or lead to death. Casualties are usually reported for annual periods and the exposure to risk is either the average number of workers employed or the time worked during the year covered by the statistics. The consequences of a casualty, however, may last beyond the end of the year : a temporary case may either last longer than one year or, as is more frequent, may begin towards the end of one year and last into the next year. The " time lost " in consequence of permanent disablement and death may be taken as the remainder of the normal or average working life or part thereof. The total time lost as a result of temporary incapacity within the year of accident occurrence can be related to the exposure of that year as it represents the time actually lost as compared with the time worked. The temporary severity rates, however, — 22 — would be inadequate as the time lost by cases lasting beyond the end of the reporting period is not comprised, and this time lost, escaping measurements, will differ in the areas or in the States compared. This could be remedied to a certain extent by computing in the rates of each period all time lost by casualties in that period, including time lost in consequence of casualties which happened in the preceding year. The time lost by cases of death and permanent disablement can only be estimated. The problem of computing severity rates, however, is one of accident statistics generally, since the data are mostly derived from insurance statistics. An American investigation into the severity of coal-mining casualties in 1925 was based on the reports of fifty-nine coal mines, furnished by the operating companies which participated in the national safety competition of 1925 *. The reports covered all " lost-time " casualties which occurred during the year, that is, those which disabled an employee beyond the day or shift on which the accident occurred. The mines were situated in eleven States and covered each at least fifty men underground. Each fatal case and case of permanent total disability was weighted by 6,000 lost days according to the scale of the International Association of Industrial Accident Boards and Commissions 2. Permanent partial disability was weighted according to the nature of the injury by using the scale of the Board. Temporary injuries were weighted according to the actual number of calendar days during which the employee was incapacitated for work. It must be noted that in the " 6,000-day weight ", the working days only are taken into consideration (20 years = 6,000 days). The time lost by temporary incapacity should consequently also include working days only 3, which might be done as in Sweden by dividing the days lost by 365 and multiplying the result by 300. The average frequency rate was found to be 123 per million hours of exposure, and the severity rate 10.214 days lost per thousand hours of exposure. The two rates do not vary in unison as the following tables show. 1 W. W. ADAMS : Accident Severity Rates for Certain Coal Mines. Reports of Investigations, Department of Commerce, Bureau of Mines, Serial No. 2783, Nov. 1926. (Typewritten report.) 2 Fatalities due to explosives were 'weighted by 12,000 days as required by the rules governing the competition. 3 It is not clear whether this is the case here or not. — 23 — F R E Q U E N C Y AND SEVERITY RATES IN U N I T E D STATES MINES, 1 9 2 5 Size of mine (number o E employees) Number of mines Hours worked 1-99 100-199 .... 200 or m o r e . . . 6 17 36 828,312 4,960,099 23,329,150 State Pennsylvania : anthracite bituminous West Virginia Utah Total Total Accident Accident number number frequency severity of of rate rate accidents days lost 71 431 3,070 19,647 85.716 23.719 54,722 86.893 11.032 223,048 131.595 9.561 Hours worked Accident frequencyrate per million hours Accident severity rate per 1,000 hours 4,890,295 3,684,991 3,946,989 5,706,078 1,327,463 1,622,929 94.677 50.746 47.631 322.638 109.984 133.093 8.871 10.772 9.114 10.453 15.022 19.528 It appears that the largest mines have the highest frequency rate but the lowest severity rate, and the smallest mines the lowest frequency rate and the highest severity rate. The report comments upon this fact as follows : It would thus appear that the larger companies are meeting with reasonable success in keeping down their death rates, but that minor injuries at their place still demand considerable attention it losses caused by them are to be reduced. It appears from the instance given that severity rates may differ very considerably from frequency rates and that their calculation would be of the greatest interest. Methods of calculating severity rates, however, cannot be discussed here. T H E E X P O S U R E TO R I S K The worker is evidently exposed to risk of accident during the time he spends at work, which, for the underground worker means the time spent underground including the time for descent and ascent. Time worked in mining differs in two respects from manufacturing industry: (1)mines do not usually or everywhere work continously throughout the year, but may work a varying numbers of days ; and (2) the time worked by the miners differs from mine to mine as well as individually. The mine worked thus depends on various factors which may — 24 — vary from country to country such as : (a) the regular (agreed or legally prescribed) hours of work, (b) the overtime or supplementary hours worked, and short-time shifts as well as winding time, and (c) the time lost by individual absences of workers employed at the mine on account of sickness, accident» holidays, or other reasons. All these factors must be taken into account in assessing the exposure to accident risk of coal miners. This, however, is done in very varying degrees in the different statistics. It is the aim of the following analysis to enquire to what extent the rates calculated in the different countries allow for the various time factors of exposure and to arrive at some conclusion as to the best method of assessing accident exposure in coal mining. In coal-mining accident statistics, the following casualty rates are calculated : (1) Rates per average number of workers (a) employed, and (b) present ; (2) Rates per number of manshifts worked, or rates per man-days worked, or per full-time worker ; (3) Rates per number of hours worked ; (4) Rates per ton of coal raised ; (5) Rates per aggregate wages paid. Rates per Average Number of Workers Rates per average number of workers are calculated by different methods in the various coal-mining countries. (a) The method employed in Great Britain and in the Netherlands, for instance, is to record the number of workers employed at certain dates and to average these figures. Such an average, however, will only be correct where the number of workers employed at the mine on these dates is representative of the daily number of workers employed. In the Dutch statistics, the number of workers derived from fortnightly statements on the wages books is divided by the periods of payment. In Great Britain, quarterly statements of the numbers on the colliery books are utilised. (b) The method employed in Belgium, France, and India is to add up the number of man-days actually worked or days — 25 — of presence and divide the sum by the number of days worked by the mine. In Belgium and France the " jour de présence " on days only when coal was wound are divided by the average number of these days. In India all " daily attendances " are divided by all days worked by the mine. The so-called " average number of workers " thus found is, however, only the average number of workers present on each day the mine worked or coal was raised. It is not the average number of workers employed ; it must be lower than this average, as man-days lost owing to individual absences of workers employed but not present on days the mine was open are not taken into account. Nor does the average number present, on the other hand, represent the time worked. The mandays actually worked, representing time worked, are divided by the days on which coal was wound, and the differences in time worked due to differences in the number of days on which the mine was idle are thus eliminated 1 . The correct method of calculating the average number employed 2 would be to include in the total number of man-days of the dividend not only those worked, but also those lost by workers employed but absent owing to sickness, leave, etc. This is done in Dutch statistics, where sick, injured and absent workers are included in the " exposed to risk ". All man-days worked and all man-days lost on all days when any work is carried on in the mine should be included in the dividend, and the sum divided by these days. Man-days lost should then also comprise man-days lost on days when only few workers worked, i.e. man-days lost owing to partial idleness of the mine, but not those lost by complete idleness of the mine â. 1 For instance : in one mine, which works four days a week, 100, 105, 120, and 75 workers are respectively present. In another, six days are worked by 100, 120, 80, 140, 60, and 100 respectively. In both cases the average number present is 100, yet the man-days worked are 400 and 600 respectively. 2 Rates of risk calculated by using this figure would be unsatisfactory, as workers who are sick or otherwise absent are not exposed to risk. 3 Moreover, the average number of workers present must be relatively smaller if man-days or manshifts on all days on which work was done in the mine are divided by these days than if man-days worked on days only when coal was wound are divided by this latter number, as fewer workers will be working on the days when no coal was raised. — 26 — The " number of men employed " in the United States coal-mining accident statistics is not based on any prescribed method, but the calculation is left to the reporting operators. According to the Geological Survey *, which prior to 1925 computed these figures, some of ;hese apply the method of sampling by taking the average of each of the pay-rolls during the year. The Geological Survey states that " this method inflates statistics to the extent that the same employees appear on the pay-rolls of two or more operations in the same period, that is, duplication is introduced in proportion to the rapidity of the labour turnover". Operators did not, however, include pay-rolls for periods of strike or complete idleness of the mine. The Geological Survey found that usually, however, the figures reported were " not the average number of men actually working at any time, nor the aggregate number of men who had been working at the mine during the year, nor the absolute average number of men on the pay-rolls, but rather the number of men commonly dependent on the mine for employment " . They represent the number ordinarily reporting for work when the mine starts plus the absentees, that is, the men who have been working recently and who will work again, but who for one reason or another are not available. Rates per Number of Manshifts (Days) or Full-Time Workers Rates per average number of workers employed may actually indicate variations of accident incidence in a coalfield or differences between several coalfields of a country in relation to employment and may be a suitable basis for insurance calculations. Such rates, however, are not suitable for the comparison of the risks peculiar to each mine or coalfield. For international comparison these rates are misleading because the time worked in one country may be longer than that worked in another country. The same number of workers employed is therefore exposed to risk during a longer time within the same period in the one mine than in the other. We have seen that the time worked by the employee in a mine may vary in two respects : (a) according to the number of days worked by each worker employed, which may differ (i) by the number of 1 1922, U.S. GEOLOGICAL SURVEY : Mineral pp. 494 et seq. Resources of the United States, — 27 — clays the mine was in operation during the period of observation, and (ii) by the number of days on which the individual worker was present ; and (b) according to the daily hours worked by each employee present. The mine may be idle during some days of the week or during certain seasons of the year as well as in consequence of strikes or lock-outs. Thus in the United States the number of days worked varies enormously from State to State. In 1925, the number of days worked by the mines varied from 112 in Arkansas to 254 in Virginia 1. These differences may be clue to natural conditions or to commercial reasons. On the whole, as was stated by the Geological Survey, which formerly computed these figures, the larger mines work more steadily than the smaller mines. In addition, the number of days lost by individual absences will differ. A measure of accident risk taking account of this time factor is the casualty rate related to the number of man-days or manshifts worked or to full-time workers based on man-days or manshifts worked. The number of normal manshifts, i.e. manshifts of the legally prescribed or agreed number of hours, and of man-days are the same where no overtime shifts are worked. Conditions, however, vary with regard to overtime. Where practically no overtime is worked, as in Belgium and France, man-days and manshifts are identical. In Great Britain, manshifts are calculated by including overtime and short-time reduced to shifts of normal length. A further difference must arise where, as in the United States, the number of man-days is calculated by multiplying the number of employees by the number of days the mines were in operation and coal was wound (tipple days). In order to compute these " a c t i v e " days, the " o p e r a t o r " reporting is asked to state the total number of full days the mine was in operation, parts of days being reduced to full days 3. 1 Coal Mine Fatalities in the United States, 1926, op. cit., pp. 65-66. The average number of days in each State is a weighted average. Days worked by each mine are multiplied by the number of men employed at that mine and the sum of these products for all the mines in the area considered is divided by the sum of the men employed. The weighted average is considerably higher than the unweighted, owing to the fact that large mines usually work more steadily than small mines. 2 — 28 — The number of man-days thus calculated, which is divided by 300 in order to arrive at full-time workers, is not necessarily a correct measure of the man-days actually worked, as mandays lost by absence of underground workers on days when the mine is working are usually included, whilst man-days worked by underground or surface workers on days when no coal is raised are excluded. The so-called " full-time worker " of Prussian statistics was for the years 1921 to 1925 found by adding up all regular manshifts worked, short-time shifts being reduced to shifts of normal length and the total sum divided by all ordinary working days of the period. Overtime shifts were left out of account. The measure thus obtained may diner according to the number of general holidays in the respective coalfields. The measure was really calculated for the purposes of wages statistics and meant to represent the full-day worker, for the calculation of which overtime shifts are irrelevant, being worked on the same days. For accident exposure, however, overtime represents additional exposure. The calculation was in fact revised in 1926, overtime being included in the exposure. Rates per Number of Hours Worked Rates of casualties per manshift or man-days worked, or per 1,000 full-time workers calculated on man-days, take account only of the number of days worked, but not of the number of hours worked. The hours worked may differ (a) with regard to the length of the normal shift worked, and (h) with regard to the overtime worked in addition to the normal shift. (a) Regular Hours of Work. — If the length of the shifts varies, the time actually worked within the period of observation may differ in two coalfields though the number of days worked is the same. The difference between an eight- and a nine-hour day, for instance, might amount to an addition of 12 per cent. to the rate in the nine-hour coalfield. In Great Britain the normal shift up to 1926 was of shorter duration than the German normal shift. The hours of work are fixed in the legislation or in collective agreements according to different principles in different countries. The question is particularly complicated as regards hours of work of underground workers. As a matter of fact» — 29 — the aggregate working time underground comprises not only hours of effective work, but also time needed for travelling from the shaft bottom to the place of work, breaks for meals, as well as descent and ascent x . It is evident that during the travelling time and breaks underground the worker is exposed to a risk of accident. True, this risk is different, if not smaller, than while at work; but since the travelling time and breaks vary considerably not only as between different individuals, but also from mine to mine, it would be impracticable to establish a distinction between the time of actual work at the face or other work, and the rest of the time spent underground. The case is somewhat different as regards the time needed for descending to the shaft bottom and ascending to the surface, i.e. " winding time ". Both during the descent and ascent the worker is exposed to accident risk, and hence the total time needed for them is to be included in the hours of exposure, i.e. the total " hours of attendance ". It is to be observed that the duration of a winding time differs for each pit, a collective winding time being taken to mean the average duration of the descent or ascent of all workers. It depends not only on the depth of the shaft, but also on the number of workers descending or ascending. In order to obtain a correct idea of the length of the time during which the underground workers are actually exposed to risk, it will be necessary to calculate an average time of descent and ascent in all cases where the regulation of the hours of work refers to miners collectively, as is the general practice. An examination of the conditions of work in mines shows that an approximately exact average is obtained if, in calculating the total hours of attendance, account is taken of half of the total time needed for the descent and ascent of all cages, i.e. of one winding time, as each worker must necessarily spend underground the time it takes to lower those workers entering after him and to raise those ascending before him. Now, the legal or contractual hours may or may not include the time of descent and ascent. In Germany, for instance, 1 For a detailed study of the method of determining hours of work in coal mining, and of the actual differences found, see INTERNATIONAL LABOUR OFFICE : Wages and Hours of Work in the Coal-Mining Industry, P a r t I, Chapter II, and P a r t II, Chapter I ; Studies and Reports, Series D, No.18 ; Geneva, 1918. — 31 — was prior to 1926 not taken into account at all, though workers are also exposed to risk during overtime. The same number of full-time workers in two coalfields represented the same normal time worked, but the overtime worked in the two coalfields might differ and the actual time of exposure represented by a full-time worker need not have been the same for the two coalfields. In those countries where practically no overtime is worked, as in Belgium and France, the number of man-days worked will be fairly representative of the time worked except for differences in the winding times. Actual rates per hours worked are calculated in the United States coal-mining accident statistics. Fatal casualties are related to 2,000-hour workers. As the actual number of hours worked by the miners is practically impossible to obtain, much coal being mined by contract at a stated price per ton, rates are calculated which are " based upon the number of hours the mines were in operation and the employees were presumably at liberty to work, rather than the number of hours the men actually worked " K The piece-workers or tonnage men are not obliged to put in a certain number of hours. Other workers engaged in maintenance, haulage and repair work are paid by the day and " their hours conform more closely to the established working day of the mine " 8. The hour-rate calculated in the United States statistics for each State is obtained by multiplying the average number of workers in mines working eight hours by eight, those in mines working nine hours by nine, and so on. The products added up give the number of hours worked per day by all employees in each State, which number is then multiplied by the average number of days worked in the State per year, the product giving the total hours worked per year. This figure, or the average for the United States, is divided by 2,000 in order to arrive at a 2,000-hour worker. This figure was chosen because the working hours per year in the different States ranged a few hours below and above 2,000, the average for the United States being 1,909 (1903-1913). 1 Coal Mine Fatalities in the United Slates, 1920, op. cit., p. 45. Mineral Resources of the United States, 1922, op. cit., P a r t II, 1925, pp. 501 et seq. 2 30 one winding time is included in the eight-hour shift;. In Belgium, the normal shift comprises both winding times according to the law, though the actual practice seems to vary in the different districts. In France, both winding times are counted within the eight-hour shift. In Great Britain, the legal hours (seven prior to 1926, now eight) are counted from the last cage down to the first cage up, that is, both winding times are excluded from calculation of the legal hours of work. The Netherlands mines count the normal shift as beginning with the descent of the first worker and ending with the ascent of the first worker, one winding time being thus included in the shift. In Poland, according to the law the shift includes both descent and ascent ; in Upper Silesia, however, only one winding time is included, according to the collective agreement. In Czechoslovakia, both winding times are comprised in the shift. To arrive at a uniform determination of the average hours of attendance as defined above, i.e. the total time spent underground, plus one winding time, the hours fixed by law or agreement must be modified so as to exclude one winding time when both descent and ascent are comprised, and to add one winding time when only time spent underground is included in these legal hours. As regards surface workers, the determination of the length of the shift does not present difficulties. The hours of exposure are those of " real work " excluding breaks. The regulation adopted in the different countries in this respect is practically uniform. (b) Overtime. — Not only the ordinary shift but also the overtime shifts worked should be taken into account in order to arrive at the time the workers were actually exposed to risk. Overtime may also be worked in supplementary hours added to normal shifts. It is evident that for each overtime shift the same question of the inclusion or exclusion of winding time arises. Overtime shifts may be worked by part of the workers only, in which case the descent or ascent will take less time than for the normal shifts worked by all workers. Such detailed data, however, will not be available. The best method, therefore, will be to reduce overtime shifts to normal shifts and consider these normal shifts as including one winding time each. In the calculation of the German full-time worker, overtime — 32 — Some instances may be given showing the differences of rates due to different methods of calculation. UNITED STATES FATAL CASUALTIES IN 1923x State Kentucky Ohio West Virginia 1 Per 1,000 men employee! Per 1,000 300-day workers Per 1,000 2,000-hour workers 2.96 1.70 2.00 2.11 2.38 2.06 3.70 3.83 3.23 4.37 4.19 4.78 2.91 6.56 2.95 2.69 3.63 3.43 3.96 2.41 5.43 Coal Mine Fatalities in the United States, 1924, op. cit., pp. 40 and 69. Washington, 1925. The variations in the differences between the rates of the several States due to the adoption of different methods of calculation are striking, especially with regard to rates per number of employees and rates per 300-day workers. If the States are arranged according to accident frequency, the order s as follows : Gates per 1,000 men employed Illinois Indiana Pennsylvania Kentucky Ohio Alabama W e s t Virginia Bates per 1,000 300-day workers Rates per 1,000 2,000-hour workers Pennsylvania Illinois Alabama Kentucky Indiana Ohio West Virginia Pennsylvania Illinois Alabama Kentucky Indiana Ohio W e s t Virginia It appears from this comparison that the change is mainly due to the difference in the number of days worked. The following figures afford an explanation, showing the varying; number of days worked in each State in 1923. States Ohio W e s t Virginia Men employed Average days worked per year 30,035 99,714 35,408 60,811 54,555 194,981 117,300 232 158 136 152 150 213 169 Total hours worked per year 60,409,784 126,184,014 38,563,616 75,198,352 65,671,800 333,880,908 159,915,743 — 33 — To conclude, it appears that the best method of calculating the time exposure to risk would be to take into account the aggregate number of hours worked, separately for surface and underground workers. In order to arrive at a uniform computation of hours underground, the normal shift might be so calculated as to include one winding time, as the first workers must necessarily descend or ascend one winding time before the last workers. If agreement were to be reached regarding standard international tables, each country could, for the purpose of filling in these tables, then reduce its overtime shifts to shifts of normal length. The total number of shifts spent underground could then be converted into hours by taking the normal shift to include one winding time. An English normal shift, for instance, would have been seven hours plus one winding time from 1919-1926, the Belgian shift eight hours minus one winding time, and so on. The hours worked in short-time shifts, however, should be calculated separately and include one winding time each. Moreover, supplementary hours should be separately assessed. The table would read as follows : Normal shifts Country or district Short-time shifts Overtime shifts Supple- Total Number mentary hours reduced to Hours • hours worked Number Hours ' Number Hours * Number normal shifts • Each normal shift and each short time shift should include one winding time. For surface workers, regular hours and overtime hours could be calculated as in other industries. Rates could then be calculated per million hours of work or per standard manshifts of a certain length agreed upon internationally. The hours might also be reduced to full-time workers by dividing the hours actually worked by some figure, say, 2,400 hours, assumed to make up a working year 1 . 1 The standard unit is somewhat arbitrary and would not be applicable to all countries, e.g. the United States, where the working year is exceptionally short. — 34 — Rates per hours worked are thus found to be the most appropriate measure of the industrial risk of coal-mining labour. In comparing the risks of underground and surface workers and of the various occupations, rates must of course be calculated for every risk group separately. The classification of the hours of exposure by risk groups must evidently conform to the classification of the casualties. For underground and surface risk, casualties are most suitably classified by the premises. The exposure consequently should relate to the hours worked underground by all workers, on the one hand, and to the hours worked above ground, on the other hand 1 . For the measurement of occupational risk, however, casualties are classified according to the ordinary occupation of the worker injured — the exposure should, therefore, in this case, include all hours performed by the workers of the occupation concerned, irrespective of whether they were spent underground or on the surface. All the rates so far discussed refer to labour exposure properly so called. Most countries, however, publish in addition rates in which the number of casualties is related to the tonnage of coal raised. Tonnage Rates and Rates per Aggregate Wages Paid Tonnage rates are obviously not a suitable measure of the accident risk of coal miners and are not meant as such. The risk can only attach to the work and is only related to the tonnage in so far as the amount of labour engaged in raising a ton of coal represents a certain risk. In the United States and Great Britain, for instance, fatal casualty rates were as follows in 1925 : Country United States. Great Britain. Production (short tons) Men employed Kate per 1,000 workers Rate per 1,000,000 tons 581,869,890 281,016,205 480,227 1,117,828 4.65 1.02 3.84 4.04 The rate in the United States is the rate per 1,000 300-day workers. As the time worked by the mines in the United 1 As with casualties, there may be a difference in the relation of underground exposure to surface exposure according tò whether the unit of exposure is the underground or surface worker, or the day or hour worked underground. — 35 — States is very irregular, the number of 300-day workers probably corresponds more closely to the British calculation of average number of workers employed than the average number of workers. The average number of days worked in the year of observation in Great Britain was 252, while in the United States it was 192. Nevertheless, the higher rate in the United States is partly accounted for by the different measure of exposure. The production of short tons per 300-day worker was 1,212 short tons in the United States ; the risk per 1,000 workers was 4.65 ; per million tons it was therefore 4.65 divided by 1.212, i.e. 3.84. The output per man in Great Britain was 251 short tons only, the rate per million tons is therefore 4.04, i.e. higher than the American rate, whilst the rate per labour unit was very much lower. If therefore the personal risk in the United States is greater, the output per man is so much higher in the United States that the accident rate per unit of coal raised is lower than in Great Britain. It might however be suggested that accidents may be related to some units representing the cost of production in order to compare the costs of accidents in different coalfields. The tonnage rate might therefore serve to compare the incidence of accidents per unit of production. A higher tonnage rate in one of two coalfields compared would then suggest that the accident compensation to be paid for each ton of coal raised was a higher proportion of total costs in the former coalfield. The output, however, is not a direct measure of the relative cost of production, which depends on the wages paid and many other factors besides, nor is the number or severity of accidents a direct measure of the costs of compensation. The tonnage rate, however, is of some economic interest as it shows the loss of life and working capacity connected with the raising of coal in different countries. A lower tonnage rate in one of two countries compared shows that the same output does not involve so much loss of working capacity as in the other country. Even if the risk of the worker is therefore higher in that country the loss of working capacity paid for each ton raised is lower. Tonnage rates are calculated in the United States, Great Britain, Germany, Canada, India, and other countries. The tonnage to which the number and importance of accidents is related can be either the coal raised including dirt, etc., or the saleable coal, which includes the coal distributed to the — 36 — miners and the coal consumed by the mine. Finally, commercially disposable coal is defined as the coal after deduction of the miners' coal and mine consumption. The measure usually employed in calculating tonnage rates is the saleable coal. This is the most appropriate measure as it represents the coal which has value and realises an income in one form or another. Another measure of " exposure to risk " is the aggregate wages paid. Such rates are published in the report of the Oregon Industrial Accident Commission. The number of accidents and the days lost per 100,000 dollars of the pay-roll are calculated. Wages paid, however, differ in various countries and coalfields and at different dates, and these differences may have no relation whatever to the differences in time of exposure or in the number of workers, but may be due to differences in the rates of wages. The wages bill will change with the changing of the time worked or the number of workers employed, but also with a change in the rates of wages. The wages bill would also change in consequence of changes in the proportions of workers in the different occupations. The rate of accidents per aggregate wages paid does therefore not measure the risk of accident of the worker. It might, however, serve as an indirect measure of the accident cost of production, indicating the accident incidence compared with the wages cost of production. In order to measure the cost of accidents in relation to the costs of production, compensation cost rates would be more appropriate. To conclude, the rates per manhours worked have been found to be the most suitable measure of the personal risk, whilst tonnage rates may compare the accident risk connected with the raising of a certain amount of coal. CHAPTER IV CLASSIFICATION BY CAUSES A classification by causes is, in fact, an attempt to trace the relation between the risk and the circumstances contributing to bring about the casualty. The term " cause " is, however, a somewhat vague notion ; strictly speaking it would refer to the last preceding event resulting in injury, but in every case there is a chain of " causes " leading finally up to the casualty. In the case of a distinct disturbing material event, such as an explosion, an eruption of water, a fall of ground, the causal relation is clearly established. There still remains, however, the question as to the cause of this event, which may be of paramount interest for preventive purposes, e.g. a defective safety lamp causing an explosion. Where a personal event only occurs — such as being crushed, or falling or stumbling — this event (the individual circumstances) is usually the result of some hazardous occurrence added to a regular process or act, by which coincidence the danger inherent in the process or act becomes effective. In these cases the event alone is not usually a sufficient indication for preventive purposes, and the work or job during which it happened should be added. The events, whether disturbances of operation or personal events, may be termed primary causes, while the circumstances which cause the disturbances or the personal events may be termed " secondary causes ". In addition, the process or nature of the work in which the casualty happened and the place where it happened may be of interest as indicating contributory circumstances. In the existing classification casualties are grouped by the four criteria : (1) Material event : (a) disturbing event, (i>) personal event ; (2) Place; — 38 — (3) Process or nature of work ; (4) Secondary circumstances. Every classification by causes is, of course, based on a distinction between underground and surface casualties. These are always classified separately as the nature of the underground and surface risks is entirely different. The classification of underground casualties is based wherever possible on the disturbing event : falls of ground, explosion of gas or coal dust, mine fires, escape of water are main groups to be found in all the more detailed classifications. Where it is impossible to state such an event, we frequently find the cases grouped by the nature of the work or the process : " explosions " means casualties occurring during or by blasting and shooting operations, or while handling explosives; the group " machinery " or " by mining tools and machinery " refers to casualties caused by a mechanical process. Special importance attaches to haulage casualties. In most countries these casualties are grouped by the process of haulage, but the local circumstances play a most important part. Haulage accidents may happen on roads and gangways, on inclines and slopes or in shafts, and their character and severity depend to a considerable extent on the part of the mine where they occur ; for most other casualties the place, though of importance, does not indicate a different kind of risk. Accordingly some countries group their haulage casualties by place of occurrence. A second problem arising in this connection is the special risk connected with the shaft. With regard to the shaft, there are different conceptions. The shaft may be regarded (a) as a separate kind of premises, like underground and surface ; or simply (ft) as a place underground where an accident may happen. Finally, (c) accidents in the shaft may be considered from the point of view of the process for which the shaft serves, the winding machinery and haulage by or during which the accidents happen serving as criterion of the classification. A review may now be made of the classifications of the United States, Prussia, and Great Britain, which may be regarded as representative of the classifications in force. Special reference is made to the methods of dealing with shafts. — 39 — 1. In the classification of the United States, the shafts have been considered as separate parts of the mine. The group " shaft accidents " , therefore, is added as a third main group to the two groups " underground " and " surface " accidents and is similary subdivided by "causes". Haulage casualties occurring elsewhere below ground are considered as a causal group of " underground accidents ". The classification reads as follows : Underground (i) Falls of roof (coal, rock, etc.) : (a) at working face, (ft) in room or chamber, (c) on road, entry, or gangway, (d) on slope. (ii) Falls of face or pillar coal : (a) at working face, (ft) on road, entry, or gangway. (iii) Mine cars and locomotives : (a) switching and spragging, (ft) coupling cars, (c) falling from trips, (d) run over by car or motor, (<?) caught between car and rib, (/) caught between car and roof while riding, (g) runaway car or trip, (ft) miscellaneous. (iv) Explosions of gas or coal dust : (a) due to open light, (ft) due to defective safety lamps, (c) due to electric arc, (d) due to shot, (e) due to explosions of powder, (/) miscellaneous. (v) Use of explosives : (o) transportation, (ft) charging, (c) suffocation, (d) drilling into old holes, (e) striking in loose rock or coal, (/) thawing, (g) caps, detonators, etc. (h) unguarded shots, (/) returned too soon, (/') premature shot, (k) sparks from match, lamp, or candle, (/) delayed blast, (m) shot breaking through rib or pillar, (n) miscellaneous. • (vi) Suffocation from mine gases. (vii) Electricity : (a) direct contact with trolly wire, (ft) bar or tool striking trolly wire, (c) contact with mining machine, (d) contact with machine feed wire, (e) contact with haulage motor, (/) miscellaneous. — 40 — (via) (ix) (x) (xi) (i) (ii) (iii) (iv) Animals. Mining machines (other than vii, e). Mine fires (burned, suffocated, etc.). Other causes : (a) fall of person, (&) machinery (other than ix), (c) rush of coal or gob, (d) falling timber, (e) suftocation in chutes, (/) hand tools, axes, bars, etc. (g) nails, splinters, etc. (ft) miscellaneous. Shaft Falling down shafts or slopes. Objects fallings down shafts or slopes. Cage, skip, or bucket : (a) runaway, (b) riding with rock or coal, (c) riding with timber or tools, (d) struck by, (e) miscellaneous. Other causes : (a) overwinding, (ft) breaking of cables, (c) miscellaneous. Surface (i) (ii) (iii) (iv) (v) (vi) Mine cars and mine locomotives. Electricity. Machinery. Boiler explosions or bursting steam pipes. Railway cars and locomotives. Other causes : (a) explosives, (ft) fall of persons, (c) falling objects (derricks, booms, etc.). (d) suffocation in chute, bin, or culm, (e) falls or slides of rock or coal, (/) steam shovels, (g) hand tools, (ft) miscellaneous. In this classification, the main groups of underground casualties refer to the disturbing event (falls of roof, etc., explosions, mine fires, mine gases), where such an event occurred, or to the process, e.g. processes connected with electricity, machinery, animals, mine cars, explosives. The subdivisions complete the main groups by being based either on the place or local circumstances (at working face, on road, etc.) ; the personal event (falling from trips, caught — 41 — between car and rib, run over, falls, etc.) ; the nature of the work (switching, coupling, drilling, etc.) ; or, on the other hand, a secondary cause of importance for preventive purposes, which is frequently an act or a failure to act (open light, sparks from match, unguarded shots, etc.). In grouping shaft casualties, similar criteria are applied as in the case of underground and surface accidents. Surface casualties are mainly grouped by the nature of the work. 2. In the Prussian classification, more stress is laid on the place where the casualty occurs. Thus, the shaft is put on an equal footing with slopes, level gangways, etc., and the grouping of casualties according to these places is predominant. Also, haulage and winding casualties are split up into three local groups : " main shafts ", small blind shafts and slopes ", and " level gangways " ; which groups, however, also contain certain casualties not due to haulage. The subdivisions usually indicate the process during or by which the accident happened. The other main groups take the event as criterion, but are to a large extent subdivided so as to indicate the place where the accident occurred. (i) (ii) (iii) (iv) (v) Underground Falls of stones : (a) in main shafts, (b) in small blind shafts and slopes, (c) in level gangways, (d) at working face, By mining tools and mining machinery. In main shafts : (a) whilst ascending or descending (on ladders), (b) on Fahrkunst (moving poles), (c) haulage of persons by machinery (permitted), (d) haulage of persons by machinery (not permitted), (e) at winding service, (/) working in or at the shaft, (g) other causes. In small blind shafts and slopes : (a) falls of persons, (b) haulage of persons by machinery (permitted), (c) haulage of persons by machinery (not permitted), (d) by winding gear, (e) other causes. In level gangways : (a) whilst passing, (b) mechanical transport of persons (permitted), (c) mechanical transport of persons (not permitted), (a) haulage by machinery, (/) horse haulage, (g) hand haulage, (h) other causes. — 42 — (vi) (vii) (viii) (ix) (x) At the working face. By explosives : (a) during storage or transport, (b) whilst shooting or blasting, (c) by explosive gases. By gases and coal dust : (a) explosions of fire-damp and coal-dust, (b) suffocation by mine gases, (c) by other natural gases. By mine fires : (a) fire in seams, (b) fire in shaft or seat, (c) fire in other parts of the mine. Other causes : (a) eruptions of water, (b) machinery, (c) other causes. Surface (i) At the pit-mouth (hang-bank), including the shaft frame and winding machinery (An der Hängebank einschl. Schachtgerüst und Fördermaschine). (ii) Whilst preparing coal. (ih) In briquette, coke or charcoal works, etc. (iv) Whilst hauling and loading : (a) by railway cars and locomotives, (i>) otherwise. (v) At boiler : (a) by explosions, (b) otherwise. (vi) By generating or transmitting power : (a) by electrical current, (b) by compressed air, (c) otherwise. (vii) Other causes. 3. In the British classification shaft accidents are regarded from the point of view of the process going on in the shaft, i.e. the haulage and winding by machinery, etc. ; and " shaft accidents " consequently constitute a main group of causes on the same lines as " falls of ground " , " explosions of gases " , and " underground haulage " accidents. The method is not always quite consistent in as far as certain shaft accidents not connected with haulage and winding are included in the group, the shaft being in these instances considered as a place. (i) (ii) Underground Explosions of fire-damp or coal-dust. Falls of ground : a) at the working face, b) on roads while repairing or enlarging, (c) on roads while otherwise working or passing, (d) in shafts. — 43 — (iii) Shaft accidents : (a) overwinding, (b) ropes or chains breaking, (c) whilst descending or ascending by machinery, (d) falling into shaft from surface, (e) falling from part way down, (/) things falling into shaft from surface, (g) things falling from part way down, (ft) other shaft accidents. (iv) Underground haulage accidents : (a) ropes or chains breaking, (i>) run over or crushed by trams or tubs : (¡) mechanical haulage, (if) horse haulage, (iii) hand haulage, (iv) runaway trams or tubs, (c) other haulage accidents. (v) Miscellaneous underground : (a) by explosives, (b) suffocation by natural gases, (c) by underground fires, (d) irruptions of water, (e) electricity, (/) by machinery, (g) other accidents. Surface (i) (ii) (iii) (iv) (v) By machinery. Boiler explosions. On railways, sidings or tramways : (a) while engaged in moving wagons, (¿>) while engaged in coupling or uncoupling wagons, (c) run over while passing along or across railways or tramways, (d) crushed between wagons and structures, (e) in other ways. Electricity. Other accidents. The fact that " falls of ground " in shafts are classified under " falls of ground " and not under " shaft accidents " confirms the view that the group " shaft " is meant to indicate not mainly the place, but the kind of work or the process during which the accident occurred. It is probably intended that haulage accidents should contain accidents due to haulage and not accidents occurring on haulage roads, and the group therefore is meant in a causal and not in a local sense. Investigations have shown that some confusion evidently arises from the combined heading " cause or place of accident " . Some people (classifying by place) regard a haulage accident as an accident which happens on a haulage road, while others (classifying by — 44 — cause) regard an accident which is not due to any haulage operation as not being a haulage accident, even though it occurs on a haulage road. For haulage accidents no further indication of the place is given. The two subdivisions, " ropes and chains breaking " and " run over or crushed by trams or tubs ", are not mutually quite exclusive x ; the one indicates the disturbing event, whilst the other takes the personal event, i.e. the individual circumstances, as criterion. " Ropes and chains breaking " therefore may also cause someone to be run over or crushed by tubs and trams. The second group, moreover, is subdivided by the nature of the process. This subdivision of haulage accidents might be considered unsatisfactory as the group " other haulage accidents " contains approximately up to 50 per cent, of all non-fatal haulage accidents. In the Belgian classification the heading " shaft accidents " does not coincide with the British group " shaft accidents ", as blind shafts and working shafts (cheminées d'exploitation) are included, and not only winding shafts, though the nature of their risks is quite different from that of main shafts. The distinction between the three is, however, made in the detailed tables 2. On the other hand, haulage casualties other than in these shafts are subdivided by the place where the casualty happens, casualties on level roads and on inclined roads being distinguished. These groups are then subdivided according to the nature of the process. As to casualties other than those due to haulage and shaft accidents, both the British and the Belgian classifications attempt to indicate to some extent the disturbing event or the secondary cause — which may be a defect or an act or failure to act — as well as the process during or in consequence of which the casualty happened. Occasionally the place of the casualty is given. The Belgian classification goes as far as to include a fourth subdivision for certain kinds of casualties and gives a detailed indication of the combined causes. Casualties due to coal gas, for instance, are subdivided in the following way: normal escape : (1) explosion and (2) suffocation ; (1) is 1 It is noted in the statistics that the group " run over or crushed by trams or tubs " does not include accidents primarily due to ropes and chains breaking. 2 Falls of ground occurring in these shafts are given in the shaft groups and not under the heading " falls of ground ". — 45 — subdivided (a) by fuse, (b) by lamps, etc., and (c) by divers causes ; group (b) is further subdivided by " open lamps " and " defects " ; sudden irruptions : (1) explosions, (2) suffocation, struck by coal or stone, etc. The differences in the classifications are thus considerable, but agreement should not be impossible. A tentative classification might be proposed on the following lines. As shaft accidents in almost all countries are considered as a special group and the risk in the shaft is of a particular character, distinguishing it from the risk on the haulage roads, the shaft could, for the purpose of grouping the casualties, be considered as a separate part of the underground premises. The shaft in this sense should include only the main shafts, that is, winding shafts and not blind shafts or slopes serving as entrances to the mine or any other underground passage 1. The shaft would therefore be the vertical entrance to the mine through which coal and persons are wound. It does not seem advisable, however, to classify shaft casualties quite apart from underground casualties and on the same footing, because shaft casualties happen to all kinds of underground workers descending or ascending the shaft or otherwise occupied near the shaft. In calculating rates for shaft casualties, therefore, all work performed underground is generally taken as exposure. The main groups of casualties could be based on an indication of the disturbing event or the process, while the subdivisions could specify the further circumstances attending the casualty by applying any of the other criteria most appropriate for the purpose : either the individual circumstances (personal event), the place, the nature of the work, or a secondary cause of importance for preventive purposes. It is of great importance to render the groups mutually exclusive. Groups based on the personal event or individual circumstances should not overlap with groups based on the disturbing event, which may easily happen, as every disturbing event also involves personal events. The following tentative classification by causes might therefore be proposed : 1 The question whether very steep slopes serving as entrances to the mine, and therefore subject to conditions similar to shafts, should be reckoned as shafts might be left to the discretion of the Governments. — 46 — I. Shaft Accidents Falls of ground, etc. , Falls of objects. Falling down shaft : (a) while working at shaft, (b) from ladders, or while othsrwise ascending or descending not by machinery, (c) otherwise. (iv) Winding accidents : (a) chains or ropes breaking, (i>) overwinding, (c) other accidents whilst descending or ascending by machi-1 nery. (d) other winding accidents. (v) Struck by cage, bucket, etc. (vi) Other shaft accidents. II. Other Underground Accidents (i) Falls of ground (coal, stone, etc.) : (a) at the working face, (b) on roads, slopes, gangways, etc. (ii) Explosions of gas and coal dust : (a) fire-damp and coal dust, ' (b) other gases. . (iii) Suffocation by mine gases. (iv) Shooting and blasting accidents (" explosives ") : ' (a) premature shot, (b) delayed explosion, (c) explosions during transport, handling or charging 1 : (i) from sparks of matches, lamps, etc., (ii) other causes, (d) explosions of previously unexploded remnants, (e) stones projected by shots, (/) other causes. (v) Mine fires. (vi) Underground haulage : (a) run over, crushed or otherwise injured by trams, tubs, etc.; (i) mechanical haulage, (ii) horse haulage, (iii) hand haulage, (vi) other haulage, (&) injured by animals, (c) falling, stumbling or slipping, during haulage operations, (d) other haulage accidents. (vii) Mining machinery 2 and tools. (via) Electricity. (ix) Irruptions of water. (x) Miscellaneous : . (a) falls of persons, (b) rushes of coal, gob, etc., (c) struck by objects or striking against objects, (d) machinery 3, (e) other accidents. 1 Exclusive of premature shot, i.e. of the intended explosion happening too. soon. : • 3 Exclusive of accidents due to electricity. .;• • * Exclusive1 of accidents due to mining machinery. •' .. (i) (ii) (iii) — 47 — (i) (ii) (iii) (iv) (v) (vi) (vii) (viii) III. Surface Surface haulage and transport accidents : (a) run over, crushed or otherwise injured by trams, tubs, etc., (b) idem, by railway cars or engines, etc., (c) otherwise. Machinery. Boiler explosions, etc. Electricity. Irruptions of water. Explosives. Falls of persons. Other surface accidents. IV. Accidents to Clerical Workers * * * In calculating rates by causes, casualties of each group underground should be related to the hours worked underground by all workers or by the ¡group concerned * ; the same method should be applied in the case of surface casualties. The classification of aricidents by causes according to underground, shaft and surface accidents, however, does, not necessarily indicate to which group of workers the accident happened. Haulage accidents, for instance, may happen to workers engaged in haulage operations or to hewers or other workers passing by the roads to the working face. Thus, the British statistics of 1924 show that, of 262 fatal haulage accidents, 165 occurred among workers engaged in haulage operations, 44 to workers :walking to and from their work, and 53 in other ways. Falls of ground are usually given separately for the working face and for the roads, but falls of ground on a road may also hit a worker walking to or from the working face. Suffocation by mine gases or injuries by explosions may happen to different classes of workers. It would therefore be of great interest to group accidents by causes also according to the occupation to which the injured worker belongs. This classification would .differ from that according to underground and surface workings in as far as surface workers may have accidents underground and vice versa ; the groups " accidents of underground occupations " and of " surface occupations " would therefore not entirely correspond to the two groups " underground " and " surface " accidents. By classifying the casualties of each occupation by causes, each of the two great groups would have to contain a heading, " surface " 1 See Chapter V, pp. 58 et seq. 48 accidents " and " underground accidents " respectively. The casualties of each occupation by each cause would then be related to all hours worked by the members of that occupation, whether performed underground or on the surface. The following classification of casualties by occupations and by causes might complete the classification by parts of the mine and by causes, and the statistics would indicate which groups of workers are mainly exposed to certain kinds of accidents. UNDERGROUND OCCUPATIONS Hewers Haulage Repair Foremen, Other All underand as- men deputies, workers ground sistants workers etc. workers Causes A. Underground : (2) Falls of ground (3) Explosions of gas and (4) Suffocation by mine gases (5) Shooting and blasting . . (7) Underground h a u l a g e . . . (8) Mining machinery a n d (9) Electricity •3.5P o» A. Surface : Surface haulage and Boiler explosions.. Electricity Irruptions of water Falls of persons . . . Other surface acciB. Underground i li Other haulage men, etc. fe>§ SE -•? OCCUPATIONS Coal pickers, etc. SURFACE ss GO 1I 4 CHAPTER V AN INTERNATIONAL COMPARISON OF COAL-MINING ACCIDENTS GENERAL RATES In view of the fact that the statistics of coal-mining accidents in the several coal-produçing countries are fairly detailed, an international comparison would seem feasible. Analysis, however, has shown that such a comparison is greatly hampered by differences in the methods of compilation and calculation. A summary of the reasons which make comparison almost impossible is given below. It might seem at first sight that fatal casualty rates would be readily comparable, as death is a unique and indisputable event. Unfortunately, however, " death by accident " is not so ; a fatal casualty is liable to different interpretations. Death by accident is not always instantaneous ; it will therefore depend on the legislative or other prescriptions to what extent deaths occurring after the accident are still attributed to the latter. In many cases the length of time elapsing between accident and death is taken as criterion. It has been seen that this question is solved in different ways in the special statistics of the several countries, the maximum period which may elapse between the occurrence of the accident and death ensuing varying between thirty days in Belgium and a year and a day in Great Britain. As no data are available for coal-mining statistics for measuring the effects of such differences in methods on the resulting figures, an example was taken from the United States railway accidents statistics for 1924, where a difference of 11.3 per cent. was found between the fatal casualties rate calculated on the basis of casualties resulting in death within twenty-four hours (0.69) and the rate including death up to the date of reporting (0.78). Even fatal casualties being not strictly comparable, it is not surprising to find non-fatal casualties altogether incomparable. On the one hand, the limitation of the period of incapacity which — 50 — must result from the casualty in order to make it reportable in statistics differs from country to country, varying from more than three days of incapacity to work in Great Britain and Prussia, to permanent disability in Belgium. On the other hand, casualties are in, most cases not classified by duration, or, if so classified, no rates are calculated for the duration groups. It is evident that rates calculated on such different bases are really measures of totally different kinds of risks. If, for instance, the total casualty rate (fatal and non-fatal) per thousand " workers present " in Belgium is found to be 1.37 (1925) while the British total casualty rate per thousand 300-shift workers is 191.7. the first rate indicates the risk of being killed or permanently disabled, while the second rate is to be the far greater risk of death or injury entailing incapacity to work for more than three days. Fatal casualties being relatively few in number, rates for non-fatal casualties are not likely to differ much from total casualty rates if the former include all casualties except those of very short duration. The difference is, however, considerable where the non-fatal group only comprises permanently disabling casualties, as in Belgium. Total casualty rates in Great Britain and in Prussia and nonfatal casualty rates in France may be considered fairly comparable. The inclusion of fatal casualties in the former countries will not affect the comparability to any considerable extent. On the other hand, the non-fatal casualties include in both countries casualties disabling for more than three days. The minimum period of incapacity in France is five days. Reference to these rates is made below. Apart from differences in the definition of reportable casualties there are other sources of incomparability. Such a source is to be found in differences in the field covered by the statistics : this may differ in two respects : with regard to the premises covered, and with regard to the categories of workers covered. The condition that the casualty must be caused by mining operations or by industrial operations connected with mining coincides closely with the condition that the casualty must have occurred on the premises of the mine. It can, however, be modified with regard to the premises by the inclusion or exclusion of certain kinds of workers employed on these premises : Prussian and Indian statistics, for instance, include casualties to workers in ancillary works such as coke ovens and briquette works, workers engaged in the dressing of coal, etc., — 51 — which in most other countries are excluded from the statistics. Casualty rates may be affected by these differences, as the risk in ancillary works is likely to be lower than the risk in the mine itself, and the inclusion of casualties in ancillary works among the mine casualties and of the ancillary workers in the exposure to risk of mine workers is likely to lower the total rate. The effect, however, will not be very considerable, as ancillary workers are likely to be few in comparison with the total number of mine workers. In Belgium, for instance, the number of workers employed in ancillary works was approximately 4.9 per cent, of the number of workers in coal mining proper in 1926 (7,919 to 160,197) ». In the Ruhr district (Niederrheinisch-Westfälischer Bezirk), shifts worked by ancillary workers in 1925 accounted for not quite 8 per cent, of shifts worked by all workers in coal mining and ancillary works (7,924,071 to 118,005,809) \ As regards the mine itself, the scope may differ in respect of the inclusion or exclusion of clerical workers ; they are excluded in the American and Belgian statistics, and included in most of the other statistics. The effect, however, of this factor can only be slight, as clerical workers play a very small part in coal mining. British statistics for 1925 show 1.7 per cent, of all persons employed in coal mining to be clerks and salaried persons (19,145 to 1,117,828). As such workers have very few casualties, however, their inclusion in the statistics will nevertheless slightly lower the rate. The differences so far discussed however, are comparatively slight. If the rates themselves are compared, it is found that the main source of in comparability lies in the methods of calculating the rates, that is, in the calculation of the exposure to risk. The data available with regard to the exposure to risk may be summarised as follows : United States of America (1) Actual average number of employees : methods of computation left to the reporting operators. (2) 300-day workers found by multiplying the number of employees by the number of days the mines wound coal and dividing by 300. 1 Statistique des Industries extractives et métallurgiques en Belgique, 1926. Tables II and IV (agglomérés et coke). Brussels, 1927. 2 The relation of the numbers of full-time workers is different, i.e. about 6 per cent. (23,039 to 374,864), as the shifts include overtime shifts, and more overtime was worked in ancillary works. 52 (3) 2,000-hour workers calculated by multiplying the number of workers in each State by the standard hours of work at the mines, and the products by the average number of days the mines wound coal in each State and dividing by 2,000. The number of man-hours worked is also utilised. Great Britain (1) Average number of workers employed during the year ascertained by quarterly statements of thè numbers on the colliery books and clerks and salaried persons. (2) Manshifts worked including short-time shifts and overtime shifts reduced to normal time. Prussia (1) Full-time workers and technical employees combined. Up to 1926 the former figure was found by dividing the sum of normal manshifts worked by the number of all ordinary working days ; since 1926 overtime shifts reduced to normal shifts are included. For the technical employees their average during the year is taken. Belgium (1) Average number of workers present, calculated by dividing the man-days worked on days when the mine wound coal by these days. (2) Man-days worked on all days the mine was open. France Average daily number of workers present found by dividing the man-days worked by the days the mine worked. Netherlands Average number of employees found bjf dividing the total number of workers ascertained fortnightly from the pay-sheets by the periods of payment, including [sick, injured and absent workers. India Daily average number of persons employed, found by dividing the aggregate number of daily attendances of persons permanently and temporarily employed by the number of days worked by the mine. Rates per average number of workers — where only rates per full-day workers are calculated, the latter have been compiled — are shown in the following table for fatal casualties, underground and surface workers combined, for 1920 to 1926. FATAL CASUALTIES PER 1,000 1920-1926 " WORKERS Rates Country 1920 1921 1922 2.90 3.78 0.88 1.13 0.97 2.31 0.98 1.06 2.42 4.20 0.87 0.89 0.94 2.00 1.35 1.18 2.35 4.90 0.95 0.93 0.82 2.05 1.13 1.01 1923 1924 1925 1926 2.98 4.65 1.02 0.92 1.18 2.70 1.07 1.33 3.32 4.50 1.08 0.99 1.03 2.30 1.00 1.06 United States of America : per average number per 300-day workers Great Britain Netherlands 2 . 8 5 3.08 4.39 4.80 1.06 0.98 1.09 1.17 0.86 0.98 1.60 2 . 2 1 1.82 1.23 1.28 0 . 9 3 — 53 — The great differences in these rates are largely due to differences in the number of days worked by the mine. This appears at once from a comparison of the two rates in the United States. The figures for 1922 will be taken as an example for the following deductions. In 1922 the United States mines worked an average of only 144 days, the ratio therefore of the two rates — the rate per average number and per 300-day worker — is 144 to 300, the second rate being 108 per cent, higher l. Some idea of the comparative value of the rates based on the average number of workers will be gained by comparing the number of days worked by the mines in a few countries : AVERAGE NUMBER OF DAYS WORKED BY T H E MINES IN 1 9 2 2 Prussia United States 302 144 Great Britain Belgium 262 295 It appears from the above figures that the rate per thousand workers is relatively much too low in the United States as compared with the other countries because the number of days of exposure of these workers is much smaller than the number of days of exposure of 1,000 workers in Great Britain, Prussia, and Belgium. The British rate is somewhat too low as compared with the Belgian rate. The Prussian rate is relatively too high as compared with the British and American rates ; the 302 days are ordinary working days, each full-time worker being supposed to have worked this number of days 2. It must, however, be kept in mind that the number of " days worked by the mine " are not calculated on quite the same methods in the different countries. Comparison might seem more fruitful if the rates per manshifts worked or per full-time worker were compared. It will appear from these rates how far the rates per thousand workers are modified by differences in the time worked. The following fatal casualty rates are given in the statistics for 1922 : 1 For pre-war coal-mining accidents see also : DEPARTMENT OF THE : Coal-Mining Statistics in the United States and Foreign Countries, by F. W. HORTON, Bulletin No. 69 ; Washington, 1913. In this study a re-calculation on a 300-day basis was made for the United States, France, and Belgium. 2 The figure refers to the Dortmund district. The numbers of working days in the other districts were as follows: Upper Silesia, 297; Lower Silesia, 305 ; Left Rhine, 305 ; Aachen, 302. INTERIOR, BUREAU OF MINES — 54 — • United States Great Britain Belgium Prussia 4.90 0.39 0.305 2.045 per 1,000 300-day workers. „ 100,000 manshifts worked. „ „ 1,000 full-time workers. In order to facilitate comparison, the British and Belgian manshifts (days) are reduced to 300-shift or day workers. The rates, however, are still not completely comparable because the British shifts include overtime and short-time shifts reduced to normal shifts, while the Prussian full-time worker represents normal shifts or days. Moreover, the ordinary working days constituting the divisor of the Prussian rate are not necessarily 300, but 302 in this particular case. The rate for Prussia has been recalculated on the basis of 300 days. The number of man-days in the United States may be overstated, as man-days lost by individual absences on coal-winding days might be included. The rates per thousand 300-day workers would be as follows : United States of America 4.90 Great Britain 1.17 Belgium Prussia 0.915 2.031 In the United States the ratio of the full-time casualty rate to the average worker casualty rate (4.90 : 2.35) corresponds to that of the number of days (300 : 144). In Great Britain, the rate per 1,000 average number employed and the rate per 1,000 300-day workers (0.95 : 1.17) do not bear exactly the same relation to each other as the days worked (262) to the 300 days of the full-time workers. This is probably due to a lower degree of accuracy in the ascertainment of the number of workers employed than in that of the number of manshifts worked. In later reports the two relations almost agree—for instance, in 1925 1 . It would be expected that the Belgian full-time worker rate would be also higher than the average worker rate as the relation of the days is 300 to 295. It is, however, lower. This result is due to the fact that the average number of workers is calculated by taking into account the manshifts on coalwinding days only, whilst the full-time rate refers to all manshifts worked, the full-time rate being lower because the difference between all manshifts worked and^manshifts worked on 1 A slight difference will always remain, as the rates per 100,000 manshifts are calculated for mines exclusive of stratified ironstone mines in Cleveland, Lincolnshire, and Northamptonshire, while these are included in the calculation of the rates per average number of workers employed. — 55 — coal-winding days is greater than the difference between the days K The rate per full-time workers based on the number of days worked, however, takes no account, of differences in the hours worked in the several countries. In considering the hours worked it must be kept in mind that all observations relating to winding time and shifts, etc., do not apply to surface workers, the hours of whom may differ considerably from those of underground workers. The estimates made below are therefore approximate only. If the winding time is taken into account, the regular working hours at the time of observation (1922) were approximately 7 1 / 2 in Great Britain, 7 hours 50 minutes in Belgium, 8 hours in Germany and 8 in most States of the Unites States, where, however, nothing is known about the winding time. In addition to these differences in regular hours, differences in overtime worked would have to be taken into account. If rates per hours worked, including overtime, were calculated, the differences in the rates would be less as between Prussia on the one hand and Great Britain and Belgium on the other, and that for two reasons : (1) The regular hours in 1922 were longer in Prussia than in the two other countries. (2) Considerable overtime is worked in Prussia, which is not included in the number of manshifts as is done in Great Britain. In Belgium overtime is not usually worked. In 1922, the total number of manshifts worked in Prussian cpal mines was 198,078,758, the number of overtime shifts was 14,407,687. Therefore, overtime worked was 7.8 per cent. of the normal and short time worked. Some idea of the change which would be effected in the rates by taking into account the hours worked will be gained if rates per thousand 2,400-hour workers are re-calculated on the assumption of 8 hours plus 7.8 per cent, overtime (=approximately 8 hours 37 minutes) in Prussia, and 7 hours 50 minutes and 7 */2 hours in Belgium and Great Britain respectively. The United States figures are re-calculated from the rate per 1,000 2,000-hour worker (4.04) 2. 1 All manshifts worked : 46,590,260 ; manshifts on coal-winding days : 45,113,970. 2 The calculations could of course be made more accurately by taking the absolute figures as basis ; the above method, however, is intended to bring out more clearly the theoretical aspect. 56 — Fatal casualty rates per 1,000 2,400-hour workers would have been approximately as follows ; the original rates per average number of workers are given in brackets in order to show the ensuing differences : United States 4.848 Belgium 0.934 1 a (2.350) (0.930) Great Britain 1.250 « (0.950) Prussia 1.886 * (2.045) The differences in the rates as between Great Britain, Prussia and Belgium are thus considerably reduced, a part of the excess of the Prussian rate being due to the methods applied in calculating the exposure. Rates ought really to be compared for underground and surface workers separately. The above calculation was not made for underground work alone because the rates per manshift in Great Britain and the rate per 2,000-hour workers in the United States are not calculated for underground and surface workers separately. It will be seen from the comparison of the underground and surface rates in the different countries that the differences as between the countries of the total rates are largely due to differences in the underground rates. This may be due either actually to differences in risks or to the fact that differences in the number of days worked by the mine and in the hours worked are more considerable for underground than surface workers, the comparability of the underground rates being thus more impaired than the comparability of the surface rates by differences of methods 3. A rather interesting confirmation of the approximation which would ensue between the rates of the different countries if similar methods of calculation were applied may be seen in the similarity of the total casualty rates of underground and surface workers in Prussia and Great Britain. In both countries the minimum period of incapacity for a reportable casualty 1 The rate is slightly lower than the 300-day worker rate because the hours taken into account in calculating the rate per 2,000-hour worker are on an average slightly more than eight. „ „ t „ ... . 1.17x2,400 0.915 x 2,400 x 60 D , . 2 Great B n t a m : Belgmm : 7.5x300 470x300 2.031x2,400x60 PrUSSla : 517X300. 3 Moreover, underground rates may be affected by the methods of allotting the units of exposure to underground and surface. In some cases underground exposure relates to all underground workers, in others to all time worked underground. — 57 — is three days. The Prussian rate refers to full-time workers, the calculation of which is based on normal shifts worked exclusive of overtime; the British rate is re-calculated per 1,000 300-shift workers from the rate per 100,000 manshifts, and includes overtime shifts reduced to shifts of normal length. The rates are as follows : Year Prussia Great Britain 1923 1924 1925 1926 65.06 176.00 199.70 215.91 201.9 188.7 191.7 201.9 The Prussian rate is even lower than the British rate in 1923 and 1924, and would be still more so if overtime and length of shifts were taken into account 1 . It must, however, be kept in mind that these rates are mainly determined by the non-fatal risk, which may be actually relatively lower in Prussia than the fatal risk. COMPARISON B Y CAUSES Comparison of the risk of coal miners in different countries is thus found to be rendered almost impossible by the differences in the methods of compiling statistics. It might seem somewhat more fruitful to compare the relative importance of the different causes of accident risk underground (a) by means of percentages, and (6) by comparison of the rates by causes within each country. Even such comparisons, however, are very much hampered by the differences in the classifications by causes. Similar headings do not always represent comparable groups. The summary table " Shaft accidents " in Belgium for instance, for which rates are calculated, covers main shafts and slopes, blind shafts and working shafts, which groups are given separately in the detailed tables only. The British and German groups refer to main shafts and the United States group to shafts and slopes. " Falls of ground " in some countries comprise all falls of ground wherever they occur, while in other statistics falls of ground in the shaft are included among shaft accidents. 1 The very low Prussian rate in 1923 may be due to curtailed production owing to the Ruhr occupation. — 58 — Haulage accidents in Belgium {circulation des ouvriers et transport des produits) cover only casualties on level gangways or slopes, but not those in blind shafts and working shafts, etc. Rates byi Causes As regards rates by causes, the in comparability is still greater than in the case of comparison of general rates owing to differences in classification. Also until recently rates for casualties by causes in the United States statistics were calculated per million hours of labour performed underground and at the surface. They were therefore evidently much lower than those in the other countries where rates by causes underground are related to underground exposure only. FATAL CASUALTIES D U E TO FALLS OF GROUND AND TO HAULAGE P E R WORKERS EMPLOYED OR HOURS P E R F O R M E D UNDERGROUND * United States Tear 1920 1921 1922 1923 1924 1925 1926 G r e a t Britain F a l l s of ground 0.970 0.350 0.55 • 1.077 0358 0.55 1.123 ' ' 0.422 ' 0 5 9 0.60 1.038 0.367 0.62' 1.053 • 0.351 0.61 0.372 1.111 0.68 F a l l s of ground Haulage Belgium Prussia France Haulage F a l l s of ground Haulage F a l l s of ground Haulage F a l l s of ground Haulage 0.24' 0.25 0.23 032 0.27 0.29 0.32 1.05 0.85 0.95 0.70 i:20 1.30 1.39 1.01 0.81 0.S6 0.60 0.90 0.90 0.93 0.50' 0.47 0.38 0.50 Ó.44 0.44 0.46 0,23 0.20 019 0.27 0.23 0.26 0.26 059 0.43 0.46 0.51 0.58 0.49 0.58 0.13 0.1,7 0 08 016 0.28 0.22 0.28 ! 1 The measures of exposure in the different countries are as follows : United States : million man-hours underground : Great Britain : 1,000 workers employed underground ; Prussia : 1,000 full-time workers ; Belgium:. 1,000 workers present; France : 1,000 workers present. But rates by causes per underground exposure, even if the' methods applied were the same, are only very approximate measures of the various kinds of risks : t h e various groups of underground workers are not equally exposed to the different kinds of risk ; haulage casualties, for instance, happen mainly to haulage workers, falls of ground to a large extent to hewers at the working face 1. 1 I n G r e a t B r i t a i n c a s u a l t i e s b y falls of g r o u n d i n 1926 w e r e d i s t r i b u t e d " a s follows : ' • ' Place Killed Injured At working face 820 23,952 On roads while repairing and enlarging . . . . 64 3,240 „ „ „ otherwise working or passing. • 45 2,667 In shafts. 1 20 Total. . 33Ö 29379 ' — 59 — By referring casualties by causes to all work performed underground, differences in the classification of the man-hours according to the parts of the mine where they are worked are not taken into account. These differences are considerable. In Great Britain, for instance, the estimated number of manshifts actually worked at the coal face in 1925 was 108,679,000, as against 108,270,000 elsewhere below ground, or more than 50 per cent, of all manshifts worked underground. In Belgium, the number of man-days worked by hewers and getters in 1925 was 6,496,605, as compared with a total number of man-days worked underground of 33,094,070, or round about 20 per cent. of the total number of man-days worked underground. In Prussia (1924) the ratio of hewers to all underground workers was 15.7 to 69.3 in Upper Silesia, 32.7 to 68.2 in Lower Silesia, 43.2 to 75.6 in Dortmund, 41.5 to 72.9 on the Left Bank of the Lower Rhine, and 39.1 to 70.9 in Aachen. The proportion of hewers (full-time workers) varied therefore between 22.6 percent. in Upper Silesia and 57.1 per cent, in Dortmund \ These proportions are, of course, not strictly comparable. For example, the British figures are based on the manshifts' worked at the coal face and all manshifts worked underground ; the Belgian on man-days worked by hewers and all underground workers ; and the Prussian figures on full-time hewers and full-time underground workers. It is known, moreover,that the term " hewer " has very different meanings in the several countries. If casualties by falls of ground are related to all work performed underground, the rates in countries with a lower proportion of hewers will necessarily work out lower than in countries with a higher proportion of hewers. The same holds good for haulage casualties, the number of which evidently depends on the proportion of haulage men, or rather on the time worked underground in places where haulage casualties may occur 2. 1 Sixth Annual Report of the Secretarti for Mines, 1926, p. 141 ; London, 1927. •—• Statistique des industries extractives et métallurgiques, 1926, Table II; Brussels, 1927.—Zeitschrift für das Berg-, Hütten- und Salinenwesen im Preussischen Staate 1924, p. 206; Berlin, 1924. 2 An instance may nevertheless be given for falls of ground. The Belgian and British rates of fatal casualties caused by falls of ground in 1925 per thousand workers underground were 0.44 and 0.61 respectively, the thousand workers representing the average number of workers present in Belgium and the average number of workers employed — 60 — Haulage casualties may happen to all kinds of underground workers, though mainly, of course, to haulage men. Moreover, the classifications in some countries are based on a local criterion, the haulage roads and blind shafts, etc. (Prussia, etc.) ; in others, e.g. in Great Britain, on the operations of haulage ; in Belgium, haulage casualties in blind shafts (puits intérieurs) and working shafts are given under these headings. . One conclusion only can be drawn from the rate by causes : a fairly great regularity within each country which is not to be found with other causes. Fatal casualty rates by gas and coal dust, for instance, show far more considerable changes from year to year. As regards causes of non-fatal casualties, the definitions of reportable non-fatal casualties are generally too different to allow of any conclusions, except for Great Britain, Prussia and France, where the limits are three or four days respectively. Rates by causes, however, are calculated only in France for non-fatal casualties, and in Prussia for total casualties. As fatal casualties are only a small percentage of all casualties, the inclusion of the former would not greatly impair comparability. It may be noted that contrary to what was found with fatal rates, total casualty rates by falls of ground are lower in Prussia than in France. In 1926, the Prussian rate by falls of ground was 83.67 per 1,000 full-time workers underground and the French rate 170.5 ; the haulage rates were 93.48 and 83.53 respectively. As the limit for reportable casualties is even lower in Prussia than in France, and as the application of similar methods for calculating exposure would rather lower the Prussian rate, it is difficult to offer an explanation. So far as non-fatal casualty rates are available, they too show a certain regularity within the countries for falls of ground and haulage casualties. in Great Britain. If Belgian rates are calculated for fatal casualties by falls of ground for hewers only (dans íes tailles, travaux préparatoires et galeries horizontales ou inclinées en veine, au cours ou à l'occasion du travail d'abattage ou de creusement) per 100,000 man-days of hewers, and British rates for casualties by falls of ground at the working face per 100,000 manshifts worked at the coal face, the Belgian rate works out at 0.58 and the British rate 0.30, the relation being contrary to that of the average worker rates. It must, however, be kept in mind that falls of ground in all kinds of shafts are excluded in the Belgian average worker rate for falls of ground. The British rates per manshifts do not include casualties of workers in ironstone mines, which are included in the rates per average number of workers. — 61 — Percentages by Causes Comparison of percentages by causes, besides being impaired by differences in classification, and though independent of the methods of calculating exposure to risk, does not afford any measure whatever of the actual incidence of risk, as a higher percentage for one cause lowers the percentages for other causes. Thus the high percentage of haulage casualties in Prussia greatly reduces the percentage for falls of ground. Two facts, however, may be derived from the table giving percentages by causes for fatal casualties in 1920, 1921 and 1922. (1) The percentages for falls of ground and the percentages for haulage accidents are highest and very constant within each country. (2) Though percentages for falls of ground and for haulage accidents vary considerably as between the different countries, the two percentages together are fairly uniform, oscillating between 60 and 80 per cent., being usually somewhere near 70. Belgium is found to form an exception, but the lower percentage for haulage casualties and falls of ground is most likely due to the inclusion of many such casualties among shaft accidents, which include all such casualties in main shafts, blind shafts and working shafts. The percentage for shaft casualties in fact amounts to between 22 and 27 per cent, in this country. It will be seen, moreover, from the national tables given in the Appendix that neither rates nor percentages for falls of ground and haulage casualties have declined. Rates for other causes as well as percentages differ widely as between countries and vary most considerably from year to year. Percentages for causes of non-fatal casualties, which have been calculated for Great Britain and France, together with percentages by total casualties in Prussia, so far as available, show much the same features. Percentages for falls of ground in the different countries agree more closely even than those for fatal casualties, and the variations within the countries both for falls of ground and haulage casualties are considerably less than in the case of fatal casualties. COAL M I N I N G : FATAL CASUALTIES U N D E R G R O U N D Percentages by Causes, 1920-1922, in Different Countries Country and year Falls of ground India 1922 United States Great Britain Gases and coal dust Explosives Shaft accidents Miscellaneous All causes Number Per cent. Number Per cent. Number Per cent. Number Per cent. Number Percent. Number Per cent. Number Per cent. 1920 United States 1,132 Great Britain 544 Belgium . . . . 55 505 7S India 71 1921 United States Great Britain Belgium Haulage All causes rate per 1,000 men 54.53 56.37 36.92 35.71 49.37 48.30 408 237 25 487 17 16 19.65 24.56 16.78 34.44 10.76 10.88 182 26 1 14 103 9 6 8.77 2.69 9.39 7.28 5.69 4.08 128 22 7 68 6 8 6.16 2.28 4.70 4.81 3.80 5.44 56 40 39 130 28 28 2.70 4.14 26.17 9.19 17.72 19.05 170 96 9 121 20 18 8.19 9.96 6.04 8.57 12.66 12.25 2,076 965 149 1,414 158 147 100 100 100 100 100 100 4.32 0.97 1.34 2.94 12.00 1.39 1,024 383 53 444 61 150 54.85 56.16 40.46 35.40 42.36 61.73 341 170 23 424 24 32 18.26 24.93 17.56 33.81 16.67 13.17 125 19 18 179 26 15 6.69 2.78 13.74 14.27 18.05 6.17 152 13 1 54 6 18 8.14 1.91 0.76 4.32 4.17 7.41 36 26 35 76 16 18 1.93 3.81 26.72 6.06 11.11 7.41 189 71 1 77 11 10 10.13 10.41 0.76 6.14 7.64 4.11 1,86 V 682 131 1,254 144 243 100 100 100 100 100 100 4.73 0.98 1.16 2.40 10.30 1.39 905 551 39 435 72 105 49.16 55.16 36.79 39.30 54.14 56.15 341 211 20 391 13 14 18.52 21.12 18.87 35.32 9.77 7.49 320 73 9 68 3 21 17.38 7.31 8.49 6.14 2.26 11.23 92 5.00 142 3 QO 108 21.70 7.41 14.28 17.11 7 84 18 8 7.71 10.81 6.60 7.59 13.53 4.28 1,841 999 106 1,107 133 187 100 1.70 41 39 23 82 19 32 2.23 17 5.51 1.07 Î.02 2.42 8.60 2.18 1 8 47 8 7 7.55 4.24 6.02 3.74 Exclusive of suffocation. 100 1ÖÖ 100 100 100 ^ C CD U s* ce CD —H S 5 O O O O O O H r l r l TH lOCO C O O O CD 00 i-i IO O S O O I O OS " * l H CO 0 0 TH T - I CM^tCM oo Tf oo io o o o •Û g £ CD C O O X 00 T t 1 - * O CO CM io e o o THCMCO iqqoo osco od O T H OS CO TH O O T T 00 CO'iO r i CM CM CM T-t - * 0 0 CO eòo-* 0) Pi COCOCO CD COCOCO io io os CO OSCO CM T H T H X! THCM s 3 T H O C M " * • * IO l-( s u O T H C D TH O O C D ±» Cß 3 £ +3 a o « w 0 o o o o o o H r l r l 0) CU 03 o Q Z ¡3 O o o o o T HoH o H mono IO e o o COCMiO THTHCM O TH TH «¿ CD I O I O O a 2 'S ÎJ - * l > IO 0 0 I - I CM CM CO OS OS OS 00 odd odo -*OsO X X O - * l O CS T H - * co 00 co co Ö CD ö l O OS 0 0 r-lLO^f O O O O CD 0- CD 04 COCM-tf CMC33 1 0 Tf 0 0 - * CMTHCO oo oo T - I 00 O O O T H CM CDiOO COTHTf TjlHOS CO CO CM t3 CD ä , Û « z o S ¿2 « I O CO CO OOSCM O T-^Tf correr THCMCM C/l W H •—, X CO r-l < D 03 <U -« U a tfl J <• « U 'cfj O P. X cu rf ed •4 t, o o H < h V¡ <s> 53 ÓÓÒ 00 i O CM CMCMIO CM-ri CM rf CMO oosoo OCM -W 50 _ ^ tí CD 00T»<IO COCOCM THCD TH O O O CD TH O O n o » O T H CO CD odd io eoo ooq odo - * ! D H THCOTH TH T-l C M O T H CM T - I I O IO co CO O TH CM CO if O M lOO^f CMCOlO OCMO CM CM CM C O O X TH TH O O C M T H T H O C O 'O o OH «M CD Ss tí ri O !*3 Z O Z •frj THOÓOÓ THCMOS l O O O 00 i o t > OcOrt< T H OSCO TfOCM a ¡z¡ 00l>co C l H l f O O S O COCO T H co OOCM^IO0t>TJH 0 0 O O COCOCD OS CM CO m ono I O O C O oô do COCOCO -* co d CO-*rf OS CM CO* CMCOCO r-l CD 0 0 COOCM OOSOS - * -* o cooco H O N - * COCO - * 0 0 Ü O t. as OQ scej fa S 3 a u « (S M O T H C M CMCMCM OS OS OS T-l I H T - I T-l TH T-l '»«"T'a «Sa Ì—I THCMCO o - * OS e o o CM CM C O - * IO CMCMCM OS OS OS r l H r l ^-^ WS >—•. uè 2 fe Ü •* -*CMrr N O - * CS a l J-" CO CS w ti o THOSCM THCMCM O T H C M CS +J CO •y c s r j O H T f • * TJI T * CO T - I OS CM CM CM Os Os OS .S'ca'új"' -*coco T i TH i>-*q 0) PH "o IOCMTH IO o C O N O <3 o •u odd IO O S O CM T-l TH ^ CD e« X CM CM CO CO O T H co u CD O Z odo 3 S3 i—l'-Ö OH odd CD s» « CM CO OS C D - * OS CO^CO .Û W H O H tó O oo osco T-l T - l T H ^ '•"î *"¡ cu +3 - j CS CS m O » 3H S ft. The one salient fact which is borne out by the statistics is the great importance of haulage and falls of ground among causes of coal-mining risks ; also there is a considerable constancy of these two kinds of risk within each country. This might appear all the more remarkable as falls of ground are natural phenomena, largely independent of the human factor while haulage casualties depend exclusively on this factor. Some suggestions, though based on very slight data only, may be added to show that this regularity is mainly one of accidents. In France it is found that the percentages for fatal and non-fatal casualties by falls of ground vary inversely : higher fatal percentages correspond to lower non-fatal percentages : Year 1920 1921 1922 Fatal 49.37 42.36 54.14 Non-fatal 34.94 43.23 40.36 Percentages of fatal and non-fatal casualties by haulage casualties also vary more or less inversely both in Great Britain and France : Great Year 1920 1921 1922 Britain Fatal 24.56 24.93 21.12 Non-fatal 27.25 22.30 27.55 France 1920 1921 1922 10.76 16.67 9.77 25.36 19.52 19.72 This negative correlation seems to point to the conclusion that the regularity of percentages for falls of ground is really one of " accidents ", the consequences of which may be fatal or non-fatal, according more or less to hazard. Finally, the great regularity of risks by falls of ground and haulage is confirmed by the British rates per manshifts which since 1922 have been calculated for casualties by falls of ground and haulage casualties. In calculating these rates, casualties by falls of ground have been related to manshifts worked at the coal face, and haulage casualties to those worked elsewhere below ground. These rates come very near to true measures of risk, and they too show a marked constancy throughout the five years of observation. APPENDIX NOTES ON T H E STATISTICS OF COAL-MINING ACCIDENTS IN DIFFERENT COUNTRIES BELGIUM Accidents in coal mines are recorded both in the Annales des Mines de Belgique and in the Statistique des Industries extractives et métallurgiques et des appareils à vapeur en Belgique, published by the Ministry of Industry, Labour and Social Welfare. Scope The statistics cover all coal mines in Belgium, exclusive of ancillary works such as washeries, briquette and cokeworks, and power stations, for which returns are given separately. Office and technical workers are also excluded. The statistics include all accidents, all fatal casualties and non-fatal casualties entailing permanent total or partial incapacity. A fatal casualty is one leading to death within thirty days from the accident. Accidents due to pathological causes are excluded. Data and Sources Accidents are reported, under the Compensation Act of 1903 (24 December), by the head of the undertaking to the mine engineers. The average number of employees for each concession is found by dividing the man-days worked (journées de présence) during the days when coal was wound, as obtained from the payrolls, by the average number of days on which coal was wound. These days are, for each mine, all days on which at least one pit was active. The total number of coal miners is found by adding up the numbers thus obtained for each concession. The average number of man-days worked for the purpose of calculating rates, however, refers to manshifts worked on all days on which the mine was open. Accidents and casualties are classified by provinces. Fatal casualties and non-fatal casualties as well as accidents are shown for underground and surface separately and for underground and surface combined. Accidents and casualties are also classified by causes. The Annales des Mines de Belgique contain, moreover, detailed reports for the several districts and for certain groups of causes. — 66 — Measures of Risk 1. Fatal casualties per 10,000 workers for underground, and underground and surface combined, are calculated. 2. Fatal casualties per 1,000,000 manshifts worked are published in the annual tables (underground, and underground and surface combined). 3. In the text, moreover, fatal and non-fatal casualty rates per 10,000 workers, underground, surface, and underground and surface combined, are calculated ; for instance, in the report of 1924. 4. The number of accidents and the proportion of fatal and nonfatal casualties per accident both for underground accidents separately, and for underground and surface accidents combined, are given in the text. 5. Fatal casualties per 10,000 workers are calculated for main groups of causes, both per 10,000 underground workers, and per 10,000 workers for surface and underground combined. TABLE I. NUMBER OF MAN-DAYS WORKED, AND NUMBER OF FATAL CASUALTIES AND RATES PER 1 , 0 0 0 , 0 0 0 MAN-DAYS IN BELGIUM, 1 9 1 2 , 1913 AND 1920-1926 (UNDERGROUND AND SURFACE) Year 1912 1913 1920 1921 1922 1923 1924 1925 1926 Man-days worked 42,517,868 43,268,158 48,510,230 48,307,370 46,590,260 48,906,910 51,429,270 48,918,080 49,304,580 Fatal casualties Kate Number 145 152 181 146 142 175 202 147 159 3.41 3.51 3.73 3.02 3.05 3.58 3.93 3.00 3.22 TABLIí II. NUMBER OF WORKERS PRESENT, AND NUMBER OF FATAL AND SERIOUS NON-FATAL CASUALTIES AND RATES PER 1 0 , 0 0 0 WORKERS IN BELGIUM, 1 9 1 2 , 1 9 1 3 , AND 1 9 2 0 - 1 9 2 6 (UNDERGROUND AND SURFACE) Year Number of workers present Serious non-fatal casualties Fatal casualties Number Number .Rate Fatal and serious non-fatal casualties Rate Number Rate 9.21 7.56 6.54 5.04 5.22 8.03 6.97 5.55 5.24 222 207 220 188 160 239 248 193 193 21.08 19.56 19.98 16.61 15.47 21.80 20.84 17.56 17.44 2.97 4.30 2.81 2.75 3.85 5.56 4.32 2.38 2.62 32 42 47 29 55 52 60 27 37 7.93 10.62 9.43 5.69 11.14 10.32 11.26 5.35 7.46 254 249 267 217 215 291 308 220 230 17.43 17.13 16.70 13.23 14.07 18.19 17.87 13.72 14.36 Underground 1912 1913 1920 1921 1922 1923 1924 1925 1926 105,324 105,801 110,116 113,191 103,444 109,639 118,981 109,916 110,615 125 127 148 131 106 151 165 132 135 97 80 72 57 54 88 83 61 58 11.87 12.00 13.44 11.57 10.25 13.77 13.87 12.01 12.20 Surface 1912 1913 1920 1921 1922 1923 1924 1925 1926 40,346 39,536 49,828 50,949 49,394 50,364 53,304 50,467 49,582 20 25 33 15 36 24 37 15 24 4.96 6.32 6.62 2.94 7.29 4.76 6.94 2.97 4.84 Underground 1912 1913 1920 1921 1922 1923 1924 1925 1926 145,670 145,337 159,944 164,140 152,838 160,003 172,285 160,383 160,197 145 152 181 146 142 175 202 147 159 9.95 10.46 11.32 8.90 9.29 10.94 11.72 9.17 9.93 12 17 14 14 19 28 23 12 13 and Surface 109 97 86 71 73 116 106 73 71 7.48 6.67 5.38 4.33 4.78 7.25 6.15 4.55 4.43 — 68 — TABLES III. NUMBER OF WORKERS PRESENT, AND NUMBER OF FATAL CASUALTIES AND RATES PER 1 0 , 0 0 0 WORKERS, UNDERGROUND, BY CAUSES, IN BELGIUM, 1920-1926 f a t a l casualties by causes Year 1920 1921 1922 1923 1924 1925 1926 Number of workers present Shaft accidents Coal gases and coal dust Falls of ground Number Rate Number Kate Number Bate 38 35 23 17 28 29 29 3.45 3.09 2.23 1.55 2.35 2.64 2.62 55 53 39 55 52 48 51 4.99 4.68 3.77 5.02 4.37 4.37 4.61 14 18 9 26 44 14 15 1.27 1.59 0.87 2.37 3.69 1.27 1.36 110,116 113,191 103,444 109,639 118,981 109,916 110,615 Fatal casualties by causes (continued) Explosives 1920 1921 1922 1923 1924 1925 1926 Number 7 1 8 8 4 6 2 Haulage Bate Number 0.64 0.09 0.77 0.73 0.34 0.55 0.18 25 23 20 30 27 28 29 Miscellaneous All causes Kate Number Rata Number Rate 2.27 2.03 1.93 2.74 2.27 2.55 2.62 9 1 7 15 10 7 9 0.82 0.09 0.68 1.36 0.85 0.63 0.81 148 131 106 151 165 132 135 13.44 11.57 10.25 13.77 13.87 12.01 12.20 — 69 — CAxXADA Accident statistics for the whole of the coal mines of Canada were not published by the Dominion Government before 1926. In 1926, the Department of Trade and Commerce, Dominion Bureau of Statistics, in its annual report, Coal Statistics for Canada, for the first time published a table showing the fatal and non-fatal casualties in the coal-mining industry, in Canada, by provinces. Casualties are classified as underground and surface and by groups of causes. The classification and the figures for 1926 were as follows : Fatal casualties Underground : Falls of roof or face Mine cars or locomotives Gas and dust explosives Explosives Electricity Timbering Mining and loading Miscellaneous I:Î 678 567 10 21 11 1 12 S) 816 73 2,121. 3.'} 14 Total Surface: Haulage and cars Machinery Miscellaneous , Granii total. y 4 97 55 :¡07 459 77 2,580 , Total Non-fatal casualties CZECHOSLOVAKIA Coal-mining accident statistics are published by the Ministry of Public Works, section V, in the publications Hornicky Vestnik (Mining Bulletin) and Uhli (Coal). Scope The former report covers all mines and ancillary works, the second coal mines (pit coal and lignite) only. Casualties reported are : (1) Fatal casualties, i.e. those resulting in death immediately or subsequently. No limitation is fixed to the period which may elapse between accident and death in order to make the casualty reportable. (2) Non-fatal casualties. These are : (a) " serious " casualties, i.e. casualties serious as regards the nature of the injury, viz. those — 70 — resulting in mutilation of the body or permanent injury to health. The days of incapacity to work are not taken into account, and serious casualties need not even entail any incapacity to work. (b) Casualties serious only as regards the duration of incapacity to work, including those which require a treatment of more than nineteen days. Casualties incapacitating for less than nineteen days have to be notified, but are not included in the statistics. Data and Methods of Computation Fatal and serious non-fatal casualties have to be notified by the management of the mines to the Mining Office of the district (Bassin). These Offices send copies of the reports to the Ministry of Public Works, which compiles the statistics. In Hornicky Vestnik, casualties are classified by causes, underground and surface separately. In a further table casualties are classified by districts (underground and surface). Moreover, casualties (underground and surface) are grouped by the kind of mineral mined, pit-coal mines and lignite mines being given separately, as well as a classification by fault. A classification by the day of the week on which the casualty happened is also given. Fatal and serious casualties in pit-coal and lignite mines are classified separately as " fatal ", " serious by nature of injury ", and " serious by duration of incapacity " (underground and surface). i Measures of Risk 1. Casualties per 10,000 workers : (a) fatal ; (b) serious by nature ; (c) serious by duration (all mines in Bohemia, Silesia and Moravia). 2. Fatal casualties per 10,000,000 hundredweights of mineral extracted (coal mines). 3. Percentages by causes (all mines). TABLE IV ( a ) *. NUMBER OF ESTABLISHMENTS, NUMBER OF WORKERS AND OVERSEERS OUTPUT IN HUNDREDWEIGHTS, NUMBER OF FATAL AND SERIOUS CASUALTIES, AND NUMBEI OF FATAL CASUALTIES PER 1 0 , 0 0 0 WORKERS CZECHOSLOVAKIA, 1919-1925, PIT-COAL AND 10,000,000 HUNDREDWEIGHTS MINES (UNDERGROUND AND Casualties Number of Year establishments 1919 1920 1921 1922 1923 1924 1925 140 154 175 151 147 143 193 Number Of workers and overseers 68,794 78,688 80,128 75.210 71,991 72,677 64,890 1 Output in hundredweights Underground Fatal Surface Seri- Serious ous by Fa- Seriby by tal nana- durature tion ture 102,542,332 153 3 9 2 113,749,539 66 4 4 5 120,232,092 95 4 0 5 104,649,897 39 371 123,472,513 55 4 2 2 151,789,419 90 573 127,544,560 70 4 5 5 1,613 16 119 91 1,375 14 2,028 4 125 4 103 2,596 8 104 2,575 7 109 3,085 86 4 2,710 I> SURFACE) Fatti easealtin Total SeriSeri- Serious by Fa- ous by ous by dura- tal na- duration tion ture 267 169 511 265 80 536 99 530 351 4 5 5 43 4 7 5 63 526 370 97 6 8 2 415 74 541 371 per per 10,000 10,000,0« work- hundred e r s weights 1,880 2 4 . 5 1,640 10.1 2,375 12.3 5.7 2,851 8.7 2,945 3,500 13.3 3,081 11.4 Supplied by the Ministry of Public Works to the International Labour Office. 16.4 7.0 8.2 4.1 5.1 6.4 5.8 — 71 — TABLE IV (b) 1 • NUMBER OF ESTABLISHMENTS, NUMBER OF WORKERS AND OVERSEERS, OUTPUT IN HUNDREDWEIGHTS, AND NUMBER OF FATAL AND SERIOUS CASUALTIES AND NUMBER OF FATAL CASUALTIES PER 1 0 , 0 0 0 WORKERS AND 1 0 , 0 0 0 , 0 0 0 HUNDRED- WEIGHTS IN CZECHOSLOVAKIA, 1 9 1 9 - 1 9 2 5 , LIGNITE MINES (UNDERGROUND AND SURFACE) Casualties Kumbir of Numbs? of Year establishments 1919 1920 1921 1922 1923 1924 1925 229 251 266 243 231 217 151 workers and overseers Output in hundredweights 46,121 56,480 56,885 51,449 46,020 45,611 43,349 173,229,608 199,566,103 213,351,276 191,742,961 162,655,298 204,596,902 187,891,420 Underground Fatal easuittin Surface Total per Seri- SeriSeri- SeriSeri- Seri- 10,000 ous Fa- ous ous Fa- ous ous work- 10,000,000 Fa- ons workers by by by by by tal natal nadura- tal na- dura- e r s durature tion ture tion ture tion 37 44 36 35 37 41 30 180 181 181 187 206 266 212 581 421 655 1276 1389 1486 1493 20 23 26 16 7 14 9 123 131 144 129 106 109 79 330 287 386 474 544 520 515 57 67 62 51 44 55 39 303 312 325 316 312 375 291 911 708 1,041 1,750 1,933 2,006 2,008 12.3 11.8 10.8 9.5 9.5 12.0 9.4 ' Supplied by the Ministry of Public Works to the International Labour Office. FRANCE The Mines Department of the Ministry of Public Works (Ministère des Travaux Publics, Direction des Mines) publishes a report, Statistique de VIndustrie Minérale et des Appareils à vapeur, en France et en Algérie. Scope Statistics relate to practically all coal mines, including lignite mines, but ancillary works are not included in the main statistics. All fatal casualties and all casualties disabling the worker for more than four days are registered. Casualties the consequences of which are unknown are regarded as non-fatal. Data and Methods of Computation Casualties are classified as fatal casualties, casualties entailing permanent incapacity, casualties resulting in temporary incapacity, and casualties with issue unknown. These casualties, as well as accidents, are subdivided as underground and surface accidents and casualties. The former (underground) are subdivided by summary groups of causes. These various groups are given separately for the different districts, but for lignite and pit-coal mines together. The number of workers employed is found by dividing the number of man-days worked (journées de présence) by the number of coalwinding days. Measures of Risk 1. Fatal casualties, underground and surface combined, per 10,000 workers employed underground and surface. 2. Casualties underground per 10,000 workers underground : (a) fatal, and (b) non-fatal. 3.2 3.4 2.9 2.6 2.7 2.7 2.0 — 72 — 3. Similar rates are calculated by causes underground for fatal and non-fatal casualties. 4. Casualties on the surface per 10,000 surface workers : (a) fatal and (¿>) non-fatal. TABLE V. NUMBER OF WORKERS PRESENT, NUMBER OF FATAL AND NONFATAL CASUALTIES, AND FATAL CASUALTY RATE PER 1 0 , 0 0 0 WORKERS, IN FRANCE, 1 9 1 2 , 1 9 1 3 , AND 1 9 2 3 - 1 9 2 6 (UNDERGROUND AND SURFACE) Year 1912 1913 1923 1924 1925 1926 Casualties Number of workers Underground Total Surface present (underground and surface) Fatal Non-fatal Fatal Non-fatal Fatal Non-fatal 202,365 203,306 263,995 286,562 299,133 306,878 44,954 44,716 61,889 74,504 81,895 91,128 268 184 186 234 277 266 34 34 42 47 77 53 6,553 6,813 10,564 11,710 11,621 11,892 Fatal casualties (underground and surface ) per 10,000 workers 14.9 1 10.7» 8.6 9.8 11.8 10.3 302 51,507 218 51,529 228 72,453 281 86,214 354 93,516 319 103,020 1 Figures calculated by the International Labour Office. * Figures recalculated by the International Labour Office. TABLE VI. NUMBER OF WORKERS PRESENT, NUMBEF! OF FATAL CASUALTIES, AND RATES PER 1 0 , 0 0 0 WORKERS, UNDERGROUND» BY CAUSES, IN FRANCE, 1 9 1 2 - 1 9 2 6 Year 1912 1913 1923 1924 1925 1926 145,573 146,568 183,463 203,444 214,831 222,983 Year Chains or ropes, etc., breaking Explosion of fire-damp, other gases, or coal dust Falls of ground Number of workers present Explosives Number Rate Number Rate Number Rate 84 101 93 120 107 131 5.8 6.8 5.1 5.9 1 á.C1 5.8 106 14 2 3 1 2 7.3 1.0 0.1 0.1 0.0 0.1 7 3 5 12 12 12 0.5 0.2 0.3 0.6 1 0.6» 0.5 Shaft accidents Number Rate 1912 1913 1923 1924 1925 1926 4 6 6 1 61 5 0.3 0.4 0.3 0.0 2.8 0.2 1 Haulage Others Sumber 17 14 31 25 29 29 . Other causes AH causes Kate Number Sate Number Hate Number Rate 1.2 1.0 1.7 1.2 1.3 1.3 28 41 30 57 48 63 1.9 2.8 1.6 2.8 2.2 2.8 22 5 17 16 19 24 1.5 0.3 0.9 0.8 1 0.9 1 1.0 Figures recalculated by the International Labour Office. 268 184 186 234 277 266 18.5 12.5 10.1 11.5 12.8 11.9 — 73 — GERMANY As the great majority of the German coalfields are in Prussia, the Prussian accident statistics published in the annual report of the '•Grubensicherheitsamt" (Safety of Mines Department), Berg-, Hüttenund Salinenwesen im Preussischen Staate, published by the Ministry of Trade and Industry, are the main sources of information. Further, provisional accident statistics are published quarterly in the report, Deutscher Reichs-, und Preussischer Sfaatsanzeiger by the "Grubensicherheitsamt". Scope Up to 1923 only fatal casualties were reported in the statistics. If the fatal issue of a casualty is reported after termination of the report no rectification is made. Since 1923 all non-fatal casualties disabling the worker totally or partially for more than three days are included. Casualties reported are those due to mining operations, and happening in or about the mine during the shift to the mine workers (Belegschaft). Thus the main statistics do not include : 1. Casualties due to physical or mental defects of the injured, for instance death due to strokes. 2. Casualties due to some event not connected with mining operations, as, for instance, lightning, suicide, and murder. 3. Casualties happening on the way to and from the shift. 4. Casualties among persons other than mine workers, for whom fatal casualties are given in a separate table. Casualties at ancillary undertakings, as coke ovens and briquette works, are included provided these works are under the control of the mining authorities. They are, however, given as a separate group among the causes of surface accidents. Data and Methods of Computation Casualties have to be reported by the head of the undertaking to the local mining police within three days after the head has had cognisance of them, according to section 1552 of the Imperial Insurance Code. The employees to which the accidents refer are the technical employees, and the wage earners 1. From 1921-1925 the number of workers was represented by "full-time" workers calculated by dividing the number of normal manshifts worked (including shorttime shifts reduced to shifts of normal length), i.e. the total shifts exclusive of overtime shifts, by the number of working days. Since 1926, the total number of manshifts worked, including overtime shifts reduced to shifts of normal length is taken as basis for the calculation, a method already applied in most districts before 1921. The number of working days consists of the calendar days after deduction of Sundays, legal holidays and church holidays on which the mine was completely idle. The data are derived from the official Prussian production and wages statistics. The tonnage to which casualties were related till 1925 was the saleable coal as reported in the statements for production statistics (Decree of 15 January 1913). 1 The " employees " include also those persons working for their own account ( Unternehmer-Arbeiter) who are insured in the trade fund (Berufsgenossenschaft) of the respective mine. Not included are persons not belonging to the mine (Betriebsfremde) and only temporarily present there. — 74 — Casualties are given for the various coal-mining districts by causes. The classification by causes was somewhat changed in 1923. Since 1926, non-fatal casualties are classified by duration of incapacity to work, as follows : casualties entailing incapacity to work of (o) more than thirteen weeks ; (b) more than four weeks and not more than thirteen weeks ; (c) four weeks or less. Measures of Risk The two main measures of accident risk employed in Prussian statistics are : 1 Totar A and fatal casualties per 1,000 men : (a) underground, (b) surface and (c) total. 2. Similar rates by causes. 3. Tonnage rates indicating the number of fatal casualties per 1,000,000 tons were calculated up to 1925. Various tables on special groups of casualties, on percentages of all casualties for fatal casualties, and percentages of total casualties by causes, are occasionally published in the text. The tables below relate to coal mines only. Insurance Statistics Accident statistics of the insurance institutions are published in the Amtliche Nachrichten des Reichsversicherungsamtes, bui; only for the mining trade unions generally, and not for coal mining in particular. Moreover, accidents are compensated for by the accident insurance institutions only after thirteen weeks of disability, as for the first thirteen weeks benefits are paid out of the sickness insurance fund. It is evident that rates must be much lower than in the coal-mining statistics. Accident rates are calculated per thousand persons insured as well as per thousand full-time workers. The full-time worker is found by dividing the days worked by 300. These rates are given separately for fatal casualties, casualties entailing permanent total and partial disability, and casualties entailing temporary disability. Under a special heading all casualties reported — that is to say, all casualties disabling for more than three days — are published separately. Casualties, though not rates, are classified according to age and sex. Juveniles under sixteen years and adults are distinguished. Casualties, moreover, are classified by causes according to the general scheme of the insurance statistics which differs from the scheme of causes for coal mining statistics. T A B L E V I I . —• N U M B E R O F F U L L - T I M E W O R K E R S A N D T E C H N I C A L E M P L O Y E E S , N U M B E R O F FATAL C A S U A L T I E S , A N D R A T E S P E R 1 , 0 0 0 F U L L T I M E W O R K E R S A N D T E C H N I C A L E M P L O Y E E S IN P R U S S I A , 1 9 2 3 - 1 9 2 6 ( U N D E R G R O U N D AND S U R F A C E ) Underground Year Casualties by all causes 1923 1924 1925 1926 Rumbar Bato 899 942 1,180 982 2.0 2.6 3.3 2.78 Surface Full-time Full-time workers Casualties workers by and and technical all causes technical employees employees 447,060 362,084 362,209 353,137 Hamber Bat« 126 148 140 111 0.65 1.10 1.10 0.91 193,188 131,153 127,254 122,181 Total Casualties by all causes Rombar lata 1,025 1,090 1,320 1,093 1.60 2.21 2.70 2.30 Full-time workers and technical employees 640,248 493,237 489,463 475,318 — 75 TABLE VIH. —• NUMBER OF FATAL AND TOTAL CASUALTIES AND HATES PER 1,000 FULL-TIME WORKERS, UNDERGROUND, BY CAUSES, IN PRUSSIA, 1 9 2 3 - 1 9 2 6 Year and severity of casualties Falls of stone Number Fatal casualties 1923 1924 1925 1926 Total casualties 1923 1924 1925 1926 Year and severity of casualties Fatal casualties 1923 1924 1925 1926 Total casualties 1923 1924 1925 1926 By mining tools and machinery Kate Number Bate 292 434 461 491 0.7 1.2 1.3 1.1 9,243 23,027 27,446 29,546 20.7 63.6 75.8 83.7 At the working face 2 0.004 1 0.003 7 0.019 4 0.011 312 618 971 1,081 By explosives, etc. Number Rate Numbei 8 27 21 25 0.02 0.07 0.07 0.07 2,805 9,285 12,045 13,729 6.3 25.6 33.3 38.9 0.8 1.7 2.7 3.1 In main shafts In small blind shafts and slopes dumber Rate dumber Rate In level gangways Number Rate 68 85 77 68 0.2 0.2 0.2 0.2 186 210 217 214 0.4 0.6 0.6 0.6 87 104 99 114 0.2 0.3 0.3 0.3 1,075 2,193 2,423 2,290 2.1 6.1 6.7 6.4 3,178 5,954 7,096 7,358 7.1 16.4 19.6 20.8 9,923 20,958 24,153 25,653 22.2 57.6 66.7 72.6 By gases and coal dust Mine fires Kate Number Rate Numbe r Rate 21 32 37 26 0.05 0.09 0.10 0.07 153 30 237 24 0.30 0.08 0.65 0.07 63 3 2 4 0.141 0.008 0.005 0.011 207 312 328 297 0.5 1.0 0.9 0.8 235 75 313 80 0.52 0.20 0.90 0.23 105 36 25 28 0.20 0.10 0.07 0.08 Other causes Number Rate 19 16 16 12 0.05 0.04 0.04 0.03 4,171 7,953 7,413 8,532 9.30 22.00 20.50 24.20 GREAT BRITAIN Statistics of the divisional inspectors and a general report with statistics for mines and quarries by the Chief Inspector of Mines are published in the Annual Reports of the Secretary for Mines and the Annual Reports of the Chief Inspector of Mines by the Mines Department of the Board of Trade, in virtue of returns made under the Coal Mines Act, 1911, the Notice of Accidents Act, 1906, and the Workmen's Compensation Act, 1923. Scope Up to 1923 the statistics covered all fatal and non-fatal accidents causing incapacity to work for more than seven days. From 1924 onwards all accidents disabling for more than three days were included in accordance with the new provision of the Workmen's Compensation Act of 1923. Moreover, accidents by which the worker is " seriously " injured have ot be reported to the Divisional Inspector of Mines irrespective of the duration of incapacity, according to the Coal Mines Act, 1911 \ Data and Sources Reports cover all mines in Great Britain and Ireland under the Coal Mines Act, up to 1921 (inclusive) and all mines in Great Britain only from that date onward. For some tables, accidents in stratified ironstone mines in Cleveland, Lincolnshire and Northamptonshire are excluded. Accidents are reported by the employers to the mining inspectors. A fatal casualty is one leading to death within not more than one year and one day or up to the date of the report. The number of employees is the average number on the colliery books as reported in quarterly statements made under the Coal Mines Acts. They include all wage earners on the colliery books at the specified dates, that is, officials (overmen, deputies, firemen, examiners, etc.) ; workers employed in raising or in handling coal or other minerals ; tradesmen (mechanics, smiths, joiners, masons, electricians, etc.) ; workers employed in washing coal on premises adjacent and belonging to the mine, and workers employed on sidings at the pit and on private branch railways and tramways. Workpeople employed at ancillary undertakings, such as coke ovens, briquette works, etc., are not included. Moreover, clerks and salaried persons ordinarily employed in or in connection with the management of the colliery, or in and about its offices, are included, such, for instance, as surveyors, undermanagers and draughtsmen, weighing clerks, despatch clerks, time-keepers and store-keepers, if wholly or mainly employed in clerical work. The manshifts worked are those furnished by the monthly ascertainments of the joint auditors to the district boards instituted by the terms of the national wages agreements. They cover approximately 95 to 98 per cent, of the industry. The total number for all mines is estimated. Manshifts include overtime and week-end shifts reduced to normal shifts. The minerals raised are the saleable coal as stated by the quarterly returns under the Coal Mines Act and the minerals raised together with coal or, in mines under the Coal Mines Act, such as ganister, fireclay, limestone and ironstone. Both the number of accidents and the number of persons killed or injured, i.e. of casualties, are published. The latter are given separately for the several districts. They are classified by causes, and since 1925 by the nature and location, as well as duration of the injury. For 1927 casualties were classified by age-groups. Measures of Risk] Three measures of risk are published in the statistics. 1. Fatal casualties per 1,000 workers employed, classified by main groups of causes, underground and surface. 2. Casualties (fatal plus non-fatal casualties) per 100,000 manshifts worked are calculated since 1922. They are classified by districts. Rates for fatal casualties separately for the whole country are published in the texts. Underground casualties per 100,000 1 The following classes of accidents are included : accidents causing fracture of head or limb, or dislocation of limb, or any other serious personal injury, accidents caused by explosion of gas or any explosive or by electricity or by overwinding, and causing any personal injury whatever. — 77 — manshifts classified by three main groups of causes are given in a special table. The three groups of causes are : falls of ground at the working face, haulage accidents below ground, and other miscellaneous accidents below ground. For the first group the manshifts worked at the coal face were till 1926 taken as exposure to risk, for the second and the third group manshifts worked elsewhere below ground. Since 1927 casualties by falls of ground and haulage are related to all manshifts worked underground, and a surface rate is calculated similarly. 3. Tonnage rates, that is, fatal casualty rates per million tons of mineral raised. Fatal and non-fatal casualties combined per 1,000 persons employed are calculated for age-groups since 1927, underground and surface. Occasionally non-fatal casualty rates per thousand employees and 100,000 manshifts have been calculated. Percentages by causes are not usually given but were calculated for the three groups of causes above mentioned, e.g. in 1923, by districts. TABLE IX. — NUMBER OF WORKERS EMPLOYED, NUMBER OF FATAL CASUALTIES, AND RATES PER 1,000 WORKERS EMPLOYED IN GREAT BRITAIN, 1 8 9 3 - 1 9 0 2 , 1903-1912, 1913 AND 1920-1927 (MINES UNDER THE COAL MINES ACTS) Underground Year Number of workers employed 1893-1902 1 1903-1912 1 1913 1 1920 1 a 19211 1922 1923 1924 1925 1926 3 1927 588,446 772,234 909,834 990,359 918,066 933,029 979,785 979,108 890,849 899,778 824,866 Casualties from all causes Number Rate 895 1,130 1,580 965 682 999 1,179 1,087 1,028 577 1,033 Total Surface 1.52 1.46 1.74 0.97 0.98 1.07 1.20 1.11 1.15 1.22 1.25 Number of workers employed 143,945 185,614 218,056 257,865 226,245 229,725 240,646 251,140 226,979 228,431 212,525 Casualties from all causes Number Bate 120 145 173 138 74 106 118 114 108 72 95 0.83 0.78 0.79 0.54 0.42 0.46 0.49 0.45 0.48 0.53 0.45 Number of workers employed 732,391 957,848 1,127,890 1,248,224 1,144,311 1,162,754 1,220,431 1,230,248 1,117,828 1,128,209 1,037,391 1 United Kingdom. ' Figures based on statistics of first nine months. * Figures based on statistics of first four months. Number of workers in March. Casualties from all causes Number Bate 1,015 1,275 1,753 1,103 756 1,105 1,297 1,201 1,136 649 1,128 1.39 1.33 1.55 0.88 0.87 0.95 1.06 0.98 1.02 1.08 1.09 TABLE X. NUMBER OF WORKERS EMPLOYED, NUMBER OF FATAL CASUALTIES, AND RATES PER 1,000 WORKERS EMPLOYED UNDERGROUND, BY CAUSES, IN GREAT BRITAIN, 1893-1902, 1903-1912, 1913 AND 1920-1927 Year Number of workers employed By explosiona of fire-damp or coal dust Nombir Rate 1893-1902 11 1903-1912 191311 1920 1921ia 1922 1923 1924 1925 1926 s 1927 588,446 104 772,234 133 909,834 462 990,359 26 918,066 19 933,029 73 979,785 60 979,108 35 890,849 29 899,778 5 824,866 72 0.18 0.17 0.51 0.03 0.03 0.08 0.06 0.03 0.03 0.01 0.07 By falls of ground Shaft accidents Haulage accidents Miscellaneous accidents Numbar Kate Number Rate Mir Rate tabir Rate 448 573 620 544 383 551 585 607 547 334 565 0.76 0.74 0.68 0.55 0.55 0.59 0.60 0.62 0.61 0.68 0.69 79 81 98 40 26 39 58 59 34 28 53 168 235 251 237 170 211 314 262 260 131 232 0.29 0.30 0.28 0.24 0.25 0.23 0.32 0.27 0.29 0.32 0.28 96 108 149 118 84 125 162 124 158 79 111 0.16 0.14 0.16 0.12 0.12 0.13 0.16 0.13 0.18 0.15 0.13 ' united Kingdom. * Figures based on statistics of first nine months. ' Figures based on statistics of first four months. 0.13 0.11 0.11 0.04 0.03 0.04 0.06 0.06 0.04 0.06 0.06 — 79 — TABLE XI. —• NUMBER OF MANSHIFTS WORKED, NUMBER OF FATAL AND NON-FATAL CASUALTIES (COMBINED), AND RATES PER 1 0 0 , 0 0 0 MANSHIFTS UNDERGROUND AND SURFACE COMBINED, BY DISTRICTS, IN GREAT BRITAIN 1 9 2 2 - 1 9 2 5 * 1923 1922 District Northumberland.. Durham Cumberland Lancashire, North Staffordshire and North Wales Eastern Division . South Staffordshire South Wales and Monmouthshire Bristol Forest of Dean . . . Kent Scotland Manshifts worked Fatal and non-fatal casualties Number Per 100,000 shifts Manshifts worked Northumberland . . South Wales and Monmouthshire Forest of Dean . . . Kent Number Per 100,000 shifts . 9,947 26,284 2,223 64.3 67.2 77.6 17,066,000 45,548,000 3,139,000 10,852 28,777 2,327 63.6 63.2 74.1 32,337,000 4,468,000 83,580,000 23,486 2,747 57,592 72.6 61.5 68.9 37,200,000 5,059,000 92,905,000 27,194 3,442 68,259 73.1 68.0 73.5 2,560,000 1,290 50.4 2,885,000 1,668 57.8 60,037,000 586,000 1,723,000 1,250,000 411,000 36,531,000 40,048 310 763 509 295 20,736 66.7 52.9 44.3 40.7 71.8 56.8 66,592,000 641,000 2,083,000 1,569,000 509,000 41,336,000 44,899 346 917 788 348 23,086 67.4 54.0 44.0 50.2 68.4 55.8 280,947,000 186,230 66.3 316,532,000 212,903 67.3 Manshifts worked 1925 Fatal and non-fatal casualties Number Per 100.000 shifts Manshifts worked Fatal and non-fatal casualties Number Per 100.000 shifts 16,132,000 44,156,000 2,934,000 9,506 22,887 2,255 58.9 51.8 76.9 13,734,000 35,699,000 2,624,000 8,381 19,385 2,004 61.0 54.3 76.4 37,431,000 5,083,000 93,821,000 22,926 2,904 68,418 61.2 57.1 72.9 32,458,000 4,532,000 90,741,000 19,539 2,521 66,121 60.2 55.6 72.9 2,997,000 1,527 51.0 3,013,000 1,336 44.3 64,185,000 449,000 2,058,000 1,685,000 354,000 40,336,000 42,004 227 878 781 297 21,305 65.4 50.6 42.7 46.4 83.9 52.8 56,572,000 425,000 1,855,000 1,465,000 514,000 35,878,000 37,478 206 850 749 398 19,507 66.2 48.5 45.8 51.1 77.4 54.4 311,621,000 195,915 62.9 279,510,000 178,475 63.9 Lancashire, North Staffordshire and Eastern Division.. South Staffordshire Fatal and non-fatal casualties 15,459,000 39,142,000 2,863,000 1924 District J * Stratified ironstone mines in Cleveland, Lincolnshire and Northamptonshire are excluded, while being included in the other tables. — 80 — TABLE XII. ESTIMATED NUMBER OF MANSHIFTS WORKED AT THE COAL FACE AND ELSEWHERE BELOW GROUND, NUMBER OF PERSONS KILLED AND INJURED, AND RATES PER 1 0 0 , 0 0 0 MANSHIFTS, UNDERGROUND, BY CERTAIN CAUSES, IN GREAT BRITAIN, • • ' •' Year 1922 1923 1924 1925 1926 1922-1926 Falls of ground at the working face Haulage accidents (below ground) Number of persons Estimated number Number of persons Estimated number killed and injured of manshifts killed and injured of manshifts actually worked actually worked at Total Per 100,000 elsewhere below Total Per 100,000 coal face ground number shifts number shifts 107,011,000 123,023,000 120,692,000 108,679,000 38,456,000 Year 1922 1923 1924 1925 1926 50,261 55,046 51,314 47,904 18,130 47.0 44.7 42.5 44.1 47.1 111,431,000 123,144,000 121,539,000 108,270,000 37,852,000 47,007 54,589 49,254 43,700 15,868 Other (miscellaneous) accidents (below ground) Number of persons killed Estimated number and injured of manshifts actually worked elsewhere below Per 100,000 ground Total number shifts 111,431,000 123,144,000 121,539,000 108,270,000 37,852,000 57,936 67,287 61,304 56,581 20,999 52.0 54.6 50.4 52.3 55.5 INDIA Coal-mining statistics are published by the Bureau of Mines Inpection in India in the reports of the Chief Inspector of Mines. These reports apply to the mines under the Indian Mines Acts of 1901 and 1923 which cover British India, including Baluchistan and Sonthal Parganas. The native States, which occupy about two-fifths of the Empire, are not included. Hyderabad has its own department of mines which publishes an annual report. Scope of Statistics The Act of 1901 applied to any mine more than 20 feet deep and extending beneath the super-adjacent ground. The Act of 1923 does not contain this limitation. A mine includes all works, machinery, tramways and sidings, whether above or below ground, in or adjacent or belonging to a mine. Since 1923 " any part of such premises on which a manufacturing process is being carried on is excluded unless such process is a process for coke making or the dressing of minerals". The statistics include fatal casualties, and casualties entailing serious bodily injury. The latter are defined as injuries which involve " the permanent loss of the use of, or permanent injury to any limb, or the permanent loss of, or injury to the sight or hearing, or the fracture of any limb, or the enforced absence of the injured person from work for a period exceeding 20 days ". 42.2 44.3 40.5 40.4 41.9 — 81 — Data and Methods of Computation Casualties of the kind mentioned above occurring in or about the mine as well as disasters, by which is meant accidental explosions, ignition, outbreak of fire or eruption of water, must be reported by the owner, agent or manager to the chief inspector through the district or subdivisional magistrate. The persons employed to which the fatal casualties are referred do not include the superior supervising staff but all subordinate officials and persons employed on sidings, loading wharfs, private railways, surface tramways and in carting. The daily average number of persons employed (or rather present) is found by dividing the aggregate number of daily attendances of persons permanently and temporarily employed by the number of days worked by the mines. The output is the coal raised, including colliery consumption and coal raised for coke making. Casualties are tabulated in three ways: 1. Fatal and serious causalties are given separately below ground and above ground for districts and mineral fields. 2. Fatal casualties are classified according to age and sex of the worker : figures are given for adult males and females over 12 years and children under 12 years. 3. Casualties are classified by causes. Measures of Risk The measures of accident risk calculated are : 1. Fatal casualties per 1,000 persons employed : (a) underground ; (b) open workings (since 1926) ; (c) surface ; (d) underground, open workings and surface combined. [Continued on next page. TABLE XIII. — AVERAGE NUMBER OF PERSONS EMPLOYED DAILY, NUMBER OF FATAL CASUALTIES, AND RATES PER 1,000 PERSONS EMPLOYED, IN INDIA, 1 9 1 1 , 1 9 1 3 AND 1920-1926 (UNDERGROUND AND SURFACE) Underground Year 1911 1912 1913 1920 1921 1922 1923 1924 1925 1926 1 Average number of persons employed dally 68,867 77,930 84,828 102,922 108,957 107,494 108,994 115,873 110,630 115.4551 Surface Fatal casualties Number Rate Average number of persons employed daily 136 141 175 143 237 185 312 203 173 147 1 1.97 1.80 2.06 1.39 2.18 1.72 2.86 1.75 1.56 1.272 37,731 43,462 48,214 73,021 81,690 76,861 73,607 71,215 62,510 55,173 Total Fatal casualties Number Rate 12 16 10 29 20 24 20 27 13 24 0.31 0.36 0.21 0.40 0.24 0.31 0.27 0.38 0.21 0.43 Average number of persons employed daily 106,598 121,392 133,042 175,943 190,647 184,355 182,601 187,088 173,140 170,628 Fatal casualties Number Rate 148 157 185 172 257 209 332 230 186 171 1.38 1.29 1.39 0.93 1.35 1.13 1.82 1.23 1.07 1.00 Including open workings, which are given separately for the first time in coal statistics for 1936. • Calculated by the International Labour office. — 82 — Non-fatal (serious) casualties per 1,000 workers : (a) underground ; (b) open workings ; (c) surface ; (d) underground, open workings and surface combined. Fatal casualties per million tons raised at coal mines. JAPAN Coal-mining accident statistics are published in the General Review of the Mining Industry by the Bureau of Mines, Department of Commerce and Industry, formerly the Department of Agriculture and Commerce. The statistics do not refer to Korea, Formosa, and Sagalia. All accidents have to be reported by the mineowner or his representative to the governor of the prefecture. There are forty-three local prefectures and three urban prefectures (Tokio, Kioto and Osaka), whilst Hokkaido forms a separate district. The statistics cover accidents on the one hand, and all casualties on the other hand, fatal and non-fatal. Non-fatal casualties are classified according to whether persons are seriously injured or slightly injured. " Seriously injured " means incapacitated for more than one week. Casualties are classified by detailed causes, and for each group of causes by age, that is to say, for persons under fifteen, under twenty, and over twenty years of age, as well as by sex. TABLE XIV. NUMBER OF FATAL, SERIOUS, AND SLIGHT CASUALTIES, BY SEX, IN JAPAN, 1 9 2 1 - 1 9 2 4 (UNDERGROUND AND SURFACE) Year Place Underground 1921 Total Underground 1922 Total Underground 1923 Total Underground 1924 Total Fatal casualties Serious casualties Slight casualties Total casualties Men Women Men Women Men Women Men Women 502 35 537 99 7 106 3,838 368 4,206 1,015 80 1,095 114,322 11,830 126,152 25,917 2,230 28,147 118,662 12,233 130,895 27,031 2,317 29,348 438 25 463 74 10 84 4,187 371 4,558 1,238 91 1,329 109,593 10,869 120,462 26,206 2,001 28,207 114,218 11,265 125,183 27,518 2,102 29,620 527 43 570 87 6 93 4,071 395 4,466 1,340 96 1,436 125,773 12,678 138,451 26,679 2,098 28,777 130,371 13,116 143,487 28,106 2,200 30,306 680 34 714 121 4 125 1,336 445 4,781 1,262 71 1,333 114,810 11,424 126,234 26,148 2,203 28,351 119,826 11,903 131,729 27,531 2,278 29,809 — 83 — NETHERLANDS Coal-mining accident statistics are published annually by the Chief Mining Engineer in the Jaarverslag van den Hoofdingenieur der Mijnen over het jaar . . . Scope Statistics cover all pit-coal mines in the Netherlands. They include all fatal casualties which have led to death before 2 June, as the report lias to be presented to the Minister of " Waterstaat " before that date. Non-fatal casualties reported are those disabling for more than three weeks. Only casualties are reported. Data and Sources Accidents are reported, under the Accidents Act of 1901, section 61, subsection 4 (Ongevallenwet 1901, Artikel 61, vierde lid.), to the National Insurance Bank (Rijksverzekeringsbank). They are classified as total, fatal, and non-fatal casualties, underground and surface. Latterly casualties entailing incapacity of more than forty-two days are separately classified. A detailed classification by causes is given, underground and surface. The number of workers employed is found by dividing the total number of workers as ascertained on the fortnightly pay sheets by the periods of payment ; sick workers, those injured by accidents, and those absent for other reasons are comprised, but not those serving in the forces. Measures of Risk 1. Fatal and total casualties per 1,000 workers employed : (a) underground and surface : (b) underground and surface combined. 2. Fatal and total casualties per 100,000 tons produced. 3. Percentages by causes for total casualties. TABLE XV. NUMBER OF FATAL CASUALTIES AND RATES PER 1,000 WORKERS EMPLOYED, IN THE NETHERLANDS, 1 9 2 2 - 1 9 2 7 (UNDERGROUND AND SURFACE) Year 1922 1923 1924 1925 1926 1927 Undergronud Surface Total Number Kate Number Rate Number Bate 20 31 26 39 30 28 1.10 1.57 1.19 1.73 1.28 1.13 6 4 2 2 4 6 0.79 0.52 0.25 0.24 0.46 0.66 26 35 28 41 34 34 1.01 1.28 0.93 1.33 0.39 0.36 — 84 — TABLE XVI. AVERAGE NUMBER OF WORKERS INSURED, NUMBER OF FATAL AND NON-FATAL TEMPORARILY AND PERMANENTLY DISABLING CASUALTIES BY CAUSES, AND RATES PER 1,000 WORKERS INSURED (ALL CAUSES), IN POLAND (VOIEVODIE OF SILESIA) IN 1926 1935 Causes Power machinery, power transmission and labour machinery . . . . Lifts, etc., winding ma- 1925-1927 1927 Injuries Injuries Injuries anent orary anent orary anent orary Fatal Fatal cases Perm- Temp- cases Perm- Temp- cases Perm- Temp- 4 16 65 8 18 77 4 20 88 5 13 70 10 10 38 5 8 18 1 2 1 2 0 1 0 0 0 2 24 20 4 10 12 6 11 21 4 78 5 93 0 404 8 98 2 79 1 378 11 77 2 82 4 349 10 6 16 9 120 84 14 13 16 16 101 180 6 3 8 15 111 223 34 3 5 47 6 44 396 25 103 29 0 7 40 3 45 322 23 51 26 0 8 39 0 20 315 18 53 275 1,293 193 239 1,184 146 Explosions of boilers, steam tanks, steam Explosions of explosives and gases Inflammable, corrosive burning substances, heated metals, e t c . . . . Falls of ground, stone etc. Falls of persons in shafts, on slopes, galleries, Loading and unloading. Haulage by machinery, animals and other means IM 205 1,200 R a t e per 1,000 workers 1.64 2.96 13.92 243 3.01 14.90 1.83 2.57 15.06 Average number of workers insured . . . . 92,869 79,44() 79,664 — 85 — POLAND Coal-mining accident statistics are compiled by two institutions : the Accident Insurance Fund of Lwow, for the formerly Austrian and Russian territories, and the Accident Insurance Fund of the District of Silesia (Voievodie) in Krolewska Huta, for the Upper Silesian part of the Voievodie of Silesia. The statistics of the latter institution are very detailed. They are referred to in the following analysis. Scope The source of the statistics being the insurance institutions, casualties reported are those first compensated for in the year of observation Data and Methods 0/ Computation Casualties are classified as fatal casualties, casualties entailing permanent (total and partial) incapacity to work, and casualties entailing temporary incapacity only. All casualties are classified by sex, civil status and age (adults and juveniles). The classification by sex and civil status is also applied to the different categories of casualties. A classification by nature and location of injury, going into great detail, is made. Casualties (total) are classified by fault. Six main groups are distinguished : (1) fault of fellow-workers ; (2) dangerous operations ; (3) insufficient measures on the part of the management ; (4) no use made of the means of protection by the injured ; (5) lack of attention and skilfulness (maladresse) on the part of the injured ; (6) misfortune and unknown causes. These groups are further subdivided, the subdivisions specifying the nature of the fault. The classification by causes, being made for insurance purposes, is not wholly adapted to coal-mining risks as regards the main groups, but is largely specialised for coal mining in the subdivisions. We have therefore in the appended classification summarised the groups of causes which are of little significance for coal miners and reproduced the subdivisions of interest for coal-mining statistics. The workers covered by the statistics are the average number of workers insured in the year of observation. Measures of Risk 1. Fatal casualties per 1,000 persons insured. 2. Casualties entailing permanent incapacity per 1,000 workers insured. 3. Casualties entailing temporary incapacity per 1,000 workers insured. — 86 — UNITED STATES OF AMERICA The United States Bureau of Mines, formerly of the Department of the Interior, and, since 1 July 1925, oi the Department of Commerce, started publishing coal-mine accident reports in 1912. In 1913 the Bureau published a report covering all fatal accidents described in the State inspectors' reports beginning with 1896. In 1916 the Bureau published all fatal accidents described in the inspectors' reports from the beginning of the inspection service. State inspection began in Pennsylvania in" 1869. Data regarding major explosions or fires in pre-inspection years back to 1839 were given for some States. From 1889 to 1909, inclusive, States reporting accidents comprised the coalfields whose production represented 90 to 99 per cent, of the nation's total output of coal. From 1910 onward the States reporting accidents include all States producing coal. Accident statistics are now regularly published in the report entitled Coal-Mine Fatalities in the United States. Scope of Statistics The United States statistics comprise fatal accidents only as the State reports are incomplete and of different scope with regard to non-fatal accidents. The inspectors' reports publish non-fatal accidents under various forms. The introduction of workmen's compensation has led to more complete records in recent years. All fatalities not directly attributable to the coal-mining industry are eliminated, such as accidents in coke ovens, briquette making, milling or smelting. Office work is excluded from the statistics. Technical employees are included if the duties of such employees expose them to mining hazards. Data and Sources Data are derived from various sources. Fatal accidents are compiled from the State mine inspectors' monthly summaries furnished to the Bureau of Mines. They include those deaths that had resulted when the State mine inspector furnished his summary. The monthly reports are revised currently during the year, and further revised after the year is closed. Yearly data on production (since 1820) on the number of men employed (since 1889), the number of days worked (since 1890) and data on mining machines in use (since 1896), and the percentage of coal mined by machine, are derived from the Annual Volumes of the Mineral Resources of the United States, published by the United States Geological Survey. This division of the Geological Survey was transferred to the Bureau of Mines on 1 July 1925. Data on mining methods were taken both from the State mine inspectors' reports and from the United States Geological Survey. Since 1912 accidents have been published for each State and for each month. They are classified be detailed causes. — 87 — Measures of Risk Four different kinds of frequency rates have been published by the United States Bureau of Mines : 1. Fatality rates per average number of workers employed. As some of the State mine inspectors' reports gave the number of employees for fiscal years and, moreover, some States included cokeoven employees, the number of employees from 1889 onward were those calculated by the United States Geological Survey, except for the years 1909-1911. 2. Frequency rates per thousand 300-day workers are calculated irrespective of the length of the working day. These rates are specified according to underground and surface working. As the United States Geological Survey did not originally publish figures for underground and surface employees separately, the proportion of underground and surface workers was calculated on the basis of the proportion found by the Bureau of Mines in 1911. Data of underground and surface workers, however, are now published separately by the United States Geological Survey. 3. Frequency rates per thousand 2,000-hour workers were calculated for 1903 to 1913 and have been continued until 1925. Now, rates per million hours worked are given. 4. Normal fatality rates by causes, per million hours of labour and per million tons "of coal which are calculated by excluding fatal accidents caused by mining disasters killing five or more than five persons, are published for bituminous coal mines. 5. Fatality rates per million hours of labour are calculated for groups of causes. Underground, shaft, and surface accidents are given separately. Accident rates are calculated, first, in relation to total hours of exposure underground ; second, in relation to total hours of exposure on the surface ; and third, in relation to total hours of exposure underground and surface combined. For 1925, rates per million men-hours worked, by causes, were calculated for each State, back to 1911. Now, rates per 1,000 300-day workers by more detailed causes are given. 6. Fatality rates per million short tons of coal raised are calculated from 1870 onwards and the production of coal per death is also computed. 7. Percentages of total fatalities by causes are published regularly. 8. Fatality rates, normal, exceptional and total, have been computed for groups of States, classified according to the percentage of coal mined by machine (see Bulletin No. 115, published in 1916). Methods of Calculation and Collection The United States Bureau of Mines obtains its data from the owners (producers') written reports, which are collected by correspondence and supplemented by visits of field agents. The Bureau of Mines has, however, no power to enforce these records. The so-called commercial mines, those producing at least 1,000 tons (per year) and working more or less steadily year after year, receive schedules of enquiry by mail. For smaller mines, the railroads sometimes supply data on the coal conveyed, from which estimates as to the number of men employed are made. For the so-called " sporadic wagon mines", satisfactory results can not be obtained by mail. The Bureau therefore asked the railroads in 1920, 1922 and 1923 to furnish lists of the wagon mines which — 88 — had conveyed coal to indicate the dates of opening and closing of the mines, and the number of car loads conveyed. In years of depression, however, most wagon mines are out of business and records are not kept. No reports as to the number of employees or the time worked by wagon mines have ever been obtained. The number of men employed is not calculated according to any common method. No records of man-days are kept by bituminous coalowners at most mines for tonnage workers, that is, those paid on piece rates. It was found that most of the figures reported represented the number of men commonly dependent on the mines for employment, that is, those reporting for work when the mine started, plus absentees, and not the average number of men actually working at any one time, nor the aggregate number during the j r ear, nor the absolute average number of men on the payrolls. The days worked by the mine are the total number of full days the mine was in operation and raised coal (tipple days), part of days being reduced to equivalents in full days. The averages for States are weighted by the number of men employed in each mine x. The hours of labour refer to the " number of hours that the mines are supposed to be in operation, cars and orders permitting, and not the number of hours the mine actually worked ". The tonnage men are piece workers and not obliged to put in a certain number of hours. The figures given therefore really indicate " the number of hours the men had an opportunity to work during a full day on the assumption that there was a full run of cars and that the market conditions were favourable to full time operation. They do not mean that all the tonnage men worked the number of hours stated ". The Bureau of Mines calculates the 300-day workers by multiplying the average number of workers in each State by the days the mine was in operation in that State during the year. The 2,000-hour worker is found for each State by multiplying the number of workers by the number of hours per day in the respective mines, that is, the number of workers in the eight-hour mines is multiplied by eight, in the nine-hour mines by nine, and so on, and multiplying these products by the number of days the mines were in operation. The total is divided by 2,000. The rates per thousand 300-day workers and per million man-hours are calculated separately for underground workers, including shaft workers, and for sarface workers. Recently, attempts have been made to calculate severitv rates. 1 UNITED STATES GEOLOGICAL SURVEY : Mineral Resources of the United States, 1922, p. 401. (In 1925 the division of the United States Geological Survey handling this work became a part of the Bureau of Mines.) — 89 — TABLE XVII. ACTUAL NUMBER OF MEN EMPLOYED, NUMBER OF 300-DAY WORKERS, AVERAGE DAYS ACTIVE, TOTAL HOURS WORKED PER DAY, NUMBER OF 2 , 0 0 0 * ÎOUR WORKERS AND NUMBER OF FATAL CASUALTIES AND RATES PER 1,000 EMPLOYEES, 'ER 1,000 300-DAY WORKERS AND PER 1,000 2,000-HOUR WORKERS IN THE UNITED STATES, 1911-1913 AND 1920-1927 Men employed Year Actual number 300-day workers Average days active 1911 1912 1913 1920 1921 1922 1923 1924 1925 1926 1927 728,348 722,662 747,644 784,621 823,253 844,807 862,536 779,613 748,805 759,033 757,000 534,122 541,997 593,131 601,283 474,529 405,056 560,646 499,896 480,227 559,426 — 220 225 238 230 173 144 195 192 192 221 — Fatal casualties Total hours 2,000-hour worked per workers per 1,000 per 1,000 per 1,000 Number empday 300-day 2000-hour loyees workers workers 6,136,211 6,149,053 6,333,895 6,309,400 6,622,765 6,805,520 6,954,302 6,288,930 6,043,406 6,121,449 — 674,983 691,768 753,734 725,581 572,869 489,997 678,044 603,737 580,167 — — 2,656 2,419 2,785 2,272 1,995 1,984 2,462 2,402 2,234 2,518 2,224 3.65 3.35 3.73 2.90 2.42 2.35 2.85 3.08 2.98 3.32 2.94 1 Calculated by International Labour Office, as discontinued in report. orked: 1.68. 4.97 4.46 4.70 3.78 4.20 4.90 4.39 4.80 4.65 4.50 — 3.93 3.50 3.69 3.13 3.48 4.05 3.63 3.98 3.85 3.72 ! — Kate per million hours — 90 — TABLE XVIII. NUMBER OF DAYS OF LABOUR PERFORMED, AND OF 300-DA WORKERS, NUMBER OF FATAL CASUALTIES AND RATES PER 1,000 300-DAY WORKER IN THE UNITED STATES, 1 9 1 1 - 1 9 1 3 AND 1 9 2 0 - 1 9 2 6 (UNDERGROUND AND SURFACI Underground Surface Year of Fatal Days of labour Number casual300-day performed ties workers 1911 1912 1913 1920 1921 1922 1923 1924 1925 1926 133,283,700 135,251,775 148,061,942 144,236,367 118,297,091 100,211,422 139,429,422 125,279,912 120,654,263 140,857,480 444,279 450,840 493,540 480,788 394,324 334,038 464,765 417,599 402,181 469,525 TABLE XIX. Kate per per of Fatal Bate Days of labour Number 1,000 1,000 300-day casual- 300-day performed 300-day workers ties workers workers 2,486 2,220 2,624 2,077 1,875 1,846 2,304 2,264 2,107 2,400 5.60 •;-.92 5.32 ¿.-.32 4.75 5.53 4.96 5.42 5.24 5.11 26,952,860 27,347,175 29,877,330 36,148,575 24,061,600 21,305,400 28,764,316 24,689,068 23,413,969 26,970,252 89,843 91,157 99,591 120,495 80,205 71,018 95,881 82,297 78,046 89,901 170 199 161 195 120 138 158 138 127 118 NUMBER OF MAN-HOURS WORKED, NUMBER OF FATAL CASUALTIES, AND RATES PER MILLION MAN-HOURS, UNDERGROUND BY CAUSES, IN THE UNITED STATES, 1 9 2 1 - 1 9 2 6 Year Falls of roof or coal Her 1921 1922 1923 1924 1925 1926 Year 1921 1922 1923 1924 1925 1926 1,025 908 1,167 1,062 1,080 1,214 Haulage Kate 1.077 1.123 1.038 1.053 1.111 1.069 Electricity Kjnber Rate fallir Rate Number 0.358 0.422 0.367 0.351 0.372 0.381 126 Sil 372 536 345 422 0.132 0.385 0.331 0.531 0.355 0.372 142 93 115 99 102 96 0.149 0.115 0.102 0.098 0.105 0.085 80 74 75 80 84 96 1 Other causes Explosives Rate tab« 341 341 413 354 361 433 Gas or dust explosions Shaft accidents All causes Number Rate Number Rate Number Rate 125 78 116 104 101 104 0.131 0.096 0.103 0.103 0.104 0.091 36 41 46 29 34 35 0.038 0.051 0.041 0.029 0.035 0.031 1,875 1,846 2,304 2,264 2,107 2,400 1.969 2.284 2.049 2.244 2.168 2,114 Rate 0.084 0.092 0.067 0.079 0.086 0.085 Man-hours worked underground 952,084,500 808,173,400 1,124,171,900 1,008,690,500 971,756,300 1,135,436,100 1.89 2.18 1.62 1.62 1.50 1.94 1.65 1.68 1.63 1.31