AN INQUIRY INTO FACTORS THAT MIGHT EXPLAIN DIFFERENCES IN OCCUPATIONAL ACCIDENT EXPERIENCE OF SIMILAR SIZE FIRMS IN THE SAME INDUSTRY Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY Y—AGHOUB SHAFAI-SAHRAI 197.1 * --.\ LIBRARY Michigan State University I IIIIII II IIIIIIIIIIIIIII III III IIIIII I L ’ 31293 00686 2985 This is to certify that the thesis entitled AN INQUIRY INTO FACTORS THAT MIGHT EXPLAIN DIFFERENCES IN OCCUPATIONAL ACCIDENT EXPERIENCE OF SIMILAR SIZE FIRMS IN THE SAME INDUSTRY presented by Yaghoub Shafai-Sahrai has been accepted towards fulfillment of the requirements for Ph.D Jegreein Management Efi/flh fié/f/{Zw/VQ’Q Major professor Date (Lee M; / 9 7/; 07639 ”SING av :5" TIME & SUNS . PM" RI-‘IUERY INC, In! -r.n7 “CHIC!” | I II‘C“: _’v . In. __*’___D' . w-a’* / \I ‘\' SJ ‘31 J ~Q“! ," ,- HI“? C ‘ S a.“ I tins“ .9 fl ,. \‘q“‘| \ ‘I ABSTRACT AN INQUIRY INTO FACTORS THAT MIGHT EXPLAIN DIFFERENCES IN OCCUPATIONAL ACCIDENT EXPERIENCE OF SIMILAR SIZE FIRMS IN THE SAME INDUSTRY BY Yaghoub Shafai-Sahrai It has long been recognized that some industries because of the nature of tasks involved, have potentially higher work injury frequency and severity rates than others. Yet, within the same industry firms experience widely dif— ferent occupational injury rates. Furthermore, within each industry, injury rates tend to be relatively higher for small size firms than large, though some of the small industrial concerns demonstrate excellent safety perform- ance records. This study was an attempt to investigate empiri— cally some of the factors that could possibly provide an explanation for differences in accident experience and injury rates of similar size firms in the same industry. A series of hypotheses were formulated to be tested Concerning correlates of each of the following factors to 7" r‘ 5::CL q :6 S . *CIK: \L'J Yaghoub Shafai-Sahrai work accidents. The factors were: 1) TOP management's support and involvement in safety; 2) Formal educational level of workers; 3) Age of the employees; 4) Company recreational programs; 5) Promoting safety through employ- ee's family; 6) Safety rules; 7) Marital status of the workers; 8) Worker's length of service in a company; 9) Company age; 10) Accident record keeping system; 11) First line supervisors' relative span of control; 12) Relative age of production machinery and equipment; 13) Physical workplace conditions; 14) Safety devices and controls on machinery; and 15) Safety committees. Selection of eleven matched-pairs of industrial firms from eleven different industries, located in 16 dif- ferent locations throughout the State of Michigan, consti- tuted the sample of this study. Each pair was composed of two similar size firms in the same industry but with one of these two having higher injury rates than the other. For collection of data, company visitation and personal inter— viewing were the principal means. The major findings of this study are as follows: 1) The hypothesis that in the firms with better safety records, t0p management is highly interested and involved in company's safety 2) 3) 4) 5) 6) 7) Yaghoub Shafai—Sahrai performance, appears to be strongly supported by this study. Formal educational level of employees in the firms studied, did not correlate with the injury rates. Higher average age of employees is positively correlated with better safety performance records of the firms. Companies that provide recreational programs and facilities for their employees, tend to have considerably lower work injury rates than the companies without such programs and facilities. The hypothesis that firms with better safety records have tried to promote safety through employee's family was not supported by this study. Quality and quantity of safety rules showed little correlation with safety performance of the firms studied. Firms employing more married workers on the average tend to have significantly better safety performance records than the firms with 8) 9) 10) ll) 12) l3) l4) Yaghoub Shafai-Sahrai fewer married employees in their employ. This study supported the hypothesis that a positive correlation exists between low accident rates and higher average employee length of service. Age of the companies showed no significant correlation with safety performance. Among the companies studied, those with a better accident record keeping system were found to have considerably lower work injury rates. The larger span of control of the first line supervisors tends to be significantly corre- lated with higher work injury rates. Relative age of the production machinery and equipment did not tend to correlate with the accident experience. Roomy, and clean shOp environment with adequate temperature, ventilation, lighting and noise level, appears from this study, to be signifi- cantly correlated with low injury rates. The hypothesis that better and more safety devices on machinery contribute to attainment 15) Yaghoub Shafai-Sahrai of better safety performance records of the companies, was strongly confirmed by this study. Safety committees did not appear from this research to be a significant factor in attainment of a good safety performance record. AN INQUIRY INTO FACTORS THAT MIGHT EXPLAIN DIFFERENCES IN OCCUPATIONAL ACCIDENT EXPERIENCE OF SIMILAR SIZE FIRMS IN THE SAME INDUSTRY by Yaghoub Shafai-Sahrai A THESIS submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Management 1971 CDCOpyright by YAGHOUB SHAFAI -' SAHRAI 1971 TO: my wife Parvin my daughter Azita my son Armin ' 1 guwlo-p,‘ Lav“ uuu . A H flrqb. 3:854 H‘:.. o a Y‘als “ .‘ IA- hi 1 33C [332'ij ' h . LanLan A: . {HAN u ’ “tn ant meat has and time ACKNOWLEDGMENTS The successful completion of a research study of any magnitude is seldom the product of efforts of a single per- son. This thesis is no exception and I wish to eXpress my deep gratitude and appreciation to the many individuals and organizations who contributed to this field study. I am extremely grateful and indebted to the People and Government of Iran specially officials of the Imperial Iranian Army who provided me the Opportunity to pursue the Ph.D program. My sincere appreciation goes to the members of my dissertation guidance committee. Professor Rollin H. Simonds, chairman of the committee, who is an internationally known authority in the field of occupational safety manage- ment has provided eXpert counsel, encouragement, assistance and time under demanding work schedules and severe time constraints for which I am deeply grateful. But the debt I owe to Professor Simonds is not likely to be fully repaid. Professor Daniel Kruger generously contributed the initial idea for this endeavor and subsequently provided invaluable advice, encouragement and assistance to which I am sincerely thankful. Professor Robert Penfield supplied the initial .impetus and contributed substantially to this work to which ii I wish t ..,. ”I ”‘4‘! riat-u-‘ . C““'" .5.“ 5‘ H. V. sue . AC”‘AQ.< “Lt“: I , G V»,- ”b IA§\ F"gm.+. ~}"'”HI.. . . 'I‘n r: I‘u H {v ('11 v I wish to extend my gratitude. I also would like to thank Dr. Dalton E. McFarland, Professor and Chairman of the Department of Management for providing advice and assistance throughout the years of my study. The author is specially grateful to the presidents of the twenty-two firms included in this study who did not wish to be identified by name. I am also thankful for the assistance received from the employees and officials of the Michigan State Department of Labor. Mr. Don F. Jones, Executive Director, Labor Safety Council of Ontario, pro- vided invaluable references on industrial accident research to which I am sincerely grateful. My special thanks go to Mr. Ramon Aldag, a doctoral candidate in the Department of Management who read the manu— script, provided substantial editorial assistance and aided in statistical analysis of the data. Although generous assistance, direction, c00peration and encouragement supplied by above organizations and indi- viduals contributed to my ultimate success in this endeavor, the content of this study and research conclusions, are of course solely my responsibility. iii TABLE OF CONTENTS ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . . LIST OF TABLES . . . . . . . . . . . . . . . . . . LIST OF FIGUMS. O O C O O O O O O O O O O O O O 0 Chapter I. II. III, INTRODUCTION. . . . . . . . . . . . . . . Discussion. . . . . . . . . . . . . . . Nature of the Problem . . . . . . . . . Purpose of the Study. . . . . . . . . . METHOD OF STUDY . . . . . . . . . . . . . Introduction. . . . . . . . . . . . . . Sampling Method . . . . . . . . . . . . Method of Collecting Data . . . . . . . Analysis of Data. . . . . . . . . . . . Limitations of the Study. . . . . . . . Summary . . . . . . . . . . . . . . . . REVIEW OF THE PERTINENT LITERATURE. . . . Introduction. . . . . . . . . . . . . . Occupational Accident Research Before and After the Industrial Revolution . Company Size and Work Accident Records. TOp Management and Safety Performance of the Firm . . . . . . . . . . . . . Influence of Age on the Frequency and Severity of Accidents . . . . . . Length of Service and Occupational Accidents . . . . . . . . . . . . . . Worker Participation and Occupational Safety. . . . . . . . . . . . . . . . The First Line Supervisory Span of Control and work Accident Prevention. Effects of Some Employee Background Variables on the Job Attitude and Safety. . . . . . . . . . . . . . . . iv Page ii vii l6 l6 17 25 3O 32 35 36 36 37 48 52 55 59 62 69 74 Chapter a) Family Background b) Marital Status and Safety at Work c) Employee Educational Level Physical Working Conditions and Occupational Safety . . . . . . . Temperature . . . . . . . Lighting . . . . . . . . . Noise. . . . . . . . . . . Accident Record Keeping System and Reporting . . . . . . . . . . . . Machine Guarding, Safety Devices and Controls. . . . . . . . . . . Status of the Occupational Accident Research. . . . . . . . . . . . . IV 0 STIIDY FINDINGS O o o o o o o o o o a 0 Introduction. . . . . . . . . . . . Specific Background Data and Statistics. . . . . . . . . . . . Testing Of Hypotheses . . . . . . . Hypothesis #1 . . . . . . . . Hypothesis #2 . . . . . . . . Hypothesis #3 . . . . . . . . Hypothesis #4 . . . . . . . . Hypothesis #5 . . . . . . . . Hypothesis #6 . . . . . . . . Hypothesis #7 . . . . . . . . Hypothesis #8 . . . . . . . . Hypothesis #9 . . . . . . . . Hypothesis #10 . . . . . . . . Hypothesis #11 . . . . . . . . Hypothesis #12 . . . . . . . . Hypothesis #13 . . . . . . . . Hypothesis #14 . . . . . . . . Hypothesis #15 . . . . . . . . OTHER FINDINGS OF THE STUDY . . . . . Summary . . . . . . . . . . . . . . V. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS. Summary . . . . . . . . . . . . . . Page 84 87 91 93 96 99 103 106 106 107 118 118 123 125 126 129 130 132 134 136 138 143 145 146 148 150 156 164 165 165 Chapter Page Some Conclusions. . . . . . . . . . . . . 172 Some Recommendations. . . . . . . . . . . 174 BIBLIOGRAPHICAL ESSAY. . . . . . . . . . . . . . . . 178 APPENDICES A. Format and Nature of Letters Sent to the Director of Bureau of Safety Regulations, Michigan-State Department of Labor. . . . 187 B. Letter Sent to the Companies as Initial and First Contact . . . . . . . . . . . . 188 C. Questionnaire Used as Interview Guide . . . 189 D. Letter of Authorization Obtained from the C0panies. . . . . . . . . . . . . . . 204 E. Statistical Tests Used for the Analysis Of Data 0 O O O O O O O O O O O O O O O O 205 vi Table I-l II-l II-2 II-3 III-1 III—2 III-3 IV-1 IV-2 IV-3 IV-4 IV-5 LIST OF TABLES Fatalities and Disabling Injuries from all Work Accidents, 1964-1969 . . . . . Total Cost of Work Accidents for the Six Year Period of 1964-1969. . . . . . . . Types of Industries Randomly Selected, from Which Eleven Matched-Pairs of Industrial Firms Were Chosen to be Studied . . . . . . . . . . . . . . . . Geographical Location of the Companies Studied in Alphabetical Order and Their Relative Distance from East Lansing . . Title and Number of Persons Participated Entirely or Partially in the Interview . . . . .'. . . . . . . . . . Distribution of Accident Research Liter— ature by Accident Category and Publication Date. . . . . . . . . . . . Injury Rates Per Million Man-House Worked and Company Size. . . . . . . . . . . . Relative Effect of the Temperature and Air on Physical Work. . . . . . . . . . Average Number of Employees at the Employ of Companies Studied by Industry. . . . The Work Accident Frequency Rates of the Firms studied 0 O C O O O O O O O O The Work Accident Severity Rates of the Firms Studied . . . . . . . . . . . . . Membership in the Safety Organization . . Percentage of Production Automated in the Firms Studied. . . . . . . . . . vii Page 23 24 29 46 49 90 108 110 112 114 115 Table IV-6 IV-7 IV-8 IV-9 IV-lO IV-ll IV-12 IV-l3 IV-l4 IV—15 IV-l6 IV-l7 IV-18 Organizational Units Responsible for the Safety of the Companies . . . . . . . . . T0p Management's Support and Interest in Safety and How it Differs in Firm's With Low and High Accident Rates. . . . . The Total Score of the Firms on T0p Management's Support and Involve- ment in Safety. . . . . . . . . . . . . . The Average Educational Level of the Workers in the Firms Studied. . . . . . . The Average Age of Employees in the Companies Studied . . . . . . . . . . . . Company Scores on Providing Recreational Programs and Facilities for Their Employees . . . . . . . . . . . . . Company Scores on Promoting Safety Through Employee's Family . . . . . . . . . . . . Company Scores with Respect to Their Safety Rules. 0 O O O O O O O O O O O 0 0 Percent of the Total Employees Who Were Married in the Firms Studied. . . . . . . The Average Length of Employees' Service in the Companies Studied. . . . . . . . . Establishment Date of the Companies StUdied O O O O O O O O O O O O O O O O 0 Accident Costs Included in the Company Accident Record Keeping System. . . . . . The Companies' Scores on Accident Record Keeping System and TOP Management's Involvement in Safety . . . . . . . . . . viii Page 117 120 122 124 126 128 130 132 134 136 138 140 142 Table IV-l9 IV-20 IV-21 IV-22 IV-23 IV-24 IV-25 IV-26 IV-27 IV—28 The Average Number of Employees Under Direct Supervision of Each First Line Supervisor . Total Score of the Firms With Respect to Physical Workshop Conditions. Actual Points Received by the Firms With Respect to Maching Guarding, Safety Devices and Controls. Work Accident Experience of the Firms After Safety Committees were Established How Firms Scored with Regards to Safety Committees . Types of the Safety Committees in the Firms Studied Ranking of the Causes of Unsafe Employee Frequency of Occurrence of Different Types of Unsafe Acts. Factors Being Least and Most Emphasized in Promoting Safety . Evaluation of the Effect of Safety Inspection on Accident Prevention. Interviews' LIST OF FIGURES Figure Page III-l Relative Interest in Three Specific Accident Research Areas Compared with Total Accident Research. . . . . . . 47 III-2 Fatalities From Industrial Accidents as a Function of Age. . . . . . . . . . . 58 III-3 Injuries From Industrial Accidents as a Function of Age. . . . . . . . . . . 58 III-4 Decline in Accident Rate as a Function of Length of Service. . . . . . . . . . . 59 III—5 Illustration of the New Traffic Lights in Montreal (1967) . . . . . . . . . . . 86 III-6 Accident Frequency Rate Variations as a Consequence of Change in WorkshOp Temperature . . . . . . . . . . . . . . . 88 CHAPTER I INTRODUCTION Discussion Even though considerable progress has been made over the past half a century in preventing occupational accidents, annual figures indicating frequency1 and severity2 rates of work injuries and their resulting costs, still represent a serious problem for managers of today's business concerns. Achievement of business goals and the maintaining of a healthy and prosperous organization will be impossible unless managers start and lead an active campaign against accidents by fully utilizing every effective resource avail- able to them in order to minimize human and economic losses resulting from work accidents. Perhaps it will be imprac- tical -- if not impossible —- to completely eliminate occu- pational accidents, but even if it were possible and/or ‘ . lFrequency rate is total number of disabling in- Juries per million employee hours of exposure. . 2Severity rate is total days lost or charged per mlllion employee hours of exposure. practical, managerial campaign would have to be continued in order to keep the accident record at that minimum level. Thus prevention of industrial accidents and achievement of occupational safety objectives is a continuous and endless challenge for managers. Many suggestions have been put forward on the industrial injury problem, including stricter legislation, a more intensified educational approach, and financial inducements. Yet none of these can successfully achieve their end without a management's earnest co-0pera- tion. It is on them that the onus rests. It should also be recognized that accident prevention is a joint problem for government, business and industrial organizations, insurance companies, employers, employees and other agencies. In other words, occupational safety is everybody's business, but to the man who sits in the manager's chair, it becomes a compelling responsibility, for safety takes account of the experience that what is everybody's business usually turns out to be nobody's business. It should be obvious that producing more products with lower costs and better quality is an ideal achievement for an industiral concern, but as Henry Ford and other 3 industrial and business entrepreneurs, and pioneers have 3National Safety Council, "50 Years Ago ... What They Were Saying About Safety," National Safety News, Nov 1969. 3 looked at the subject matter, production without safety is inefficient. Accident prevention is absolutely an essen- tial part of the controlling of the industrial Operations. Minimizing industrial accidents is a significant means of reducing costs. If a business concern is proud of providing goods and rendering services to the community to which it belongs, it should also consider the fact that no greater service to humanity can be given than that which seeks to prevent accidents and occupational diseases. The National Safety Council reports4 that between 1912 and 1969, accidental work deaths per 100,000 pOpula- tion were reduced 67 percent from 21 to 7. This is an indi- cation of a steady and impressive progress which has been made in reducing work injury rates. But as the challenge continues to further reduce occupational injury rates, it becomes increasingly difficult to maintain the pace of yesterday's steady achievement. Today companies are gen- erally more safety oriented than in the past, but many show little effect of the safety movement. Despite tremendous achievements in the way of pre— venting occupational accidents, for the individual employee who as a result of today's work injury suffers from 4National Safety Council, Accident Facts (Chicago, Illinois, 1970), p. 83. 4 impairment and economic loss, these achievements do not mean much and, in fact, he may even feel much worse off and more unfortunate than his fellow worker who suffered from the same work injury half a century ago. While the worker is almost daily witnessing scientific achievements of un— paralleled magnitude, he sees no corresponding improvement in his immediate work environment, while he watches lunar— orbiting astronauts and is confident of their safe return, his assurance of a safe return home from the job has not kept pace. What is required is a basic reorientation of societal interests, backed by social concern and encourage- ment, the space program received the requisite financing and manpower to accomplish tasks of unforeseeable complexity. Yet, lack of interest in the simpler problems of occupa- tional safety has left a discouraging scar of human and economic losses on the face of such progress. Illustrative of this dearth of interest is the obviously tremendous im- balance between, for instance, concern for sporting events and for occupational safety. It is not difficult at all for anybody who wishes to observe this pr0position for him— self in every day life. It was surprised to notice that except in a very few cases, I could not meet anybody in the University, my residential area, or different companies 5 studied that had a fair knowledge about the magnitude of accident toll, especially occupational accidents. Among twenty—five students in the MGT 413 class that I taught in fall term 1970, no one had seen nor heard about "Accident Facts" which is published annually by the National Safety Council. Of course twenty—five students are not too many but if we consider this as a sample, it may very well repre- sent the majority of peOple in this educational society. Anybody can easily find out for himself the validity of this statement. He can simply ask the peOple whom he knows, he sees, he meets, and he will be surprised to find even one person who has seen and has looked at the content of "Accident Facts". It seems to me that "more than fourteen thousand fatalities" and more than two million disabling injuries at work, in a year, should be considered very important and should be given fair attention by everybody in the society. Why not follow safety records, strive, support and urge pe0ple and agencies involved to get the first place in the occupational safety race and accident prevention game. This interest and concern must be taught and promoted continu- CNJsly as has been, still is and will be done in the case <31? sports and society's other favorite events. It is my true belief that work accidents are absolutely unnecessary 6 and can and should be prevented. As time passes, and tech— nology advances, it may be an even more difficult task to further reduce fatal and disabling injuries and thus society's concern with safety must be more active and cre- ative than ever for technology is not going to slow down and wait for safety to catch up. There is a lot to be learned and discovered in the field of occupational acci- dents and the management of safety in work environment. It is one of the long recognized and obvious principles of war that a successful fight against any enemy requires perfect knowledge and information about him, and yet such a know- ledge is not available about work accidents. Thus, fight- ing against this enemy will result in failure unless neces— sary knowledge is collected and fighting strategies are reoriented and based on these facts. It is in recognition of this fact that this study is being undertaken. Nature of the Problem According to estimates reported by the National Safety Council,5 occupational accidents claim more than 14,000 lives and cause over 2,000,000 disabling injuries in the United States each year. Although continued efforts gradually have reduced accidents and their seriousness, 5Accident Facts, Op. cit., p. 23. 7 the improvement rate -- as predicted by one of the experts in the field of Safety Management6 —- has been decreasing in recent years. Table I-l indicates the magnitude of the work accident toll for the six year period of 1964 to 1969.7 TABLE I-l FATALITIES AND DISABLING INJURIES FROM ALL WORK ACCIDENTS 1964-19693 Total Total Deaths Total Disabling Time Lost Number of per 100,000 Year Deaths Injuries Man-Days Workers Workers 1964 14,200 2,050,000 235,000,000 67,600,000 21 1965 14,100 2,100,000 235,000,000 69,700,000 20 1966 14,500 2,200,000 255,000,000 72,600,000 20 1967 14,200 2,200,000 245,000,000 74,700,000 19 1968 14,300 2,200,000 245,000,000 76,900,000 19 1969 14,200 2,200,000 245,000,000 79,000,000 18 Even though the total number of deaths and disabling in- juries are steady over the period covered by Table I-l, 6Rollin H. Simonds, John V. Grimaldi, Safety Manage- ment (Homewood, Illinois: Richard D. Irwin, Inc., Revised Edition, 1963), p. 3. 7These figures have been collected and tabulated as shown in this table from 1965, 66, 67, 68, 69 and 1970 edition of the Accident Facts, pp. 23-39, published annually by the National Safety Council, Chicago, Illinois. 81bid., p. 25. 8 because of an increase in total number of workers, the death and disabling injury rate per 100,000 workers have been gradually decreasing. However, the cost of accidents as shown in Table I-2, estimated by the National Safety Council, for the same period has been increasing at a much higher rate. These figures do not include the value of prOperty damage in all accidents, except direct losses in fires. TABLE I-2 TOTAL COST OF WORK.ACCIDENTS FOR HE SIX YEAR PERIOD OF 1964-1969 % Increase From Cost per Worker Year Total Cost of Prior Year to Industry 1964 $5,200,000,000 + 4% $75 1965 6,800,000,000 + 7% 90 1966 6,800,000,000 + 6% 90 1967 7,300,000,000 + 7% 95 1968 7,900,000,000 + 8% 100 1969 9,000,000,000 + 9% 110 91bid., p. 25. There is no agreement among safety men with respect to what is causing the work accident cost boost, but insur— ance companies which pick up a growing tab, present two causes for the cost gains. First, the area of new chemical materials alone introduce 4,000 new compounds a year, many of them toxic. They include radioactive material, plasma, ore, super speed cutting tools, and toxic metals like beryllium, which are in growing use. The second reason given is that even though accidents are fewer, their 10 Obviously the direct effect of the severity is higher. continuous rise in the wage level and medical and workmen's compensation costs cannot be ignored when one tries to find an answer to the question of "why accident costs rise"? Of course, there are some other factors that play an important role in these cost gains —— like worker's greater awareness of labor-protecting laws and workmen's compensa- tion provisions and so on -- that could be added to the list of causes. Nevertheless, the major purpose of this section is to indicate the magnitude of human and economic losses resulting from work accidents so that the problem can be felt and recognized by: (1) those Who have the 10For further discussion see: "Rise in Plant Accident Costs Spurs Industry Safety Drive," Special Report, Iron Age, Vol. 188, No. 18, Nov. 2, 1961, p. 41. 10 capability of conducting a scientific inquiry to further discover unknown areas of work accidents; (2) those who can do something about preventing these unnecessary mishaps, namely those who design machinery, make, install and Oper- ate it: (3) those who provide the physical as well as the psychological environment in which these accidents happen; (4) and, finally, those who ultimately pay for these acci- dents -- namely end-users or consumers. So far, figures that have been given were at the National level which include Michigan as well as all other states in the union. However, Michigan, being one of the largest industrial states has contributed to these national figures, and more specifically the Michigan State Department of Labor Annual Report11 (fiscal year 1968-69) indicates that work load of the bureau of workmen's compensation for the fiscal year 1969-70 is: 53,365 injured workers and 376 fatalities. These are only the compensable injuries that were reported to the Department of Labor and the actual number of first aid cases, doctors cases, and "lost time" injuries that are not compensable -- namely cases that cause less than 7 days lost time and no permanent disability -- is much greater than those reported here. After all, 11Michigan State Department of Labor, Annual Report - Fiscal Year 1968-69, Vol. 4, 1969, p. 53. 11 what should these figures and numbers mean to any one who is exposed to them? What kind of message is supposed to be conveyed to a reader or a listener by these facts? I believe that the best possible answer for these and any other similar questions has been expressed by Professor Simonds in the following words:12 "These facts show that occupational accidents present a problem that qualifies eminently on two counts as one warranting study and signifi- cant constructive action in business. It in- volves consequences of great importance, and it is something over which it is possible and practical for man to exercise control." Purpose of the Study Occupational safety studies and work accident re- ports have revealed that some industries have higher acci- dent rates than others. Yet, it has long been recognized that not only do different industries have different acci- dent frequency and severity rates, but also within the same industry firms experience widely different work injury rates. The purpose of this research study is to conduct an inquiry into some of the factors that might eXplain why similar firms in the same industry experience different occupational accident rates, and ultimately to make some useful generalizations about the significance of the role 12Simonds, Op. cit., p. 4. 12 that some of these factors might play in causing different (lower or higher) work accident rates. Having this know- ledge will be helpful in winning the campaign against in— dustrial accidents and eventually to further prevent the occurrence of human and economic losses. In the belief that certain factors of managerial attitude toward safety, physical working conditions, char- acteristics of employees (with respect to age, marital status, number of years spent with the company, attainment of formal education) first line supervisor's relative span of control, accident record keeping procedures, quality and quantity of safety and control devices on machinery: recre- ational facilities and programs for employees; availability and provisions of medical personnel and facilities, relative age Of equipment and machinery in the firms, safety rules, interest of employee families in over all company programs, personalized safety training programs, and company age may be significant in the causation of different accident eXpe- rience in different firms, the writer formulated the follow- ing hypotheses concerning correlates of accidents and subjected each to an experimental design for testing. The hypotheses were: 1. In the firms with lower work injury frequency and severity rates, tOp management is highly l3 interested and involved in the company's over- all safety program and actively participates in and supports safety activity. In the firms with better injury frequency and severity rates, workers have higher average attainment of formal education than in firms with relatively poor work injury recOrds. In the firms with better safety records, average age of the employees is higher than the firms having relatively high accident rates. Firms with better safety records have provided recreational programs and facilities for the employees, and by doing so, have helped to bring employees together. This, in turn, has contributed to the establishment of friendly relationships among them and created a favor— able work environment in which employees may be more safety conscious. Firms with low work injury recordshave tried to promote employees' safety interest through their families. Firms with low work injury experience have established comprehensive safety rules covering all Operations, have made sure that all employees 10. 11. 12. 14 understand them and have consistently enforced them. Average number of married employees is higher in the firms with better safety records than firms with poor safety records. Average number of years spent with the company is higher for employees in the firms with low work injury records than for employees in the firms with high work accident records. Older companies have lower work injury fre- quency and severity rates than relatively newly established companies. Occupational accident record keeping systems in the firms with better safety records are considerably more comprehensive and efficient than in firms with poor safety records. Relative span of control for every first line supervisor is wider in the firms with high work injury recordsthan in firms with low work injury records. Firms with newer machinery and equipment expe- rience lower work injury frequency and severity rates than firms with older machinery and equipment. 15 13. In the firms with good occupational safety records, physical working conditions such as relative roominess, lighting, visibility, ventilation are better than firms with poor safety records. 14. There are more and better safety controls and de— vices on machinery in the firms with better safety records than the firms with poor safety records. 15. Firms with better safety records have established safety committees through which unions and/or employees aid and advise management on matters of worker safety whereas in the firms with poor safety records such committees have not been established. Summary It was emphasized in this chapter that despite consid- erable progress made over the past half a century in the field 0f occupational accident prevention, annual work injury and fatality figures still represent a serious challenge and .problem for managers. Ever increasing costs of accidents were discussed and estimated cost figures for the period of the past six years were tabulated. The purpose of study was dePicted and a number of hypotheses were formed to be empiri— cally tested. CHAPTER II METHOD OF STUDY Introduction This chapter is an attempt to explain the general approach taken in conducting this research project. The hypotheses stated in Chapter I were used as a basis to determine: (1) what data collection techniques should be used to eliminate the gathering of useless information while insuring that relevant data would be collected and that initially unforeseen requirements could later be made; (2) what kind of sampling method must be employed in order to prevent possible sampling errors and to enable the writer to select those industrial concerns which best represent the kind of firms assumed in formulating hypotheses; and (3) what kind of data analysis should be used to test hypotheses, so that evidence can be developed to support generalizations to be made as a result of findings of the Study. The research embodied: (a) Sampling, (b) Data cOllezction; (c) Data classification; (d) Analysis of data; (e) aand interpretation of data which will be discussed in 16 17 this chapter and in the same order. SamplingpMethod Since the purpose of the research was to identify some of the key factors that cause apparently similar busi- ness firms in the same industry to have markedly different accident experience, samples selected had to be a number of matched-pairs of industrial concerns. Each of these pairs was comprised of firms of similar size in the same industry, K with one of the two having considerably more work injuries than the other. At least two pieces of data and information were needed in order to select a proper sample for the pur- pose of testing the hypotheses of this research study. These were: a. Information about availability, number, and location of firms of similar size in the same industry in the State of Michigan; b. Actual work accident records of these firms in order to match the firm with a very good safety record with the firm with a relatively poor safety record in the same industry. For this study twenty-two industrial firms in the Statezof Michigan were selected. These constituted eleven ma’CChed-pairs of similar size firms in eleven different industries but with one of the firms in each pair having a —— _!-4 18 relatively much better occupational accident record than the other firm. Relatively small size industrial firms were selected for study because existing knowledge in this field strongly indicates that smaller size industrial concerns have a higher rate of work accidents than larger ones.1 By smaller companies, for operational purposes of this research study, it is meant the firms which employ between 80 and 650 employees. Having access to information concerning availability, number, location and other general characteristics (namely, name of company president and officers, number of employees, address, kind of product and so on) of the companies repre- sented no major problem to this writer since it could be gathered from different available directories.2 But the major question before any other step could be taken toward sampling, was how and from whom to get the actual work injury records of the firms to be included in the sample. Selecting any two firms of similar size in the same industry in the State of Michigan for the purpose of this study would prove 1For reasons given and more detailed discussion see: Chapter III of this work where the subject is elaborated. 2The following directory was most helpful to cOllect this kind of information: Michigan Manufacturers and IFinancial Record, The Directory of Michigan Manufac- I§E§§E§_(Detroit: Manufacturing Publishing Co. 1969). 19 to be a useless effort unless there was a way of determining their annual occupational accident rates and of knowing that two firms selected in each pair have markedly different accident experience from each other. Search for a publication or reference that carried the actual work injury rates of individual firms, by name, in the State of Michigan revealed that such a publication or reference does not exist. However, I was able to locate where such information might be available and filed. According to the Michigan State Occupational Safety Standard Act,3 all employers should submit annual reports of all disabling work injuries to the Michigan State Depart- ment of Labor. My initial contact with the Department of 4 from Professor Simonds, Labor started through a letter chairman of my dissertation committee to the Director, Bureau of Safety Regulations in which the Bureau was asked 3Section 15.(1) of Michigan State Occupational Safety Standards Act, Act 282, 1967, p. 573:Imd. Eff. Aug. 1, stipulates that: "To assure the availability of accu- rate, timely statistical data concerning occupational safety, all employers having one or more employees simul- taneously employed shall submit annual reports on a firm and in a manner prescribed by the director, of all disabling work injuries as defined and in accordance with the "Standard method of recording and measuring work injury eXperience" (216.1, latest edition) of the United States of America Standards Institute. Reports shall not call for employees to be identified by name." 4See "Appendix A" for Format and Nature of this letter. 20 to provide me suitable names of companies to be included in the sample. It was also mentioned in that letter that I would probably need more than two names in each industry to start because some would likely not find it convenient to let me get the information. Later, the Director of Bureau of Safety regulations invited ProfeSsor Simonds and the writer to a meeting during which he informed us that because of the provisions of the Occupational Safety Standards Act5 the Bureau could not release actual work injury information of individual firms by name unless prior permission had been obtained from the employer in writing. However, the Bureau provided only the names of 4 to 7 companies in each of eleven different industries which were selected randomly from the list of industries used by the Department of Labor. The list did not include work injury information of the firm nor did it indicate which firm had a better record than the other. Nevertheless, this list of names had an advantage over any other publicly available list. I was sure that these firms would have different work injury experience and record. The next step was to gather general information 5Section 15. (4) of the Act stipulate: "The sta- tistical summary shall not be prepared, published or released in such a manner as to disclose information con— cerning any individual employer unless prior permission has been obtained from the employer in writing":... 21 about these firms concerning number of employees, address, name of company president and the kind of product, which was done by referring to different directories. Especially, the DirectGQ{of.Michigan Manufacturers was heavily consulted. The third step was to write a letter6 to the presidents or in some cases to the executive vice—presidents of these companies to eXplain the purpose of this research study and to gain permission to visit the company and interview some of their peOple who are involved in safety. One week after mailing of the letters, a telephone contact was made with each of these individuals. If their response was positive, they were asked, in addition to the above requests, to pro- vide me a letter of authorization in order to have access to their actual work injury records in the company and in the Department of Labor. Eleven pairs of companies that were most similar in number of employees, kind of product and, if possible in location, but not in work injury records were selected. At first, I was not sure how many of the firms would agree to participate in this research project and from what I had observed and heard from other candidates in this stage of their program, I expected a high probability 6See "Appendix B" for the format and nature of the letter. 22 of being turned down by a majority of these companies.) This was especially true since the subject of occupational acci- dents is often touchy. This could represent a serious difficulty -- if it were true —- since the number of firms that the Department of Labor had provided were few, and if the majority of them were not willing to participate, the problem of finding adequate pairs of firms to be matched and of pursuing the study the way it had been designed, would be impractical. In order to test the effectiveness of the strategy taken -- so advised by Professor Simonds -- I decided to contact first only six companies in three different indus- tries. Letters were sent to these six companies. When I called the persons to whom the letters were addressed, a week or so later, I was astonished to receive six positive and very enthusiastic responses. This experience confirmed the effectiveness of the strategy, so I forwarded letters to other companies. Overall, I contacted twenty—six com- panies of which only four could not participate in the project for obvious reasons which were communicated to me through very nice and polite letters from these four com- panies. Table II-l shows eleven different industries which ‘Mere randomly selected and from which eleven matched-pairs (Df industrial firms were chosen to be studied. Industry 23 Groups Classification as shown in this table conforms to the definition of the Standard Industrial Classification Manual7 which are also used in the Michigan State Depart- ment of Labor for industrial accident analysis and record keeping purposes. TABLE II-l TYPES OF INDUSTRIES RANDOMLY SELECTED, FROM WHICH ELEVEN MATCHED PAIRS OF INDUSTRIAL FIRMS WERE CHOSEN TO BE STUDIED Industry Standard Industrial Number Classification Code Industry 1 201 Meat products 2 202 Dairy products 3 203 Canning, preserving 4 251 Household furniture 5 i 265 Containers and boxes 6 332 Iron and steel foundries 7 335 Non-ferrous metal, rolling, etc. 8 339 Other primary metal products 9 346 Metal stampings 10 354 Metal working machinery 11 371 Motor vehicles and equipment 7Office of Statistical Standards, U.S. Bureau of the Budget; Standard Industrial Classification Manual, 1967 Edition. 24 Table II-2 shows the location of these companies studied and their relative distance from East Lansing. Dis- tances shown in this table are not official and were taken by this writer through the observation of odometer readings. TABLE II-2 GEOGRAPHICAL LOCATION OF THE COMPANIES STUDIED IN ALPHABETICAL ORDER AND THEIR RELATIVE DISTANCE FROM EAST LANSING Distance from Number of Location East Lansing r the Firms (A11 in Michigan) (one way) in Miles 1 Belding 75 1 Berrien Springs 120 1 Bridgeport 70 1 Chelsea 44 l Chesaning 45 2 Detroit 85 1 East Jordan 225 1 Grand Haven 96 2 Grand Rapids 64 1 Hamtramck 80 1 Inkster 75 1 Imlay City 82 5 Lansing & vicinity 4 (on the average) 1 Mason 15 1 Muskegon 102 1 Sparta 80 25 Method of Collecting Data For the collection of data, two general approaches were taken. First, an attempt was made to gather necessary information from secondary sources, namely publications of Federal, State and Local Governments, private institutions and findings of other researchers conducted in this field or any other field from which findings could be useful in test- hmghypotheses of this research project. But, because of the rather confidential nature of statistics pertaining to the work injury records of industrial firms, these secondary sources provided very little data needed, and a lot had to be collected originally from the companies studied. Three alternatives were available that the writer could use in collecting data. Those three alternatives were: (1) con- ducting a personal interview; (2) mailing out a question- naire and (3) telephoning. Even though a larger number of peOple could be reached through the use of the second and third alternatives, conducting a personal interview was pre- ferred. This method of collecting data was employed not because it is the most flexible means of obtaining data,8 nor because non-response could be easily reduced to none, 8J'ulian L. Simon, Basic Research in Social Science - The Art of Empirical Investigation (New York: Random House, Co., 1969), p. 253. 2. j nor because the identity of respondent is known to researcher, but for two stronger reasons. One was the nature of some of the hypotheses formulated. In order to collect useful data pertaining to most of the hypotheses, I had to personally visit the company. An example of this would be the case of physical conditions of the plant with regard to such items : as roominess, visibility, temperature, noise, lighting and so on, which could not be fulfilled through use of other _ - methods of collecting data. Another important reason was to Q have the opportunity to examine company accident records and data. Using this technique of data collection gave me a fair chance to collect some very important vieWpoints and also to recognize some other factors that practitioners of safety felt might contribute to a high or low accident rate in industrial firms which were not included or intended to be collected in the original plan for this research project. Some of these viewpoints and facts about actual industrial work environment are reported in Chapter IV without inter- pretation, since empirical data was not collected to reject or confirm them, but they could be used in formulating hypotheses for future research projects. The nature of the interview was highly structured, and in order to ensure that every respondent was asked the same question and in the same Order, a questionnaire was —%_¥ 27 deve10ped to be used as an interview guide.9 Some of the questions were of a statistical nature and at the time of telephoning the companies -- a few days after the initial contact letter was sent -- if they said they were partici- pating in the project, I would ask them to prepare statis— tical information about employees, supervisors and the com- pany so that at the time of interviewing these statistics would be ready and little interviewing time would be used in figuring out every single question which was of a statis- tical nature. As examples of those statistical questions, the average number of employee, average age of employees, average number of years spent with the company and so on, can be mentioned. Visiting every company took place in two separate phases. Phase one constituted the interviewing. The inter- view itself was administered in two different stages. Stage one was very structured and the interviewee was restricted to direct answer to direct questions. In stage two the interviewee was asked to feel free to point out any point of view that he felt had some importance to the subject which was not included in the structured interview. They were also asked to give their Opinion on subject matters __ 9For the nature and format of questionnaire (inter- View guide) see: Appendix C. 28 pertaining to safety and to present any solution for the existing problems. In summary, in the first stage the interviewee was asked to answer only those questions the writer had asked, but in the second stage I was a listener and allowed the interviewee to express himself. This en- abled me to be exposed to many interesting viewpoints, some of the more important of which are reported in Chapter IV, under "other findings" of this study. The second phase of my visit was to have a tour in the plant area and observe the actual working environment. For me, this was one of the most beneficial parts of the research, for it enabled me to visit twenty—two different firms in eleven different industries and to observe in real world situations examples of many concepts, like "potential hazard”, "unsafe acts", machine guarding and so on that sound rather intangible in theory. This observation was also necessary for com- paring and evaluating physical working conditions Of two matched firms in each industry in order to provide empiri— cal evidence of the direct relationship between a good safety record and good physical conditions which are ex— plained in Chapter IV. Visiting each company also gave me a chance to meet and talk to different peOple, besides the original inter- viewee which would be almost impossible if I had used other 29 methods of data collection. Table II-3 shows the number and title of individuals who participated in the interview and/or touring. The total number of these persons exceeds twenty-two, because in some companies more than one person participated. TABLE II-3 TITLE AND NUMBER OF PERSONS PARTICIPATED ENTIRELY OR PARTIALLY IN THE INTERVIEW Title Number President of the company 6 Treasurer . 1 Vice-president of the company 3 Production manager 2 Personnel manager or director 8 Employee or industrial relations manager 4 Plant manager, superintendent or engineer 3 Safety director 3 Company nurse ‘_2 Total 32 Upon completion of touring the plant, I returned to the office where the interview toOk place and looked at the company records to collect any missing data. Before leaving 30 each company I requested the authorities to provide me a letter of authorization so that I could have access to actual accident records in the Bureau of Safety Regulation, 10 At first I was rather Michigan State Department of Labor. uncertain that companies would provide me the letter of authorization, but I can happily report that all twenty-two companies responded positively to my request and generously gave authorization immediately upon hearing my request. Now I have twenty-two letters of authorization in my pos- session as an indication and evidence of human generosity and mutual trust and also one of the best rewards that could ever be given to me, which would always remind me of. a remarkable series of contacts. In addition to what has been said about the data collection process, I would like to mention that I have also called some of the companies for some additional infor— mation and they have kindly provided me with whatever piece of data I had asked for. .Apalysis of Data After all necessary data was collected by use of the techniques explained earlier in this chapter, ¥ 10For the nature and format of this letter of authorization, see: Appendix D. 31 application of a suitable statistical method to analyze the collected data constituted the next major step in the pro- cess of this research study. This was necessary so that appr0priate conclusions could be drawn from the data and their statistical significance could be reported. Since the nature of this study was mainly "qualitative" research, descriptive analysis and interpretation of data was also employed to explain the extent, nature and direction of the effect these factors tended to have on industrial accident experience of the firms studied. Selection of prOper statistical tools to be used in the analysis of data, was determined by the nature and type of scale employed in measurement of data which in turn was dictated by the kind of data collected. For measurement of the data mainly two scales, i.e., the ordinal or ranking scale and the interval scales, were used.11 Data measured by these scales had to be analyzed by non-parametric 12 or distribution-free statistical methods. A non-parametric __l 11The ordinal or ranking scale refers to a level of measurement when objects in various categories of scale stand in some kind of relation to the categories. An ex- ample of this relation among classes are: higher, more pre- ferred, more difficult and etc.; The interval scale refers to a.level of measurement when a scale has all the charac— teristics of an ordinal scale; when in addition the dis—‘ tances between any two numbers on the scale are of known size. 12Sidney Siegel, Noneparametric Statistics for the .EEDaVioral Sciences (New York: McGraw-Hill BoOk Co., Inc., 1956) I pp. 29-30. 32 statistical test is a technique which does not make strin- gent assumptions about parameters of the pOpulation from which the sample was drawn. In other words, they do not depend on such assumptions as normality of the pOpulation distribution.l3 Among non-parametric tests available for statistical analysis of data, which would be most apprOpri— ate in the case of two related (matched) samples, the sign test and especially the Wilcoxon matched-pair signed rank test were employed.14 Limitations of the Study It will be advantageous, at this point, to explicitly comment upon some of the major limitations and difficulties encountered in this study so that before the reader is ex- posed to the findings of the research and conclusions drawn from them, he will be well aware of the limitations of this study. This knowledge will assist him to make prOper use generalizations provided by this endeavor. Some of the 13For more detailed discussion in this subject see the following references: (a) Sidney Siegel, op. cit., pp. 30-31; (b) Delbert C. Miller, Handbook of Research .Qesign and Social Measurement (New York: David McKay Company, Inc., 1964), pp. 76-78. 14For more information on these tests see: Appendix E of this work or refer to: (a) Sidney Siegel, Op. cit., pp. 68-83; (b) Allen Wallis & Harry Roberts, .§§atistics: A New Approach (Illinois - Glenco: The Free Press, 1956), p. 598. 33 major limitations can be summarized as follows: Because conducting of personal interviews and com- pany visitations was chosen as the main technique of data collection, a relatively small sample was selected. Sample was composed of eleven matched—pairs of similar size indus- trial firms in eleven different industries in the State of Michigan. Even though it cannot be considered as a limi- tation by nature, but this writer wishes to mention that no grants or funds were available and associated in any form with this study nor did I ask or look for any, instead, personal funds were used, but as was very well put by another researcher,15 the fact that no biased parties had any chance to influence the results or procedure taken, ended up as a net gain. Another limitation to be mentioned is that as far. as accuracy of data is concerned, I had to depend on the data provided by companies and answers given by the inter- viewees to direct questions. Even though according to provisions of the Section 15(1) of the Michigan State Occupational Safety Standard Act, all employers having 1 or more employees simultaneously 15Paul E. Sands, Accident Prevention and Govern- mental Control in the Construction Industry in Michigan and Ohio. Unpuplished doctoral dissertation (East Lansing, Michigan, 1964), p. 40. 34 employed are required to submit annual reports of all dis- abling work injuries, such a report is not being reported to the Michigan State Department of Labor where I was ex- pecting to get most of the data pertaining to the actual accident records of the firms. While frequency rates could be determined for all companies by using available data in the companies, severity rates could not. Because most of the respondents did not feel that severity rate is a very good indicator of the nature of seriousness of accidents, so they did not keep severity records. Some of the reasons given for this were: (a) that for the same disabling in- juries, a markedly variable recovery period is required for different employees; (b) chance factors cause severity rate to fluctuate greatly from year to year and in many cases, such as deaths resulting from heart attacks, the company can exert no control yet finds the severity record greatly effected; (c) for some disabling injuries, different doctors prescribe considerably different periods of treatment: (d) some of the companies bring injured employees to the firm and make them to do anything they can, like dusting company library books, telephoning and so on in order to minimize the number of days lost, which in turn effects severity rate. Since the severity rate becomes increasingly valua 1 large ‘ avail 35 valuable only if it is considered and studied over a fairly large number of years, and because pertinent data was not available for more than one or occasionally two years in the companies, nor in the Department of Labor with regards to number of days lost, frequency rates were used and relied upon in the classification of firms into those with relatively good or poor safety records. It should be recognized that the above limitations are not only associated with this particular study, nor should they greatly effect the results of it but are general to any endeavor of this nature and in this field of study. Summary In this chapter, the general approach taken in con- ducting this research project was explained. After an intro- duction to the method of study, the sampling method was reported. The sample included eleven matched-pairs of similar size firms in eleven different industries in the State of Michigan. It was also said that for collection of data, personal interviews and company visitations were mainly employed and for the analysis of data apprOpriate statistical tests, namely Wilcoxon matched-pairs signed- ranks test and signed test, were chosen. Finally, major limitations encountered in this research were commented on. CHAPTER III REVIEW OF THE PERTINENT LITERATURE Introduction A relatively extensive review of the existing liter- ature on occupational safety and work accident research was inevitable and a must in the course of this study. One reason for this need was that, after the nature of the prob- lem was defined, and the purpose of study determined, then a search in the literature was necessary to make implicit or explicit deductions which would lead to identification of some of the factors that could qualify as candidates for the cause of the stated problem for this research project. Identified factors would be used as a basis to formulate workable hypotheses to be empirically tested so that the significance of the effect of each of these factors on the causation of the problem could be determined and reported. Secondly, before formulated hypotheses could be empirically tested, a review of the literature was needed to: (1) Avoid duplication of efforts that were very well validated; (2) Consider other research findings in the sphere of this study so that reconciliation and coordination 36 between f i :zons, anc anuzrles "fin‘r‘l‘h)"‘ ,. ‘Vle ‘UUL‘ Kit direcj . recently CE ' Y‘ ‘ " \ :13.. 3: pie accident re \‘ 4" ...ed and 37 between findings of this and other related studies could be made; (3) Minimize communication deficiencies, by using --as much as possible —- the same words, expressions, defini- tions, and concepts which were most widely used in other inquiries in this field; (4) Identify areas in which more contributions were needed in order to orient this study in that direction; (5) Include in this research the issues recently developed and questions raised which were not a part of the previous studies; (6) know where occupational accident research presently stands and be aware of trends in this field; (7) And finally, be able to make prOper recommendations for future research in this field. This chapter reviews, in summary, the historical deveIOpment of occupational safety and industrial accident research to the present time with more emphasis on the subjects most directly related to the variables included in this study. gkxzupational Accidents Research Before ggd.After the Industrial Revolution As far as systematic and scientific study is con- cerned, for the period of pre-industrial revolution, to- day's knowledge on occupational accident research is very limited and practically non—existent. However, students of this field of human studies have tried to establish evidences which CO: a tan: ‘ thuhse A. n 1,. V A. “. 38 which could be used as a starting point or a base for fur- ther systematic discoveries. The purpose of those endeavors is not and will not be an attempt to prove that work acci- dents existed even before the industrial revolution —- because it would be a mistake to think that there were few or no injuries1 -- but to determine whether or not these accidents were apparent or not and how they were dealt with in those days. It is obvious that from the very beginning -- wherever and whenever it was -— man has found it necessary to survive in a hostile environment. For survival he has worked for himself or someone else, alone or with others, voluntarily or forced despite his wishes. But at any rate work has provided him a means of survival while at the same time presenting serious dangers to that survival. Today there is no record of the Caveman's work injuries but at the same time it is neither difficult nor unrealistic to say that those who used a stone hammer, un- doubtedly hit their thumb several times or drOpped it on their foot, as is true even today with much more advanced tools and equipment. As man broadened his knowledge about his environment 1Simonds, Op. cit., p. 16. «reek m y. 1311’; 1‘ V. s F? ‘ ‘ lne ..e m ‘1 39 he was able to better protect himself from hazardous work conditions. An example of this would be the case of Roman— Greek miners who would put goat bladders on their faces to protect themselves from the smoke and fumes which existed in the mine.2 Employers, of course, did not provide this primitive personal protective equipment for the workers. Probably the only group of, so called, employees that were provided the best available protective equipment, were soldiers who received helmets and armor from their employ- ers. While professional soldiers were relatively treated very well, from the safety on the job point of view, the remaining work force of those days, especially workers engaged in heavy works like mining, had no protection at all. We know that minerals were known and used to make war material for soldiers, but we also know that there was no steel chisel or drill to break rocks in order to extract the minerals. What a miner did in those days was to light a fire around a rock until it warmed and expanded. By pouring large quantities of cold water on the warmed and expanded rock, he would then make it contract. The sudden change in temperature created strong stress which caused 2Leo Greenberg, A Railing on Your Roof (Israel, Haifa: Publishing House of the Student Association, Technion, Israel Institute of Technology, 1969; Preliminary edition reproduced for educational and research purposes for this writer's personal use.), p. 4. :Iie bode“ very dange ventilatic Considerir '35 the leg I] 40 the boulder to crack. This process does not seem to be very dangerous and hazardous unless we begin to think that it was performed underground where lighting, visibility, ventilation and temperature were absolutely inadequate. Considering what is known today about the toxic prOperties of the lead and mercury which were mined, one writer con- cludes3 that: "it is to be expected that the miners died like the veritable flies." Of course, mining and other heavy works were performed by, so called, slaves, most of whom were caught in the process of wars with other tribes or nations. Not only was there no thought of protecting enemy agents from work hazards but it would even be con- sidered as a betrayal. Because of wasteful usage of this work force, constant replenishment of the slave supply was necessary. This explains, in part, the constant need to conduct wars -- to capture new slaves. Even non—slave workers of those days had no better fate than slaves. Greenberg mentions that it was normal for the wives of the miners of Joachimstaal -- presently in Czechoslovakia -- to be widowed seven times in the course of their lives 4 which, themselves, weren't very long. 3Ibid. 41bid., p. 5. q q e-3tiV€ y 5 formed t? - 1 ulna; cri dRV : .IOH ,- u stri ‘ ~ ’1’.” to uu {a \— or in i 4 11 "‘ a 'vA V rkers t-n M, :E‘JAK ‘ Q DOt a: 41 It would be apprOpriate at this point to examine some of the possible reasons why occupational safety was relatively unimportant, at least to the chroniclers. Firstly, heavy and hazardous work was mostly per- formed by enemy personnel captured as slaves or convicted criminals and there was no need to protect these workers from any work accident. Secondly, there was no organized industrial activity in a large scale as it exists today and major industrial professions (mining, smithing and carpentry) were performed by individuals in relative roominess and with a slow pace of work. Thirdly, tools used by these workers were primitive and simple, and were not terribly hazardous by nature. Fourthly, communication was so limited that the news of work injuries and fatalities could not diffuse beyond small communities, and aggregate statistics for these work accidents were not therefore available in any kind for any given period. Thus, single events occurring in those small villages or similar com— munities could not impress and draw the attention of the peOple. A fifth reason would be that those who engaged in heavy and dangerous occupations were categorized as low class peOple who were not of great concern to those who made and/or wrote the history.5 5Ibid. were an‘. 0.: ‘ . WATCH WC be cone] re'v'Olut: "919 cm tiOf‘. Wag 42 For the same reasons mentioned, there is no empiri- cal research evidence, nor should it be assumed, that there were any kind of provisions or pre—determined arrangements which would regulate worker-employer relationships. It can be concluded that work accident research and occupational safety practices are the consequences of the industrial revolution which is considered as a time when great hazards were created for the working man. The industrial revolu- tion was accompanied by a substantial increase in the fre— quency and severity of occupational accidents. A review of reasons given for this increase is essential since they provided a basis for the study of work accidents and in fact determined its direction. Greenberg6 listed five major reasons for the increase of accident rate, as a result of the industrial revolution. Those reasons can be summarized as follows: 1. Crowding, as a result of high concentration of machines in a small area. Crowding in itself is conducive to the occurrence of accidents. 2. High concentration of human beings in a small working area, with its social and psychological implications which made Operators apt to have accidents. 6Greenberg, Op. cit., p. 6. 43 3. The pace of work was substantially increased. Because steam power was used instead of human energy which made it possible to Operate at any speed and for any length of time that the em- ployer desired. Thus, the number of exposures to possible hazards, per unit of time was greatly increased. 4. The injury causing power was much greater. 5. The simplification of Operations to the extent that almost anyone could be taught to perform the given task, eliminated the requirement of passing through a selective and arduous period of apprenticeship, journeyman and master in the Guild System. As a result of this, many young peOple -- in some cases women and children -- who had no experience, were required to Operate hazardous machines. A closer examination will reveal a similarity be- tween the above listing and some prevailing work accident problem areas of today's industrial concerns. Since the beginning of industrial accident research, these problem areas have been under scrutiny by different researchers and it appears that still a lot has to be discovered if frequency and severity of occupational accidents are to be w.. P I y ~ tn. A\\ ”J an ~.\ .. U. as» .1\ \d« I h \ 44 further minimized. According to Jean Surry7 who recently completed an extensive study in this field, systematic accident research could be said to have been first established with the work of the British Industrial Health Board between the two 8 made a statistical world wars. In 1919 Greenwood and Woods analysis of injuries in a munition factory. Following this study, in 1926, Newbold9 conducted a very broad study of thirteen factories, the findings of which confirmed earlier results. Jean Surry concluded that: "The effects of these two papers on accident research were far-reaching, further in fact than the authors intended." It should be pointed out that these two studies mainly dealt with the concept of accident proneness which still is subject to controversy among students of the field of industrial accidents. Since this concept was not in- cluded, in nor examined by this study, elaboration of the 7Jean Surry, Industrial Accident Research, A Human Engineering_Appraisa1 (Toronto: University of Toronto, 1969). 8M. Greenwoods; H. M. Woods, A Report on the Inci- dence of Industrial Accidents Upon Individuals with Special Reference to Multiple Accidents (London: British Indus- trial Latigue Board, No. 4, 1919). 9E. M. Newbold, A Contribution to the Study of the Human Factor in Causation of Accidents (London: British Industrial Health Research Board, No. 34, 1926). 45 subject has been avoided. Of course, findings of other studies pertinent to the variables of this study will be discussed under related titles in the remaining parts of this chapter. Before specific literature pertinent to the areas covered by this study is examined, it seems to be benefi— cial to look at the relative position of the industrial accidents research compared to other kinds of accidents. Jean Surry's10 findings indicate that in recent years the prOportion of efforts spent on industrial accident research has declined while study of performance influencing factors has flourished. Table 111-1 shows the number of accident research papers published from 1920 to 1966 by accident category. The distribution of accident research papers shown in Table 111-1 reflects the fluctuation in public interest in different categories of accidents. This tabulation may not be free from bias. Jean Surry recognizes this fact and mentions the possible biases which may have influenced the making of this table. How- ever, she argues in her own words: "Despite these causes of bias, it is felt that this ampling of the literature was reasonably representative." 10Surry, Op. cit., p. 17. 46 TABLEtIII-l DISTRIBUTION OF ACCIDENT RESEARCH LITERATURE BY ACCIDENT CATEGORY AND PUBLICATION DATE11 Accident 1920 1930 1940 1950 1960 1965 Total Class 1929 1939 1949 1959 1954 1966 General 1 O 5 11 9 2 28 Industrial 6 11 9 22 19 9 76 Road 1 13 10 44 37 21 126 Aerospace - - 3 4 2 4 13 Child - 1 2 5 7 2 17 Poisoning - - - 4 2 1 7 Home - - - 3 6 1 10 Recreation and sports - - - - 4 - 4 Hospital - - - - 2 1 3 Farm - - - 2 2 - 4 Total 8 25 29 95 9O 41 288 Annual Publication Rate 0.8 2.5 2.9 9.5 18.0 20.5 6.1 By plotting the same information on a Chart, as shown in Figure III-l, it can easily be recognized that after a small start road accidents research has reached a point where it constitutes about 50% of all accident research for the period of 1920-1966. Family (home and Child) accident research has been rising but industrial llAdOpted from: Ibid., p. 150. 47 FIGURE III-1 RELATIVE INTEREST IN THREE SPECIFIC ACCIDENT Specific Accident Research All Accident Research RESEARCH AREAS COMPARED WITH TOTAL ACCIDENT RESEARCHl2 134~ — 102'- .- 100. .1 0.8” ‘ I l 0.6 0.4— “‘* . 002’ .. 0 l l I 1 A 1900 1920 1940 1960 Publication Date 0 Industrial Accident Research * Road Accident Research 0 Family Accident Research accident research has been decreasing in relative importance since it started the interest in accident research in the twenties. This decrease shouldn't be interpreted as lack of interest for industrial accidents in recent years since it may stem, in part, from the fact that early industrial acci- dent studies were most successful in reducing the frequency and especially the severity of accidents. This concludes a brief review of occupational lzIbid., p. 152. acciden examina the re: e in perfoz 11 size i] la ing kn: l W 48 accident research in general which provides a background for examination of literature on specific areas that constitute the remaining part of this Chapter. CompagyL§ize and Work Accidents Records A point was made earlier in Chapter II, that exist- ing knowledge in this field strongly indicates that smaller size industrial firms tend to have a higher accident rates. The interested reader was then referred to Chapter III for possible reasons behind this statement. Before these rea- sons are discussed, the writer wishes to point out that pgp necessarily all smaller size companies have poor safety records. In fact, some of them demonstrate excellent safety performance. This may result from the inherent personal contact between the manager and his employees in a small firm which may tend to keep the accidents down while the middle or large size industrial firm may be too large for personal concern. But generally speaking, the majority of larger firms have attained what appears to be an irreduc- ible minimum as far as accident rates are concerned. It is estimated that 70 percent of the accidents occur in small businesses and firms without adequate safety programs.13 13Simonds, Op. cit., p. 9. 49 A study by McFarlandl4 showed that manufacturing companies employing more than 500 employees, tend to have about a quarter of the accident rate of those with under 50 employ- ees. Table III—2 shows injury rates per million man-hours of all manufacturing and three particular industries. TABLE III-2 INJURY RATES PER MILLION MAN-HOURS WORKED AND COMPANY SIZE15 Particular Industries Number of All Pulp & Employees Manufacturing Paper Electrical Chemical Under 50 12.8 17.0 3.4 7.1 50-100 10.9 12.4 7.7 8.9 100-500 9.6 10.2 4.7 6.4 Over 50 3.6 5.3 1.4 2.4 Why larger firms have lower accident rates, has been explained by different authorities in this field. Professor Simonds' explanation for this phenomena, in his own words is that:16 14Ross A. McFarland, Thengidemiology of Industrial .Accidents (Harvard: School of Public Health, 1965). 15Adapted from: Ibid . l6Simonds, Op. cit., p. 9. 50 "Usually, those large concerns, employing more workers in staff capacities, are in a position to see the economic importance of accident pre- vention and have developed the best safety pro- grams." It can also be argued that in a firm of 30 employees with a frequency rate of, for example, 12.8, the employer wit— nesses one disabling injury a year, whereas the employer of 1000 men with such an injury frequently rate witnesses 30 injuries. It is possible that the fewer injuries per firm, 17 with accident the less concerned the management will be records. The safety performance of small business organizae tions has been a matter of concern for government agencies as well as private or educational institutions. In fact, one of the most comprehensive listings of the factors that could possibly explain why small size firms tend to have higher accident rates than larger ones, is provided by one 18 of the government publications. These factors can be summarized as follows: a. Small firms cannot (or do not) employ full— time safety personnel. b. The executive of small business carries a com- plex load and has no technical staff to assist him. l7Surry, Op. cit., p. 9. 18U. S. Department of Labor, Bureau of Labor Stand- ards, Safety Subjects, Bulletin No. 67, revised 1956 (Wash- ington, D.C.: U.S. Government Printing Office, 1958), p. 11. 51 c. He rarely joins any safety organization or attends any safety meetingscn'conferences. d. Costs of accidents are not known because small firms don't have detailed cost accounting systems. e. As there are few employees, the accident rates must be extraordinarily bad to yield a flow of injuries sufficient to arouse a management immersed in its manifold problems of sales, finance and production. f. Small companies usually can not afford expendi— tures for which immediate and prompt return is not highly expected. g. Small businesses are so great in number that communicating the "gospel of safety" to them is impractical by using the promotional method of attack that has been so widely and effec- tively used with employers of larger companies. Most of the above mentioned factors were confirmed by the findings of this study, since the companies selected were all in the category of small firms. These findings are discussed in Chapter IV, but as an example, it may be mentioned here as well that none of the tOp executives in twenty two firms studied were a member of any Safety Organization. Obviously any large reduction in the industrial accident rates, will have to come mainly from improvement <3f safety performance of the small firms. This is where more research is needed. I believe, it was in recognition of this fact that Professor Simonds, very well aware of this ne size fj C . *— everyti as IePJ CQEPEEE this IE of the 52 this need, advised this writer to include relatively small size firms in the study. Top Management and Safety Performance of the Firm In any organization, the legal responsibility for everything that happens belongs to tOp management, either as represented by a board of directors or as vested in the company president.19 In the case of safety, not only has this responsibility been assigned to t0p management as one of the activities of the organization, but in many cases it has been singled out and emphasized more exPlicitly. This statement can be very easily supported by many evidences provided by the literature on and related to safety. Most important of all would be the reports of actual success in accident prevention in which almost every report attributes this achievement to tOp management's active support and par— ticipation. In a report of "12,000,000 accident free hours", the Visking Company's answer to "How is it done"? was that this outstanding safety record did not result from some Inagic formulas. Rather, it was the result of many efforts —‘ 19Paul Pigors and Charles A. Myers, Personnel lggndnistration, A Point of View and A Method (New York: McGraw-Hill BoOk Co., Fifth Edition, 1965), p. 579. 20National Safety Council, "12,000,000 Accident- Free Hours - How's it Done?", National Safety News, V01. 68' NO. 2' AugUSt 2’ p. 30. ’71 to pret ciples one pri 53 to prevent accidents, underlying those efforts three prin- ciples were identified and reported. Of course, the number one principle was: tOp management's interest and support." Another report of success in reducing accidents coming from Packaging Corp.21 concludes that: "a successful safety pro- gram requires three components: 1) tOp management backing and participation...." It is also emphasized in that report that only when tOp management places the same emphasis on safety as on quality and quantity of production, sales and other manage- ment functions, then safety efforts will be carried out as they must be successful. Another report from Dupont Co. (U.K.)22 indicates that two divisions -- Operation and construction -- at the Company's Maydown plant, have each completed more than one million injury-free hours, since the beginning of 1965. This report also attributes the success to t0p management's participation and support. It elabor— ates that many suggestions have been put forward on the industrial injury problem but none of them can successfully achieve their end without tOp management's earnest COOperation. * 21William H. Stubbs, "How a Mill Safety Program Cuts Injuries," Pulp and Paper, Vol. 43, No. 3, March, 1969, p. 146. 22Special Report, "The Key to Safety," Personnel Magazine (U.K.); Oct. 1967, p. 25. achie piece Rent 50 Oh actua forma Quoti SEIVe of to that: OD \i 54 When occupational safety literature is reviewed, one observation is obvious and inevitable. It is impossible to find any argument or even comment which would deny the positive effect that tOp management's involvement has in achievement of a remarkable safety record. In almost every piece of work done in the field, the need for this involve- ment is either explicitly emphasized or is considered to be so obvious as to need no elaboration. So far, some of the actual reports from practitioners on outstanding safety per— formance of the companies have been mentioned. I believe quoting Professor Simonds on this pafticular subject will oz serve as the best indicator of whatgacademic world thinks of top management's responsibility for safety. He contends that:23 "The Chief executive ... should be accountable for a poor accident record and might well point to a good accident record as one among many of the evidences of the company's success in pro— gressing toward its major goals." In addition to practitioners and academicians, government agencies24 too, have tried to promote the idea that if really good practice in reducing preventable injuries is to be achieved, tOp management must accept full and definite 23Simonds, Op. cit., p. 44, 24As an example see: U.S. Department of Labor, Op. cit., pp. 69-70. share which as t1: occur The f tiOns BIUrE and ( wait 01 H :1 {.1 CL tn f. l rn t-J 55 responsibility and pay considerable attention to the task, just as it does to any other functions of vital importance. In reviewing this part of the literature a good share of attention was applied to the search for material which discussed and/or predicted future executives' role in safety. This writer was curious and interested to find out, as the technology advances, what changes -- if any -- might occur in the management's overall responsibility for safety?, The following words typically represent the kind of predic— tions and expectations which appeared in the recent liter- ature. Management's concern with safety must be more active and creative, for technology is not going to slow down and wait for safety to catch up.25 Influence of Age on the Frequency and Severity of Work Accidents Since the beginning of systematic research on occu—- pational accidents, there have been a number of studies on the nature and extent of the relation between age of workers and work accident rates. Most of these studies have shown that accident rates decline with age since increasing age is 26 usually accompanied by increased experience. A study by 25Joseph R. Shaw, "Safety in Changing World," National Safety News, October, 1969, p. 50. 26Norman R. F. Maier, Psychology in Industpy (Boston: The Riverside Press, 1955), p. 535. Newboll pariso Eering with i kind < advan. rate. gard O f WC ShOW1 [T] l A): 56 Newbold27 in the metal industries revealed that even if com- parisons are made between men of equal experience, but dif- fering in age, accident rates would still tend to decline with increased age. It has also been found that for every kind of industry there is an optimum age point beyond which advancing age is accompanied by an increase in the accident 28 with re- rate. Maier has reviewed studies done by Vernon gard to establishing the safest Optimum age for certain kinds of workers in coal mining. The findings of these studies showed that for coal—face workers, the safest age is between 30 and 39, and for other undergrOund workers it is between 20 and 29. For the transportation industry the Optimum safe age is estimated at 65, beyond which a decline in safety is expected to occur. Findings of a study by 29 attributed this change in accident rates in the Browne, later life of the worker to such accompanyments of age as slight deterioration of eyesight, hearing, and general steadiness and balance. Elderly people commonly have transient attacks of 27E. M. Newbold, A Contribution to the Study_of the Human Factor in the Causation of Accidents; Industrial Health Research Board, No. 34). 28Maier, op. cit., p. 535. 29R. C. Browne, Health in Industry (London: Edward Arnold Publishers LTD). YO‘JI‘. Olde men Wflet to“ 111- aCci 57 dizziness and sometimes are not able to coordinate hand and eye. 30 indicated that Studies summarized by Jean Surry younger employees have the highest injury rate, but the older apparently have as high a death rate from injuries. It is not Clear whether these differences are due to more exposure to hazardous conditions because customarily younger men are assigned to more adventurous or arduous roles or whether in precisely the same role, the younger man tends to make more errors than his elders. Figures III-2 and III-3 illustrate fatalities and injuries from industrial accidents as discussed by Jean Surry. King31 has studied the age - accidents relationship in detail and specifically from three different points of View. He analyzed actual injury data with the intention of showing possible differences in the causes of accidents, parts of body injured, and types of injury with differences in age. His findings indicated that the actual accidents sustained, their causes, their nature (type) and parts of body injured_differ with age. He then suggested that 3OSurry, op. cit., pp. 11—12. 31H. F. King, An Age-Analysis of Some Agricultural Accidents, in William Haddon and et al (eds) Accident Research Methods and Approaches (New York: Harper & Row, 1964), pp. 41-47. Death ra per 100, ezployee InjUry 7- Per 1,00 EZPIOyEE 58 FIGURE III-2 FATALITIES FROM INDUSTRIAL ACCIDENTS AS A FUNCTION OF AGE (Ontario 1965)32 30- 20v Death rate per 100,000 employees 10 - 0"” I l l l O 20 40 60 80 Age FIGURE III-3 INJURIES FROM INDUSTRIAL ACCIDENTS AS A FUNCTION OF ACE33 60- 50- Injury rate 40. per 1,000 employees 30 20 - 10. ’1 Age 32Adapted from: Ibid., p. 12. 33Ibid. V! findin. LL) :1chijt Per 100 E H, ard A 59 different measures of accident prevention are needed for younger and older persons. Length of Service and Occppational Accidents A number of studies have revealed an inverse rela- tionship between the length of service, which implies work experience, and work accidents. A study of occupational accidents in a large industry by Browne34 indicated a very steep fall in the rate of accidents after the first few months of work and a gentle rise towards the end of the working life. Figure III-4 is an illustration of Browne's findings. FIGURE III-4 DECLINE IN ACCIDENT RATE AS A FUNCTION OF LENGTH OF SERVICE 180 14m. Accident rate 100, per 100 employees 60. 20‘ I 1;. "WWI 5 16 26 3O 46 56 Length of Service 34R. C. Browne, Health In Industry (London: Edward Arnold Publishers LTD, 1961), p. 31. 60 A statistical analysis of 227 serious work injuries requiring extended treatment made by McKinnon and others of 35 indicated General Electric Company summarized by DeReamer that lepgth of service is one of the factors influencing the occurrence of accidents which result in injuries. This analysis indicated that workers have more than one and a half times as many injuries during the first year of ser- vice as would be expected. Other studies by VanZelst36; 37 and Hanna have shown that the workers with one Chaney months experience have rates 1% to 2 times their accident rate after six months. And for the first day, accidents appear to be a hundred times as high. Early studies by Fisher and Ghiselli, reviewed by Dr. Simonds,38 have indi— cated a significant relationship between experience and accident rates. Despite the findings of studies reviewed, there have been some other endeavors which have shown 35Russell DeReamer, Modern Safety Practices (New York: John Wiley and Sons, Inc., 1958), pp. 37-39. 36R. H. VanZelst, "Effect of Age and Experience on Accident Rates," Journal of Applied Psychology, Vol. 38, 1954, pp. 313-317. 37L. W. Chaney and H. S. Hanna, The SafetyyMovement 3E9 the Iron and Steel Industry, 1907-1917, U. S. Bureau of Laborer Statistics, No. 34. 38Simonds, Op. cit., p. 404. 111C be Cdn not min: rate haze ...at IECL and 61 increases in work accident rates with increased experience.39 40 this contradiction may But according to Professor Simonds be due to variances in the groups studied. So thorough understanding of the particular work situation under study is a necessity before any generalizations or comparisons can be made. While reviewing the literature, Jean Surry has noticed that high itinerant population of the lumbering, mining, and construction industries have very high accident rates. She contends that of course the inherent higher hazards of these tasks are recognized but it is suggested that more permanency in these industries would considerably reduce rates of accidents.41 From the literature reviewed and summarized, it can be deduced that in an industrial organization with a stable work force accident rates are expected to be found lower due to the fact that this kind of firm has prOportionally more workers with considerable length of service, who are relatively more experienced. An industrial concern with a high turnover rate and unstable employment, will have in its employ more young employees with insufficient or no experience which most likely will constitute the makings of a poor safety 39For reference to these studies see: Ibid. 4OSimonds, Op. cit., p. 404. 41Surry, Op. cit., pp. 11-13. perforr Worker_ 62 performance and records. Worker Participation and Occupational Safety The concept of employee participation has attracted considerable amount of attention from the students and prac- titioners of management and human relations in general as well as of occupational safety management in particular. Keith Davis defines participation as: ... "mental and emotional involvement of a person in a group situation which encourages him to contribute toygroup goals and share responsibility in them".42 In the case of safety, group goals would be prevention of work accidents. Occupational safety is one of the functional areas in organizations in which employee participation has been widely practiced. But there is no general agreement on the consequences nor on the degree and kind of effective worker participation in overall safety programs. The existing literature indicates that more than any other methods of participation such as, suggestion systems, employee Opinion surveys and so on, the safety committee system has been used in practiCing 43 Tworker participation in safety. Davis reports that a large aircraft manufacturing company with a total employee * 42Keith Davis, Human Relations at Work. The Dynamics 0f Organizational Behavior (New York: McGraw-Hill Book Co., 1967), p. 128. 43Ibid. | 0bserv~ records He the' tude ‘ 63 pOpulation of up to 20,000 used a safety committee system in which each department was represented by one of its workers. During a ten year period not one worker had a disabling injury while serving as safety committeeman. He observed a significant difference in safety performance records of workers as committeeman and non-committeeman. He then concluded that part of this difference surely come from the fact that the committeemen were "responsible participating persons with regard to safety. This writer would suggest that only "being responsible" should not be given the full credit for success in a participative system. There must be some other and more important factors which were not ignored by Davis and that influence workers atti- tude toward safety while they are being reSponsible com- mitteemen. In this context, three important ideas can be deduced from the above definition of participation given by Davis. Participation results in ego-involvement which in turn provides partial satisfaction for workers'social and psychological needs. It gives workers a chance to release their own resOurces of initiative and creativity toward safety of the organization. By being in the committee they are in a position and are motivated to contribute to the Situation. The third factor is that participation encour- ages workers to accept responsibility and to become self— invol tion sure: then: they #0 I h .0 “I (I) 64 involved in dealing with the problems brought to the atten- tion of the committee. This sense of self-involvement con- tributes to successful functioning of committees. It in- sures the prOper implementation of decisions, since workers themselves had a share in making them and want to see that they work successfully. It should be recognized that not necessarily all participative activities will result in favorable results. Rather, they are subject to some prerequisites, which have a direct and determinant effect on the success of partici— pation. In general, the following condition must be present if a participative program should work successfully. Some of these conditions must be provided by the individual, and some by the environment in which participation should take place. Presence of these conditions may mean that partici- pation will work better in some situations than others and in certain situations it won't work at all. Tannenbaum44 and his associates have listed the prerequisite conditions for participation. In brief, some of these conditions are as follows: (a) from the cost point of view the value and results of participation should be justifiable, (b) the ‘ 44Robert Tannenbaum, et al., Leadership and .QEQanization: A Behavioral Science Approach (New York: .Mc—Graw-Hill Book Co., 1961), pp. 88-100. parti: and Rf tion I unit 0 for pa should . i,‘ fivhe I" k the 65 participant should have the ability, such as intelligence and knowledge to participate, (C) the subject of participa- tion must be relevant to the participant's organizational unit of interest, (d) mutual communication must be possible for participant, (e) neither party (management or worker) should feel that his position is being threatened by par- ticipation, (f) since participation is hardly apprOpriate in emergencies, its effectiveness is questionable in these situations. There must be enough time to participate before action is required. In addition to the above conditions, there are some other limiting factors that must be considered before par- ticipative methods are adapted. It has been pointed out that today, the complexity of organizations and technology involved make it difficult for workers to participate successfully. Difficulties especially arise when workers make proposals in areas in which they are not competent.45 When these prOposals are rejected, no matter how apprOpriate the alternative courses of action are, they will be subject to worker's resistance. Some workers even build up a habit to be consulted on every issue, and when they are not con- sulted they become resentful and alienated. It is difficult 45Davis, Op. cit., p. 140. to de ticipe tion, partic ticipa Na Rage 66 to determine the amount of interpersonal trust between par- ticipating parties which is prerequisite for Open communica- tion, which in turn is a necessary condition for effective participation. Perhaps the most important issue with par- ticipation is that it provides a means of manipulating em- ployees, not necessarily by management but by union or other groups. I was told by some interviewees that the union sometimes uses safety issues -- in the safety committee meetings or after -- as a base to influence management on other issues. When the other issue is settled, then the union rarely shows interest on the original and unsettled safety question. Despite some obvious merits inherent in participation, not all organizations have had favorable results from safety committee meetings. Quoting Dr. 46 Simonds: "As a means of administration, committees are usually too cumbersome and slow moving to be very effective." Stronger expressions have been used by practitioners. As an example, the following statements may be quoted: "Safety committees are 'for the birds'. The com— mittee defeats the primary purpose of the manage- ment approach. It places responsibility for the safety program in the hands of a random group."47 46Simonds, Op. cit., p. 68. 47Laurence F. Mihlon, "What is Wrong with Safety .Management," Factory 119:91, Sept. 1961, p. 241. O err" On the that 'C! (a / I (‘1 I C1 67 Another writer contends that: "Safety committees (a form of participation) are not acceptable for they specifically emphasize the question of safety and thus prevent it from being seen as a part of normal Operating pro- cedure. On the other hand, the rather stronger position held by others who talk favorably about participation in general and safety committees in particular, leads to the belief that in fact it is a motivating factor and is being suc— cessfully used. Earlier eXperiments by students of man- agement such as Raethlisberger, Bavelas, and Coch and French -- as reviewed by Davis49 -- indicated that even though participation has limitations, but when its pre- requisites are met and the right kind and amount of parti- cipation is selected to match the needs of a particular situation, it offers potential for higher productivity, higher morale and other benefits. Talking about the merits of committees, Dr. Simonds contends that: "They are often an effective means of interesting or educating a large number of peOple in an activity. They may also be a good method for securing co-Operation coordination and an exchange of ideas among peOple not otherwise regularly thrown together:.... 48Evan Stallcup, "A Fresh Look at the Safety Pro- gram; When Enough is too Much," Personnel, Vol. 38, 1961, P. 27. 49For more information seequavis, Op. cit., PP. 133-146. 50Simonds, Op. cit., p. 68. ’24- 68 It seems to me that failure or success of the safety committees depends largely upon the quality of planning for it. Not all types of committees can be used successfully in all organizations. If an organization fails to determine and/or establish the type of safety committee which is best suitable to its needs, discouraging results shouldn't be attributed to ineffectiveness of the technique itself. The literature witnesses various ways of Classifying safety com- mittees. Among those four main types may be identifiable: 1. Labor - management committee - including union representatives,worker representatives and management. 2. Central safety committee - consisting of all department heads - they normally make policy decisions and the frequency of meeting is monthly. 3. Departmental safety committee - including all supervisors and their foremen. Normally this type of committee meets twice a month. 4. Foremen's safety committee - consists of all men under each foremen. Frequency of meeting is normally once a week. Actually, there may be some other kinds of commit- tees such as technical and special or a committee which is a combination of all four types mentioned above. There is no One best way of organizing a safety committee which will insure positive results in every organization. Committees must be tailored to fit each organization's needs. Failure just I make {‘1' impact tional the ex The Pi! and 0c fOreme 69 is inevitable if the organization tries to reorient its safety problems to adapt a particular type of committee just because it worked well elsewhere. It is difficult to make any generalizations based on present knowledge on the impact of employee participation on prevention of occupa- tional accidents. Further research is needed to determine the extent of its effect and to specify its direction. The First Line Supervisory Span of Control and Work Accident Prevention No other issue with regards to accident prevention and occupational safety has been emphasized as much as has foremen's importance.51 Although there is no general agreement on most issues related to safety at work, however, the "foreman's key role" in accident prevention is one of the rare in- stances where students and practitioners of this field would vote favorably and unanimously. Dr. Simonds position on this issue is best expressed with the following words: "The immediate supervisors of the workers, more than any others, are the key men in accident prevention."52 51In this dissertation the "first line supervisor" and "foreman" are used interchangeably. 52Simonds, Op. cit., p. 13. 70 Heinrich, another contributor to the field of accident prevention contends that: "The first line foreman, moreover is in a peculiarly strategic and tremendously important position so far as attaining results in accident prevention is con- cerned....53 In discussing the same subject, DeReamer refers to the study of accident case histories which has revealed why a super- visory approach to accident prevention is too effective. He concludes in his own words, that: "The supervisor is the only person who can control men, machines, and working conditions on a daily, full-time basis. The supervisor is Closest to the person most likely to get hurt. He can take direc— tion action."54 While enough emphasis and consideration has been given to the importance of first line supervision, its limiting com- ponent, namely relative span of effective control is not discussed in the safety literature. It must be recognized that the foreman has limited resources. The number of hours that he can work during a day or week is limited. His physical ability, willingness, energy, knowledge, mental ability and experience to work on many problems at 53H. W. Heinrich, Industrial Accident Prevention (New York: McGraw-Hill Book Co., 1959), Fourth ed., p. 47. 54Russel DeReamer, Modern Safepy Practices (New York: John Wiley & Sons, Inc., 1958), p. 5. ODCG his a :I.‘ (D ’O‘ O I-h r‘f 71 once have limits. Most important of all, he is limited to his attention to only few things at a time. If the number of workers assigned to each foreman exceeds his abilities to supervise attentively then he won't be much of an effec- tive agent in accident prevention. Based on limitations recognized for human beings,. there have been many attempts to determine the maximum number of peOple to be assigned to each supervisor. Prob- ably the most discussed concept in, so called, classical management and organization literature involves the idea of span of supervision. For example, Graicunas, hypothesized that an arithmetic increase in the number of subordinates results in an exponential increase in the relationship.55 He prOposed a mathematical formula to determine the number of these relationships as the number of subordinates in- creases. He identified three kinds of relationship called: a) direct single; b) direct group; and C) cross relation- ships. According to Graicunas, the total number of rela- tionships under each span of supervision can be computed by using the following equation: r = n(2n"1 + n + l) 55V. A. Graicunas, Relationship in Organization, in Luther Gulick and Lyndall Urwick (eds.) papers on The Science of Administration (New York: Institute of Public Administration, 1937), pp. 52—57. where of re Cg. 72 where n = number of subordinates supervised and r = number of relationships. As an example, if the number of workers under each foreman is 12, the number of relationships will be 24,708.56 DevelOpment of this theory was mainly based on the consideration that a foreman not only should supervise individual subordinates but also the interaction among individuals and groups. From the safety and accident pre- vention point of view, this is still true, even if it is labeled as "Classical." Further studies by neo—Classical theorists while continuing to consider the classical con- cept as a basic guide -- have provided evidence to reject the idea that a specific number of subordinates to be effectively supervised can be determined. However, it shouldn't be impractical for any single organization through consideration of the type of work to be carried out, the quality of foreman as well as of workers, and the environmental factors in which supervision takes place, to come up with number of workers that should be assigned to each foreman in that particular concern. Once this 56This concept is discussed in most of the organi- zation and management literature. For more detailed dis- cussion see: a) JOseph Massie, Management Theory in James March, Handbook of Organizations (Chicago: Rand McNally & Co., 1965), pp. 398-399; and b) Rocco Carzo, Jr. and John Yanouzas, Formal Organizations, A Systems Approach (Home- wood, Ill., Richard D. Irwin Inc., and The Dorsey Press, 1967), pp. 44-48, 79—96. ordi: 4.87 there also < Span < Ordin. 73 objective is achieved and proved to be efficient, it is less likely that it will be subject to a radical Change as the organization grows. A recent study by Professor Simonds indicated that from 1958 to 1966 the average number of sub- ordinates reporting to the Chief executive increased from 4.87 to 5.56 which was not a marked increase and neither there was found evidence of such a trend.57 This study also confirmed the idea that there is no one best or ideal span of control in terms of the specific number of sub- ordinates. The concept of span of control still remains one of the main areas of controversy among theorists. Not only is occupational safety literature in particular, almost empty with regard to study of the relationship between supervisory span of control and safety, but empirical research on the concept of "span of control," in general, I 58 also remains scarce. 57Rollin H. Simonds,"Is OrganizatiOn Structure Reflecting New Techniques and Theory?", M.S.U. Business TOpics, Vol. 17, No. 3, Summer, 1969, pp. 65-71. 58Massie, Op. Cit. Effec on J; cover direc study these rience 74 Effect of Some Employee Background Variables on Job Attitude and Safepy Obviously, there are many factors which can be covered under this title. But only the literature most directly related to background variables included in this study is reviewed. It should be pointed out that some of these variables —- namely age, length of service and expe- rience of the workers —- have been separately reviewed earlier in this Chapter. Thus, the remaining background variables are discussed in the following order: a) family background; b) marital status and c) formal education. a) Family background: It has long been recognized that a worker's family background has an important influence in shaping his work values. For example, results of an empirical study showed that if a boy is raised in a materialistic atmosphere, he will view work as a means of obtaining economic and material returns from a given job.59 If the boy grows up in a cohesive family group and in a cultured atmosphere, he is most likely to consider work important for it provides him the Opportunity to make social 59For more detailed discussion see: John F. Kinnane and Martine W. Pable. "Family Background and Work Value Orientation," JOurnal of Counseling Psychology, Vol. 9, Nov. 4 (Winter, 1962), pp. 320-25. and a: 75 and cultural contributions.6O Specific implications of findings of this and other similar studies for the field of occupational safety would be that if management knows how its employees' work values have taken shape then it can most effectively influence them in the desired direction. For example, if a company launches a safety program to promote employees' safety interests, emphasizing economic losses from work injuries would be more effective for a worker who has raised in the former family environment. Of course, for the employee who is the product of the latter type family mentioned above, emphasizing the desire for protec- tion of others, desire to excel and be outstanding and in some cases, the desire for leadership, would be more effec- tive. Costello and Zalkind suggest that the Chief influ— ence on the develOpment of the individual attitudes is his 61 Based on research findings reviewed by these family. writers, they conclude that many specific parental views on the ways they act and think become part of the employee's attitudinal framework. Obviously, attitude toward safety 60Leon C. Megginson, Personnel - A Behavioral Approach to Administration (Homewood, 111., Richard D. Irwin, Inc., 1967), pp. 113—115. 61Timothy W. Costello and Sheldon s. Zalkind, Psychology in Administration - A Research Organization (Englewood Cliffs, N.J.; Prentice-Hall, Inc., 1963), p. 261. CCU; thee. and s ees' desir 76 could not and should not be considered as an exception to these findings. Family background investigation should help organizations to find workers with desired sets of values and should also serve in influencing and modifying employ- ees safety, as well as general work attitude, in the desired direction. b) Marital status and safepy at work: Not only does the parental family life have a considerable share in develOping employee attitudes, but the conjugal family has an even greater influence on the behavior of the employee towards and at the work. As far as safety is concerned, in general, married workers tend to be more safety conscious than unmarried ones. Part of this difference may come from the difference in age since on the average married employees are older than unmarried. But a large portion of difference in behavior of these two groups of employees is attribu- table to concern for and influence of the family of married employees. Based on the findings of a recent study it was reported that the financial demands and responsibilities upon the worker will generally increase -- if not with marriage, then certainly with the coming of Children.62 62John H. Goldthorp et al., The Affluent Worker: Industrial Attitude and Behavior (Cambridge: University Press, 1968), pp. 147-150. 77 The employee realizes that a work injury will result in economic loss and suffering which is absolutely unnecessary. In many cases, work injuries result in permanent impairment which is accompanied by loss of job, decrease in income and many other unanticipated consequences. A research study by Ginzberg showed that during the extended depression of the 1930's when a man lost his job and was put on relief, his whole family's attitude toward him frequently Changed.63 He was often considered less of a man than he had been while employed. Often his children declined to accept his dis- cipline or advice. The incidences such as these indicate the significance of work to a married employee. A study by Wight Bakke reviewed by Carvell,64 has Confirmed the findings by Ginzberg and has revealed that loss of job not only altered employee's economic status, but also affected his social status, family affairs and friendships. These findings may suggest that a stable and continuous employ- ment has more vital importance to a married employee than to a single worker. Being so dear married employees are expected to be more careful about their employment and avoid any incident such as work injury that may cause 63Eli Ginzberg, The Unemployed (New York: Harper & Row 1934). 64Fred J. Carvell, Human Relations in Business (The Macmillan Co., 1970), pp. 83-84. his (1’ answe jobs eight to mal job. liked enthus Karrig lfllOw who h and f fied (I) *1 [U (L 'U "U 78 discontinuance. Earlier studies reviewed by Herzberg and his associates, Showed that 3 studies provided a "yes" answer to the question of "do married workers like their jobs more than single workers?" Only one study said no and eight other studies indicated that marriage does not appear to make a difference in whether or not a worker likes his job.65 A study by Harris,66 indicated that married workers liked their jobs more than unmarried employees. A study of enthusiasm toward teaching, conducted by Chase revealed that married teachers are more enthusiastic than single teachers.67 Inlow studied satisfaction on the job among employed peOple who had graduated from college over a period of 20 years and found that married employees tended to be more satis- 68 fied with their jobs than those who were not married. A recent study provided evidence the amployees who were in- cluded in the study, have taken their present jobs, in 65Fred Herzberg et al., Job Attitudes: Review of ggesearch and Qpinion (Pittsburgh: Psychological Service ‘of Pittsburgh, 1957), pp. 23-24. 66Frank Harris, "The Quantification of an Indus— trial Employee Survey, I. Method," JOurnal of Applied jEsychology, Vol. 33, 1949, pp. 103—111. 67Herzberg, et a1, op. Cit. 68Gail M. Inlow, "Job Satisfaction of Liberal Arts <3raduates," JOurnal of Applied Psychology. Vol. 35, 1951, most tend fell: heh w nific dents Studi the a 79 place of lower paying but intrinsically more rewarding ones, chiefly as a result of feelings of family responsibility or of wifely pressure.69 The same study also revealed that the most "significant others" in the married worker's life will tend to be his wife and his children rather than his mates, fellow unionists or leisure companions.7O Looking at driving behavior and accidents, unmarrieds have been shown to be sig— nificantly more often involved in fatal motor vehicle acci- dents than the marrieds who are similarly exposed.71 These studies suggest that marriage influence to a certain extent the attitudes of peOple. This being the case, any attempt to modify workers' values and attitude in the desired direc- tion should prove positively effective if it is exercised through his immediate family. Davis contends that there is a general agreement that a worker's on-the-job performance, is affected by off-the—job influences and one of the most 72 significant is his family. This influence can be either in the direction desired or undesired by the organization. JDuring World War II, in order to prevent or minimize —_¥ 69Goldthorpe et al., Op. cit., pp. 34-36. 7°Ibid., pp. 148—149. 71Haddon et al., Op. cit., p. 608. 72Davis, Op. cit., p. 347. preo sold to t: more, grati prob: Stud ‘- LOd the 80 preoccupation with family concerns, the families of German soldiers were given strict instructions to avoid references to family deprivations in letters to the front.73 Further— more, the family ties served to keep the army from disinte- gration. Soldiers were warned that disertion (as it was a problem toward the end of the war) would result in severe sanctions being inflicted on the deserter's family.74 The implication of these studies for safety would be that not only may married employees because of concern for their family be more safety cautious, but also their attitude to- ward safety on the job can be influenced by their family. Being aware of the significance of this "family effect" students and practitioners of human relations urge managers to deve10p special approaches to integrate families into 75 Different the organizational communication system. methods have been suggested to achieve this objective, such as mailing of important communications to an employee's family, Open house practices, contests, family picnics and 73Edward A. Shils and Morris Janowitz, Cohesion and Disintegration in the Wehrmacht in World War II, in Walter and Douglas Egan (eds), Readings in Organization Theory — A Behavioral Approach (Boston: Allyn and Bacon, Inc. 1968), p. 311. 74Ibid., p. 312. 75Davis, Op. cit., pp. 347-348. en kn I b. 81 so on. Obviously, more research is needed to draw more specific generalizations about marriage and family influ- ence on safe behavior at work. It appears that existing knowledge provides an encouraging background for it. Employee Educational Level Does educational attainment of workers have a sig- nificant effect upon a firm's productivity and profit- ability? There is no general agreement on the nature of the answer to this question. Findings provided by research studies are heterogeneous and don't allow the formulation of a comprehensive generalization which would be applicable to every organization with any type of work force. However, a review of these findings would reveal the weaknesses and strengths of generalizations made. Some writers and per— sons interested in the subject have hypothesized that there is a direct and positive correlation between educational level of employees and the firm's productivity.76 It should be pointed out that "productivity" is used here in the general meaning of the term which would include safety performance of the firm. For a firm can hardly claim to be profitable and productive if its safety record is poor. A number of studies have been conducted to examine 76See for example: Meggison, Op. cit., p. 158. 51 IE me st CI wil ho an th; “ere 82 the possible relationship between educational level and work attitudes of the employees. Out of 13 studies reviewed by Herzberg and his associates, 5 indicate no difference in job attitude among workers differing in education; 3 studies show an increase in morale with increased education; but the remaining 5 studies reveal that the workers with higher edu- cational level tend to have lower morale.77 It should be mentioned that the type of workers included in these 13 studies varies considerably. A study by Centers and Control showed a drOp in the employee's liking for his job with in- 78 Mann found that blue collar workers creased education. with high education have low job satisfaction. Similar findings also were found and reported by Neilson at Westing- house, Mossin for a group of female sales Clerks, and Scott and Hayes for workers involved in a routine task. The three studies which showed increased morale with increased education level were conducted by the American Vocational Association which sponsored the study of home economic teachers; by Kessler who studied Job Satisfaction of- Veterans rehabilitated under public law 16; and by Scott and Hayes who related retardation in school to job 77Herzberg et al., Op. cit., pp. 15-16. 78For reference to this and other studies referred here see: Ibid., pp. 27-35. (D (I) NC me me It f0; 130: mg! 83 satisfaction and found less retarded workers to be more satisfied. The five studies which found no differences in job satisfaction among workers differing in amount of edu- cation were mostly conducted in industrial concerns. "No difference" findings were reported by: Ash, who studied a large sample of steel company employees, by Quayle, who studied a group of stenographers, Kornhauser and Sharp whose sample was a group of factory girls, Cain who inten- sively studied susceptibility to monotony of factory workers, and, finally, by Fryer who studied a group of men with varied occupations who came to an employment agency for assistance in getting jobs.79 Considering the findings of studies reviewed, it will be clear why general statements about the role of edu- cation in job performance of the workers in general and safety performance in particular cannot be made. It is possible that the difference in findings of these studies may be related to other variables of the samples studied. It has been shown that job satisfaction and safety per- formance goes up with increasing age. Thus, it is very possible that workers with low educational level whose morale was found to be high were also among the older 79Ibid. wor 10% you the rat} 9 Phys K OCCL but eXis V i V E its dang thrc rue: Cans inte acci h Patil 84 workers; on the other hand, workers with high education but low morale and relatively poor performance were among younger workers. It can be deduced that in these studies the determinant variable may be age or length of service rather than educational level. Physical Working Conditions and Occupational Accidents The human body conceived as a system is flexible but a limited system in its interaction with other elements existing in the environment. Obviously the need for sur- vival necessitates this interaction and the same need determines and regulates the extent, duration and kind of flexibility that the subsystems of the human body should exercise in order to allow Optimal interaction and maximize its gains while providing maximum comfort. But if varia- tions in the environmental factors exceed the flexibility limits of the human body, then that survival will be in danger. Interaction with the environment takes place through sensory systems. Any environmental factor inter- rupting proper functioning of these sensory systems may cause a total break down in the system. The concepts of interruption and break down are better known as work accident and disabling injury. In the literature of occu- pational safety, accidents are attributed mainly to unsafe RI. he Q» 85 acts and/or unsafe conditions. But as one writer contends,80 unsafe conditions are to some extent, created by human action or inaction. He elaborates that if an unsafe condi- tion exists but nobody is aware of it, then it can't be removed. And if it is known to a person and he does not take any action to correct the unsafe condition, any possible accident and resulting injury must be attributed to his action or inaction rather than to the unsafe condition. This being the case, this writer would suggest that there are two ways that an unsafe condition can be identified. One way would be careful investigation of accidents after they occur. This may reveal possible unsafe condition which existed before and resulted in the mishap. The second way would be through extensive research to discover limitations of the human body, dangerous prOperties of substances used, etc. Research may begin after occurrence of an accident to scrutinize evidences that could not be revealed by normal investigation. Also, independent from a particular inci- dence, studies may be conducted to discover conditions that are hazardous but not known to human beings. An example of this would be the research done on the field of human vision. 80Don F. Jones, Human Factor - Occupational Safety, A report to the Labour Safety Council of Ontario, Ontario Department of Labor; no date, p. 10. It and ahe; tra: new cal. In fa which 86 It has been found that man can see about 2000 horizontally and 140° vertically when focussed on an object straight ahead.81 Being aware of these findings, peOple involved in traffic safety of Montreal (Canada) have started to install new traffic signals which are horizontal rather than verti- cal. Figure III-4 is an illustration of these new lights. FIGURE III-5 ILLUSTRATION OF NEW TRAFFIC LIGHTS IN MONTREAL (1967)82 \ o 0 Green Red In fact the major aim of the concept of "human engineering" which began during the first world war but continued in earnest as "ergonomics" by a group of scientists during the second world war, is to enable man to work with Optimum physical and mental comfort and use his special senses with best effect. In brief, ergonomics deals with the interaction between the worker, his job and his working 81Surry,,Op. cit., p. 47. 82This illustration is adOpted from: Ibid., p. 49. ent of: ODS ca. hum tak: deal 87 environment. It combines different scientific disciplines of: a) anthrOpometry which supplies the data of anatomy and human measurements; b) physiology, which involves the calorific requirements of work and the functioning of the human body such as reception and processing of stimuli and taking of effective action as a result; C) psychology which deals with much that appertains to working life and safe working; d) engineering that attempts to collate the infor- mation provided by the above mentioned disciplines and build accordingly.83 While human factor engineering remains possibly the most potential part of this field for further develOpment, careful consideration of the previous research findings should prove as effective in prevention of work accidents as will possible future contributions. It seems most appropriate at this time to review the literature on dif- ferent factors of physical factors of physical working environments and their effect on occupational safety. Temperature. It has long been recognized that the temperature affects the efficiency of various human 83For further elaboration on the concept of "ergonomics" see for example: a) A. H. Hands, Ergonomics, in William Handly, (ed.), Industrial Safety Handbook (London: McGraw-Hill, 1969), pp. 307-314. b) Alphonse Chapanis, Man-Machine Engineering (Belmont, Ca1if., Wadsworth Publishing Co., Inc. 1966). ac: ter ICE H1113. wor‘ ill) in u..& the SR 88 activities. The relation between working environment temperature and accidents has been subject to empirical research studies. An early study by Osborn and Vernon84 indicated that accidents in the munition factory under study, were at a minimum when the room temperature was 67.5°F. and when the temperature deviated from that Opti- mum, the accident rate increased for both male and female workers in prOportion to that deviation. Figure III-5 illustrates the relative accident frequency with variation in the temperature of working environment. FIGURE III-6 ACCIDENT FREQUENCY RATE VARIATIONS AS A CONSEQUENCE OF CHANGE IN WORKSHOP TEMPERATURE85 Relative Accident Frequcncy 150- 100 0 50 60 70 80 Temperature (Dry bulb degrees farenheit) 84 E. E. Osborne, et a1; Two Contributions to the Study of Accident Causation (London: Industrial Fatigue Research Board Report No. 19, 1922). 85This illustration is adOpted from: Ibid. she 89 A study of 44 shOp departments in a tractor factory showed that comfortable shOp environment appears to be a 8° It was also elabor— major determinant of safe behavior. ated that workers with low "physical tolerance" under inade— quate temperature and noise level, probably have a somewhat higher accident frequency rate than others. A number of studies have been conducted to determine the apprOpriate temperature level for different types of works and different geographical areas.87 For example, for sedentary work of office type, the desirable temperature in Britain has been suggested at 65°F. And if temperature increases to above 68°F people tend to feel sleepy and efficiency declines. In mining, a progressive increase in minor accidents rates have been observed in pits with different temperature level ranging from 620 to 85°F. Accident frequency was 3 times more in pits with high temperature than pits having low 88 It also has been found that accidents of temperature. male and female workers are affected similarly by variations in temperature, except that female workers are somewhat 86Vernon Keenan et a1; Psychological Climate and Accidents in an Automotive Plant, in Haddon, Op. cit., pp. 309-313. 87See Browne, Op. cit., pp. 64-71. 88Maier, op. cit., p. 511. lee Ne:- by pa: SUT‘Z abo Temp: 90 less affected by high temperatures than are males.89 The New York Ventilation Commission's investigation, reviewed by Maier, indicate that physical work was definitely im- paired by high temperature and stagnant air. Table III-3 summarizes findings of an experiment which was a part of above investigation. TABLE III-3 RELATIVE EFFECT OF TEMPERATURE AND AIR ON PHYSICAL HORK9° Fall in Fall in Units Production Production of Work Due to Due to Increase Temperature Air (100 Optimum) Stagnant Air in Temperature 680 Fresh 100.0 - - 68° Stagnant 91.1 _ 8.9 75° Fresh 85.2 14.8 75° Stagnant 76.2 8.6 14.5 Findings of another study suggest that the peripheral vision deteriorates with increasing temperature.91 This may well 89Ibid. 9°This table is adapted from: Maier, Op. cit., p. 552. 91A. E. Bursill,"The Restriction of Peripheral Vision During EXposure to Hot and Humid Conditions: anrterly Journal of Experimental Psychology, Vol. 10, Aug. 1958. Pp. 113-129. le ti ab. re of fat fir the Yfl'ri UEg Una? mat. take king the 91 lead to the occurrence of an accident. It should be men- tioned that extreme cold temperature also creates an inade— quate physical environment in Which the human body is un— able to prOperly function. This disfunctioning may be a result of sluggish movement, reduction in degree of feeling of fingers of toes, slow unusual response and so on. That may lead to occurrence of an accident which in turn may result in a physical impairment. Review of the work accident literature reveals the fact that findings of all endeavors in this field have con- firmed that a direct and positive relationship exists be- tween accidents and extremely inadequate temperature in the work environment. No one study -- as far as this writer's search could cover -- reports an indifferent or negative effect. Lighting. Sight undoubtedly is the most important sensor in most occupations. Yet, without proper illumina- tion seeing becomes difficult and in some cases impossible. Unable to see, a worker is expected to have an accident no matter what other personal protective measures have been taken. Obviously natural daylight is the most preferable kind of light to human beings. Although people differ in the amount of light which is most desirable for them, a Ge: is 011. 0‘th fac ti: (is 92 92 study by Feree and his associate showed that 65 percent of the subjects studied preferred the intensity of light for reading to be between ten and thirty foot candles.93 It should be pointed out that the amount of illumination depends upon the task involved. 30 foot candles are recom- mended for reading, whereas for precision die-making an intensity of 650 foot candles is apprOpriate.94 A report of 32 percent reduction in accident rates was made by General Electric Co. as a result of raising illumination level 5 to 20 foot candles in the ecroction shOps.95 Not only is the intensity of light an important factor in overall illumination of the working area but distracting factors such as glare, shadows, color contrasts and reflec- tions must be eliminated. These factors interrupt prOper functioning of eyes by fatiguing, discomforting, annoying. 92C. E. Feree and G. Rand, "Good Working Conditions for Eyes," Personnel Journal, No. 15, 1937, pp. 333-340. 93A 1 Candella Source delivers 1 foot candle of illumination on a surface when the surface is 1 foot away from the source. For more information in this subject see: a) Browne, Op. cit., pp. 55—63: b) Chapanis, Op. cit., pp. 53-63. 94DeReamer, Op. cit., p. 125. 95E. J. McCormick, Human Factors Engineering (New York: McGraw-Hill Book Co., 2nd Edition, 1964). .r. Cu be cie des sho .5 sit e e . e r\ bk 8» D. p 93 They can be eliminated by means of relocating light source, shielding and Changing light source. A report by Gary reviewed by Dr. Simonds revealed ‘that according to an estimation made by Travelers Insurance Company, 24 percent of all accidents were. due to poor lighting.96 This figure is self explanatory and indicates the importance of prOper lighting in workplaces. It should be recognized that illumination, both natural and artifi— cial, is a highly specialized function and the original design has to be checked by an illumination expert before it is installed.97 Hazards of imprOper artificial lighting shortld be carefully considered, for a survey in England showed an increase of 25 per cent in overall accidents and 75 per cent in accidental falls as a result of the neces- sity for artificial lighting.98 Noise. The review of literature indicates that Potential hazards of imprOper industrial noise level in workplaces have long been recognizes. And yet. increasing 9°Simonds, Op. cit., p. 407. 97For an extensive elaboration of this subject See: Harry Judson and James Brown, Occupational Accident W (New York: John Wiley & Sons, Inc., 1944, PP- 90-100 . 98Simonds, Op. cit. 94 annual compensation Claims of loss of hearing from exposure to industrial noise reveals the prevalence of this hazard99 which requires further study of the problem, as well as con— trol and enforcement of established rules and standards in the firms. Studies indicate that in addition to preventing fatigue, inefficiency and accidents can be reduced by controlling noise level.100 loss of hearing, It was revealed that the rhythm of the worker's movements would be dis- turbed by loud sound. As was observed through reviewing slow-motion plant movies, movements became more irregular as the sound turned louder, increasing the possibility of Other Kornhauser101 a sudden jerk that could bring about an accident. Studies have reported different findings. investigated the effect of noise on the level of produc- tiVity of four typists working in a quiet and noisy office. He found no difference when errors, amount typed and number 0f discarded papers were observed. Studies by Gary and 99David Swankin, " 3% Inspected, 90% Unsafe, in AFL- meIpdustrial Union Department: Danger: Men at Work (washington: Industrial Union Department, 1969), p. 20. 1°°This discussion is based on a condensed report Prepared by a periodical from another publication, see: "ErQ‘Onomics - New Angle on Employee Health and Safety," MEI$3ement Review, Vol. 56, No. 12, Dec. 1967, p. 48. 101A. W. Kornhauser, "The Effect of Noise on Office output," Industrial Psychology. No. 2. 1927. pp- 621-622- f 0L inv abi Ver‘ Side V651 95 102 indicated that Kryter reviewed by Professor Simonds steady or expected voices do not have a significant effect on psychomotor activities whereas studies by Farmer and Stevens and his associates have shown that average response time, gross number of errors and number of errors per unit of production increased by the presence of noise with high 103 In studying the effect of noise, frequency and level. four different aspects may be identifiable and should be investigated. The damaging effect of the noise on hearing ability, its effect on work efficiency, interference with verbal communication and finally its annoying effect. Con- sidering all these aspects in the study, may help the in- vestigator in making more workable generalizations which can be applied to the majority of working environments, if not all. Although existing knowledge in this field does not allow such generalizations, reviewing study findings tends to leave the impression that adequate noise level may be associated with safe behavior at work. No research finding has shown an increase in the occupational accident irate by reducing noise to an adequate level. 102Simonds, op. cit., p. 407. 103Ibid. /\ 1H”— wh it de< fa: the are hal hee 0n in: the 96 Accident Record Keeping System and Reporting It is obvious that no organization can survive with- out keeping track of its market share, production, sales, costs, profits and losses. Accurate and up-to—date infor— mation on these areas is needed to provide a basis for man- agement to evaluate past performance of the organization which it serves, to determine where it stands today, and its expected position in the future. The nature of infor- mation available to the manager affects the quality of his decisions which in turn determine the degree of success or failure. For this vital reason, organizations have included the book keeping function to other important functional areas of their organization. This being the fact, it can hardly be accepted that a firm is serious about safety and. health, if it does not have at hand the factual information on company's occupational accidents. An early estimation indicated the probability that not more than 5 percent of the separate business concerns keep records which will give 'them a clear picture of their accidents.104 One would Eissume that today the situation is much better than the i:ime when the above estimation was made. There exist many treasons for this expectation. One reason might be the 104John M. Roche, Safety and the Foreman (New York: ltational Foreman's Institute, Inc., 1951), p. 111. ac w}. Co he 90‘ rat ab: me; tr( Ste COR CON C QM ._\nm\ U 97 existence of state laws requiring all employers to keep records of their work injuries and to submit certain reports to state authorities and also state workmen's compensation laws. The second reason would be availability of improved and standard methods of measuring and recording work acci- dents, such as Zl6.l and 216.2; and standard method of accident cost analysis develOped by Professor Simonds, which was approved in 1954 for publication by the President's 105 Conference on Occupational Safety. The third reason may be the Federal laws such as the Walsh-Healey Act, requiring government contractors to keep records of injuries in a rather detailed and very specific way. In addition to the above reasons, safety promotion activities by many govern- mental and private institutions and availability of elec- tronic data processing systems can be mentioned. Although statistics showing the actual number of companies having a annprehensive accident record keeping system do not exist, (mumments made by one government official106 reveals that ‘ 105Rollin H. Simonds, Estimating Costs of Industrial ACCidents, U. S. Department of Labor (Washington, D. C.: U-S. Government Printing Office, 1955). 106 . . . . Mr. DaVId Swankin 18 director of the Bureau of Labor Standards, U.S. Department of Labor who delivered speech in 1969 Occupational Safety and Health Conference of the AFL-CIO Industrial Union Department. For his entire sPeech content see: Swankin, Op. cit., pp. 20-26. 'C (II it: qus def Far Mil 98 despite the points discussed above, many companies have ignored the advantageous practice of a prOper accident record keeping system. His point was that despite provi- sions of the Walsh-Healey Act, perhaps 75 percent of com- panies visited for the first time by Department of Labor authorities have not computed their accident rates or have improperly done so. It should be pointed out that sheer collection and filing of accident data would be a waste of resources. Collected data should be analyzed and interpreted in ade- quate time intervals, to find out where, when and why acci- dents occur in workplaces, and who is mostly involved. Farmer107 reports that as a result of analyzing the records, Miles and Eyre were able to greatly reduce the number of break- ages in a group of teashOps by finding out when and where they Occurred. Then, by rearranging the delivery and removal Innocess waitresses were exposed to less mental strains. By analyzing injuries as a result of weight lifting, Overton108 was: able to reduce the number of injuries by paying atten— tion to the prOper ratio between body and weight to be 3LLfted. It should also be mentioned that the considerable 107Eric Farmer, The Causes of Accidents (London: 3117 Isaac Pitman & Sons LTD., 1932), pp. 73-74. 108For more detailed discussion and reference to this study see: Ibid. 5.. t3“. Cd? ac: «flu a 99 progress made over the past half century in preventing occu- pational accidents at the national level is partially due to the collection and analysis of accident data which dir- ected action in the right direction, where it was needed so that progress could be made in preventing accidents. At the company level too, it is difficult to believe that a firm, without knowing the frequency, nature, severity, cause, place and especially the cost of its occupational accidents, could provide a fairly safe working environment- for its members nor could it maintain an impressive safety performance record. It was on the basis of this belief that a record keeping system was included as one of the variables to be investigated in this study. Machine Guarding, Safety Devices and Controls Occasionally, discussion has appeared in the litera- turwa that most occupational accidents are due to human faults and this argument has been used by management against pro— Posals for guarding of machinery and equipment in the plant. It 11218 also been mentioned that even though lack of prOper machine guarding may have caused just a few accidents, the analysis of accidents nevertheless show that machines are the source of about 10 percent of fatalities and permanent tOtal disabilities, and of nearly 25 percent of accidents (13" (h ung 100 causing permanent partial disabilities.109 Some complaints have been made that supervisors resist the idea of machine guarding and other safety devices 110 because they feel that production will be interferred with. Today there is no logic for such an attitude for it is ob- vious that production will suffer if the worker has to divide his attention between the immediate task and the unguarded or imprOperly guarded machine with which he works. Of course, machine guarding or a device which is partially safe, may actually increase the number of acci— dents because the Operator may overestimate the safety of the machine and become less cautious than he was before the safety device was installed or guarding was provided. Maier gnaints out that men will adjust themselves to the degree of apuoarent danger, and if that danger becomes less apparent due: to safety devices or guarding, then it may result in an 112 increase in the number of accidents. A study by Vernon in England indicated that when traffic density increased it made the danger more apparent which consequently decreased ¥ 109Judson, Op. cit., p. 101. 110DeReamer, op. cit., p. 130. lllMaier, Op. cit., p. 507. 112H. M. Vernon, Accidents and Their Prevention (London: Cambridge University Press, 1936), p. 120. (1‘ mo: exi wit enc unr Pro. 101 the number of traffic accidents. But when roads were empty and straight, the danger became less apparent and drivers underestimated the hazard of speed which led to an increase in the number of accidents. Another example of this would be the difference between the number of urban and rural. automobile accidents. It should also be pointed out that, installing a fairly perfect safety device or machine guarding may pro- tect the operator from one hazard and make him unaware of others. A study by Keenan and his associates in an auto- motive factory showed that where an obvious danger factor exists, the accidents which occur tend not to be identified with the obvious danger.113 They also cited that the exist- ence of an impressive obvious hazard seems to contribute to unrelated accidents by delimiting attention and encouraging proneness to involvement in the non-obvious hazards.114 Some criteria have been suggested for prOper guard- ing and safety devices which may be listed in order of :importance as follows:115 a) Impervious - fully protecting from all possible energy release; 113Keenan et a1, op. cit., pp. 312-313. 114Ibid. 115Surry, Op. cit., p. 121. [CIS 102 b) Non interfering (with the man's sensory - control functions required for the task; c) Fool proof, protecting under all task and stress conditions; d) Comfortable, no irritation or need for aWk- ward movements; e) Operator independent, to avoid deliberate removal: f) Inexpensive. Machine guarding compared to safety devices and equipment has been more successful. It has been suggested that for further improvement in designing these devices and equipment user Opinion, past experience and human factors engineering technology should be considered.116 It is also recommended that studies should be conducted by psycholo- gists, sociologists, marketing research specialists, and human factor engineers in order to explore reasons why many excellent devices now available are not being used. In general, existing knowledge in this field sug- gests that machine guarding and installment of safety devices are effective means of accident prevention if the criteria mentioned above are met. 116International Occupational Safety and Health Information Center: Ergonomics of Machinnguardigg (CIS Information Sheet 10, Geneva 1964). Mn {1‘ .-.J Oh I. ma 0C MOI Yf'vor are '1 fiere 103 Status of the Occupational Accident Research Review of the literature provided a research based background for the factors included in the study. This was the main contribution expected from the content of existing literature. But in addition to extraction of specific study findings, exposure to accident research literature enabled this writer to formulate an overall idea about the present status of research in this field. This ideation may be manifested by saying that accident research in general and occupational accident research in particular is highly heterogeneous in content and emphasis. There are many kinds of accidents occurring in many different places. Numerous factors are candidates for the cause of each accident and often blaming a specific factor is quite arbitrary. Scientific investigation of accidents ranges from intensive study of the effect of specific factor to broad-scale study of accidents. The scarcity of research is clearly felt but lack of coordination among existing research findings is more tangible. It should be pointed out that research in some areas is much more advanced than in others. Engineer- ing has already contributed considerably to reduction of ‘work accidents whereas findings of psychological studies Eire not comparable partially because psychological variables Vvere not so readily identifiable as are technical factors. r# W) (’t (I) re has 104 But it appears that effective accident prevention depends upon findings of more disciplines than the two just men- tioned. Medical research is needed to investigate the physiological and pathological effects of environmental and technological factors and the physical circumstances con- ducive to accidents. Statistical research is required to ascertain what kinds of accidents occur in what numbers to what types of people, in what Operations and from what causes. Need for further educational research is obvious since more effective ways of teaching safety are yet to be discovered. More information is needed about calorific requirements of work and the functioning of the body which has to be provided by further physiological research. There is a lot to be learned about prOperties and characteristics of harmful materials, better machine guarding, more effec- tive safety devices and controls, better personal protective «equipment and so on, which necessitates further technical :researches. The last area which is in need of research 19erhaps much more than any other, is psychology since a lot luas to be discovered on psychOIOgical patterns conducive tC> accidents. Obviously, this is not a claim for an exten- Sinve list of fields which need further improvement in their StEite of existing knowledge nor it is a complete list of the diSciplines which could contribute to the prevention of 105 accidents. Other important areas such as safety regulation, laws, standards are not excused from further improvements. It should be pointed out that even if needed research in the above areas were undertaken, unless the findings are prOperly integrated, their utilization can not be Optimized. This integration has been missing to a great extent in the accident research so far. But the recent develOpment of the concept of "ergonomics" which embodies most of the essential fields discussed above, appears to be a positive and promising solution to this problem. ’4 t] 86 ir 0i CHAPTER IV STUDY FINDINGS Introduction This chapter is an attempt to report the findings of the field study and is composed of three major sections. The first section includes specific background data and statistics such as company accident frequency and severity rates, number of employees and so on which were determined in the company location through interviewing, reviewing com- pany records or using data obtained from the Michigan State Department of Labor. The second major section deals with the results of testing hypotheses of the study. The third‘ section is a report of findings which were not basically included in the hypotheses of the study but in the course of research were found to be of considerable importance to the field of occupational accident research either by this writer or by interviewees who felt very strongly about the. importance of some of the factors. The main body of the chapter follows the same order mentioned above. 106 (n rt wa Whe den 7a,. man. aCQj 107 Specific Background Data and Statistics The reason for enclosure of specific company data and statistics at this point in the report is that interpre- tation of findings as a result of testing hypotheses will be more meaningful if this background data is considered and kept in mind. The first statistical data to be men- tioned is the average number of employees at the employ of each firm since this information was directly or indirectly consulted or used as a basis for computation of other sta- tistical data such as accident rates. Table IV-l shows the average number of workers employed in the companies studied, at the time of this writer's visit. The total number of employees of all 22 firms studied was 6,100. Eleven firms with relatively high accident records employed totally 3,195 workers and eleven companies with low accident rates had 2,905 workers at their employ. It is also interesting to note from Table IV-l that the total number of female employees of the firms with high accident rates constituted approximately one fourth of the total employee pOpulation for this latter group of companies whereas total women employees for the firms with low acci- dent rates was approximately one-sixth of the total employ- ment figure. This is not an attempt to attribute higher accident rates to work force composition of the companies 11 To 108 TABLE IV-l AVERAGE NUMBER OF EMPLOYEES AT THE EMPLOY OF COMPANIES STUDIED BY INDUSTRY Firms with HIGH Accident Rates Firms with L9H Accident Rates Col. SIC No. Industry Code* Total Male Female Total Male Female 1 Meat products 201 300 250 50 275 215 60 2 Dairy products 202 80 73 7 190 150 40 3 Canning preserving 203 190 70 120 125 35 90 4 Household furniture 251 225 175 50 195 105 90 5 Paper containers and boxes 265 155 130 25 80 3O 50 6 Iron and steel foundries 332 675 574 101 350 336 14 7 Non-ferrous metal, rolling, etc. 335 160 104 56 350 339 ll 8 Other primary metal products 339 375 357 18 275 260 15 9 Metal stamping 346 260 220 40 290 270 20 10 Mbtal working machinery equipment 354 225 212 13 275 265 10 11 Motor vehicle 371 550 220 330 500 488 12 Total 3,195 2,385 810 2,905 2,493 412 *Standard Industrial Classification Code. (+ F! n: *(1 n: da ta‘ 109 with respect to employee sex since that hypothesis was not tested in this research, but it leaves an Open and poten— tially fertile ground for further study. The next statistical data is the companys' accident rates. As was pointed out earlier, frequency rate of com— panies was determined by using information and data pro- vided by the companies as well as some data collected from the files of the Michigan State Department of Labor. Table IV—2 indicates accident frequency rate of the companies studied. These rates are for 1969 and since generally a firm's accident frequency rates may not change drastically in a relatively short period of time or even in a year to year basis, there is a good probability that these rates may represent companies' safety performance in the recent past and also near future, though they are based on 1969 data. The industry average frequency rates shown in this table are taken from Michigan work injuries COOperative survey1 and are for the year 1968. It also should be pointed out that these industry average rates represent only those companies that participated in the survey which was on a voluntary basis. It could be speculated that rates would tend to be higher if the average was taken for all the g 1Michigan Department of Labor, Michigan Work Injuries COOperative Survey, 1968. A M Cc SE fc C3c 110 TABLE IV-2 THE WORK.ACCIDENT FREQUENCY RATES OF THE FIRMS STUDIED Firms With Industry Average Firms With SIC LOW Accident Frequency Rates HIGH Accident Industry Code Rates (Michigan)* Rates Meat Products 201 12.0 37.0 51.7 Dairy Products 202 25.2 36.1 37.0 Canning Preserving 203 18.8 24.8 61.9 Household Furniture 251 25.2 29.9 89.4 Paper Containers and Boxes 265 14.5 17.6 45.1 Iron and Steel Foundries 332 15.5 18.2 69.67 Non-ferrous Metal Rolling, etc. 335 34.1 21.6 45.0 Other Primary Metal Products 339 41.7 32.2 71.1 Metal Stamping 346 16.0 13.8 65.2 Metal Working Machinery 354 44.2 11.1 173.0 Motor Vehicles and Equipment 371 51.6 3.5 128.0 companies in each industry group. The reason for this speculation could possibly be that those firms volunteering for participation may be more safety minded and might have done more in the way of accident prevention and safety. Industry average frequency rates for firms in Michigan were (). (‘7‘ .3" in CC Ye be fa alt 111 not available for 1969 in any form at the time this study was conducted. Although data needed for computation of the accident frequency rates were not readily available in the firms, the necessary figures to calculate accident severity rates were even more scarce. In some companies these data did not exist and search for them in other places proved to be un- productive. It was mentioned earlier in Chapter II that most of the firms did not consider the severity rate as a good indicator of the company's safety performance, so they did not keep a record of severity rates. The reasons given by the firms are listed in Chapter II and are not repeated here. Wherever possible, data were gathered and the severity rates of the companies were computed. It is obvious that severity rate becomes a better indicator of safety performance when it is calculated and considered for a number of years rather than for only one year. The longer the period under study, the smaller will be the possibility that sheer chance factors, such as fatality resulting from heart attack and so on, may have altered the firm's accident severity rate. In spite of the :fact that the Table IV—3 shows severity rates for a period. CJf only one year, in most cases it coincides with the direc- tzion shown by frequency rate in Table IV-2. In other words, 112 TABLE IV-3 THE WORK ACCIDENT SEVERITY RATES OF THE FIRMS STUDIED SIC Firms with LOW Firms with HIGH Industry Code Accident Rates Accident Rates Meat Products 201 No data No data Dairy Products 202 No data 751.8 Canning and Preserving 203 988.0 1383.0 Household Furniture 251 185.0 243.2 Paper Containers and Boxes 265 616.0 1033.0 Iron Steel Foundries 332 804.7 1034.0 Non-ferrous Metal Rolling, etc. 335 588.8 734.2 Other Primary Metal Products 339 1613.0 899.0 Metal Stamping 346 580.0 1789.7 Metal Working Machinery 354 581.1 430.4 Motor Vehicle and Equipment 371 815.0 4130.0 most firms having low frequency rates, tend to have rela- tively low severity rates as well. By comparing these two tables, it is noticeable that in most cases when frequency rates of the firms are close figures, then their severity rates too, tend to be similar. When frequency rates of two firms differ considerably, then their severity rates too, appear to follow a more or less similar difference pattern. Another item of background information collected 113 from the firms deals with companies' membership in safety organizations since this could be an indicator of a firm's safety mindedness. According to statistics presented in the "Accident Facts," companies that are a member of the National Safety Council tend to have considerably better safety performance records than non-member companies. As it is shown in Table IV-4, out of eleven firms with low accident rates 8 (or 72.7%) firms are members of a safety organization and only 3 (or 27.3%) hold no membership. From the total of eleven firms with relatively high accident rates 7 (or 63.7%) are not a member of a safety organization and only 4 (or 36.3%) of these firms are members. These figures indicate a direction that could be used for the formulation of a hypothesis that: "The firms that are mem- ber of a safety organization, tend to have considerably better safety records, than do non-member firms." Obviously, valid generalizations can not be made at this point with, respect to this "membership effect" unless the above hypothesis is empirically tested in a number of firms in different industries which is subject to further research. The degree of automation of production was also considered to be an important background factor especially from the safety on the work point of view. Previous research studies have indicated that the greater the degree of manual 114 TABLE IV-4 MEMBERSHIP IN THE SAFETY ORGANIZATIONS Industry Firms with LQW_ Firms with glgg Code Accident Rates Accident Rates 201 Yes No 202 Yes Yes 203 No No 251 No No 265 Yes Yes 332 Yes No 335 No Yes 339 Yes No 346 Yes Yes 354 Yes No 371 Yes No Total 8-Yes 3-No 4-Yes 7-No Percent of Firms 72.7% 27.3% 36.3% 63.7% (Nsll) (N=11) (N=11) (N=11) effort involved, the higher accident rates tend to be on 2 the average. As is shown in the Table IV-5, out of the (eleven matched-pairs of companies studied the degree of auto- rnation of production was the same for 6 pairs and, 3 firms k 2See for example: Keenan et al.; Op. cit., p. 312. 115 TABLE IV-S PERCENTAGE OF THE PRODUCTION AUTOMATED IN THE FIRMS AT THE TIME OF STUDY ,Industry Firms with LOW Firms with HIGH Code Accident Fr. Rates Accident Fr. Rates 201 0-10% 0-10% 202 75-85 41-50 203 41-50 41-50 251 0-10 0-10 265 0-10 41-50 332 0-10 10-20 335 51-60 21-30 339 0-10 0-10 346 51-50 0-10 354 0-10 0-10 371 0-10 0-10 (Mdn=0-10%) (Mdn=0-10%) with relatively low accident rates had a higher degree of automation than their respective matched pairs. Also in two industries, two firms with relatively high accident .rates also had a higher degree of automation. Even though tilese data are not statistically evaluated, it appears that (tifferences in accident rates of the firms studied can luxrdly be attributed to the difference in the degree of SQ Ct Se PE 116 their automation in this study. Obviously this discussion does not provide, by any means, evidence to make generali— zations about the effect of automation on accident preven- tion but rather an effort to show how firms included in this study compared with respect to their degree of automation. It should also be pointed out that the degree of automation was determined through a direct question to inter- viewees.In some cases the answer was given right away, in some others, the interviewees consulted a more informed person in the matter. The last background information collected from com- panies to be mentioned in this section has to do with the safety unit or organization. As shown in Table Iv—6, except in one case, none of the firms included in this study had a separate organiza- tional unit specifically responsible for carrying out the safety function. In most firms the safety function was concentrated in the personnel departments which were ob- served by this writer to be clearly overwhelmed by routine personnel problems as well as other activities. Also, in some cases a member of the personnel department was intro— duced as the person in charge of safety or as the safety director, but further questions revealed that safety was only a sub-function and not their main responsibility. Sn 117 TABLE IV-6 ORGANIZATIONAL UNITS RESPONSIBLE FOR SAFETY OF THE COMPANIES Firms with LQW_Accident Firms with §I§5.Accident Fr. Rateg Fr. Rateg Industry Safety Org. Responsibility Safety 01g, Responsibility Code Yes No Center Yes No Center 201 X Industrial Rel. X Personnel 202 X No specific unit X Personnel 203 X Plant Management X Plant Management 251 X Personnel X Personnel 265 X I Safety Branch X Production 332 X Employee Relation X Personnel 335 X Plant Engineer X Personnel 339 X Personnel X Personnel 346 X Personnel X Industrial Rel. 354 X Personnel X No specif unit 371 X Personnel X Employee Rel. Total 1 10 - 11 For this reason they were not considered as representing an independent or separate functional unit in the organiza- tional structure. Among 22 firms only one had a safety organization which was staffed and Operated on a full time basis within the personnel department. By virtue of the fact that a great majority of smaller size firms have serious work accidents, this writer's In It» Irr: IP- 118 contention would be that the establishment of a safety organization should be given mOre'emphasis by practitioners and experts in the area of small business administration. Further, its size, structure and relations with other functional areas must be determined by considering the extent of Operations, type of business and extent of safety problems prevailing in each business concern. Testing_of the Hypotheses For the analysis of data collected for testing of hypotheses of this study the Wilcoxon matched-pairs signed rank test was primarily employed. In some cases where data type was not amenable to evaluation by that test, the sign test was used.3 Instances where the latter was employed will be mentioned in the text of reporting results of hypothesis testing. For all hypotheses where no such specification is made, the former test has been used. Hypothesis #1: In the firms with lower work injury frequengy and severity rates, top management is highly interested and involved in the company's overall safety programs and actively participates in and supports safety activity. 3For explanation of the nature of these tests see: Appendix E. ff) 1 A. re 119 The results of testing this hypothesis provided a positive and strong support for confirmation since a signi— ficance level of (a = 0.05) was obtained. Table IV-7 shows, in summary, how tOp management's support and involvement in safety differed in firms with low accident rates compared to firms with high accident rates. Although differences exist in most of the activities listed in the table, three areas represent the largest difference. It appears that tOp management in the firms with better safety records, puts more emphasis on personal audit and inspection than does tOp management of the firms with relatively poor safety records. Also, not only does tOp management of the firms' with better safety performance show more interest in plans for achieving certain safety objectives but it also holds review and analysis sessions to ascertain that those plans are being carried out prOperly and according to projected objectives. Obviously, this practice enables tOp manage— ment to take timely and constructive corrective measures. In the firms with relatively high accident rates this prac- tice is not being given much emphasis and, as it is shown in Table IV—7, considerable difference exists between firms with low and high accident rates in this matter. The third large difference is in the area of including safety reports, figures, and achievements on the agenda of board 120 TABLE IV-7 TOP MANAGEMENT'S SUPPORT AND INTEREST IN SAFETY AND HOW IT DIFFERS IN FIRMS WITH LOW AND HIGH ACCIDENT RATES Firms With Firms'With Description of Top Management's Lgfl_Acc1dent .§I§§_Accident Fr. Rates Fr. Rates Activity with Regards to Safety (N=111, N=111, Yes No Yes No 1. Does he attend any safety meetings in the company? 8 3 6 5 2. Does he chair any of these meetings? 3 8 2 9 3. Does he regularly receive safety reports? 11 0 11 0 4. Does he personally conduct any safety audit or inspection 9 2 4 7 5. Is he a member of any safety organization? 0 11 0 11 6. Does he regularly attend any safety meetings or conference outside the company? 2 9 0 11 7. Does he emphasize plans for achieving certain safety objectives? 10 1 8 3 8. Does he actively participate in execution of safety plans? 10 1 7 4 9. Does he hold review and analysis sessions in order to compare the results of carrying out safety plans with projected objectives? 9 2 3 8 10. Are safety figures, reports, achieve- ments included on the agenda of company board meetings? _8 _3_ _3_ _§ Total Score 70 40 44 66 t \Qza 121 meetings. This is being practiced in the firms with better safety records to a greater extent than in the firms with poor safety records. It appears that improvement in the above three areas may contribute to fill the gap which exists between firms with markedly different work injury rates. Table IV-8 shows how matched-pairs of firms in different industry groups differed in "score"4 with regard to tOp management's support and involvement in safety. These are actual scores which were used to test the hypothesis and determine its significance level. The most interesting and important observation in this table is that where great differences exist between firms with respect to their score of tOp management's support and involvement in safety, a great difference is also observable in their respective accident frequency rates. This point can be clearly observed when firms in industries 201, 265, 346, 354 and 371 in the above table are scrutinized. When scores are close, the respective frequency rates also tend to be close. In brief, not only did statistical testing of this ihypothesis show a very strong significance level, but a 4In determining these scores, a total of 10 ques- ‘tioms were formulated and asked. For each question an answer of "Yes" (with value of l) or "No" (with value of Zero) was required. These scores represent total of "Yes" answers. See Appendix C for the nature of questions asked. 122 TABLE IV-8 THE TOTAL SCORE OF THE FIRMS ON TOP MANAGEMENT'S SUPPORT AND INVOLVEMENT IN SAFETY Firms with LOW Firms with HIGH Industry Accident Fr. Rates Accident Fr. Rates Code gore Fr. REE; m 201 7 12.0 2 51.7 202 9 25.2 7 37.0 203 2 18.8 2 61.9 261 4 25.2 4 89.4 265 7 14.5 2 45.1 332 6 15.5 5 69.67 335 4 34.1 6 45.0 339 8 41.7 7 71.1 346 7 16.0 4 65.2 354 9 44.2 1 173.0 371 _7 51.6 _4 128.0 Total 70 44 (Mean = 6.3) (Mean " 4) ' comparison of firms' scores while considering their acci- dent frequency rates provide evidence that leads strongly to belief that tOp management's earnest support and involve- ment in company's safety is undoubtedly one of the determi- nate factors in achievement of good safety performance records. Where this support and involvement is missing, Im In: In: lo an hi It fi 81:1 ave to tha 123 firms tend to suffer greatly from high work accident rates. Hypothesis #2. In the firm's with lower injury frequency and severipy ratesL_workers have higher average attainment of formal education than in the firms with relatively poor work injury records. The results obtained from statistical testing of this hypothesis are inconclusive. By using the sign test, it was revealed that the educational level of workers ig not significantly related to accident rates of the firms included in this study. Thus the hypothesis was not con— firmed. Table IV-9 shows the average educational level of the workers in the companies studied. Although firms have markedly different accident rates, educational level of their workers in most cases are ties. In industries 251 and 354, firms with higher accident rates have also the highest educational level of workers among all 22 firms. It is interesting to note that most of the workers in the firm having high accident rates in industry 251 are college students, since that firm is owned by a university. Thus, average age of the workers which is in 18-23 bracket seems to be affecting accident rates along with other variables -- to be discussed in remaining parts of this chapter -- rather than the educational level. m We to 124 TABLE IV-9 AVERAGE EDUCATIONAL LEVEL OF THE WORKERS IN THE FIRMS STUDIED Industry Firms with LOW Firms with HIGH Code Accident Fr. Rates Accident Fr. Rates 201 10-12 Grade 10-12 Grade 202 10-12 " 10-12 " 203 10-12 ” 10-12 " 251 10-12 " l3 & over " 265 10-12 " 10-12 " 332 10-12 " 10-12 " 335 10-12 " 7-9 " 339 10-12 " 10-12 " 346 10-12 " 7-9 " 354 10-12 " 13 & over " 371 10-12 " 10-12 " (Mdn = 10-12) (Mdn = 10-12) It must be noted that this is not, by any means, to generalize that the educational level of the worker is not a significant factor in any situation. It may be a deter- minant variable where differences in educational level of workers vary greatly which is subject to further research. But as far as the firms included in this study are referred to, the educational level of the workers did not prove to be 125 a significant variable affecting accident rates of the firms. Hypothesis #3. In the firms with better safepy records, average gge of the employees is higher than in the firms having relatively high accident rates. By a significance level of (a = 0.01) this hypothesis was confirmed as a result of statistical testing. Based on the evidences provided by findings of this study, age of the employees is significantly related to the occupational acci- dent rates of the companies. In most industries of this study, where accident frequency rates of the companies vary considerably, the average age of the employees also more or less follows the same pattern. Table IV—lO is an illustra- tion of the average age of workers. This can be clearly observed by considering firms in industries 201, 251, 339, 346, 354 and 371. This correlation between age and acci- dent rates may stem from the fact that normally, lower age. is associated with a lesser degree of experience. It may. also result from the attitude of young men in regard to risk taking. The younger worker's voluntary involvement or arbitrary assignment to more hazardous tasks in the com- panies could possibly explain the nature of the relation— ship existing between age and accident rates. l<: Im In: 11: 01 126 TABLE IV-lO THE AVERAGE AGE OF EMPLOYEES IN THE COMPANIES STUDIED Firms with LOW Firms with HIGH Industry Accident Fr. Ratep, Accident Fr. Rates Code Average Age , Average Age Bracket Fr. Rate Bracket Fr. Rate 201 35~40 years 12.0 29-34 years 51.7 202 41-46 25.2 41-46 37.0 203 35-40 18.8 35-40 61.9 251 35-40 25.2 18-23 89.4 262 35-40 14.5 29-34 45.1 332 35-40 15.5 35-40 69.67 335 35-40 21.6 35-40 45.0 339 41-46 41.7 35-40 71.1 346 41-46 16.0 29-34 65.2 354 41-46 44.2 29-34 173.0 371 35-40 51.6 29-34 128.0 (Mdn = 35-40) (Mdn = 29-34) or 37.5 or 31.5 Hypothesis #4. Firms with better safety records haveppro- vided recreational prpgrams and facilities for the employees, and byydoing so, have helped to bring employees together. Statistical analysis of data collected for this hypothesis showed a considerably strong significance level of (a = 0.01). Table IV-ll summarizes how firms in eleven 127 different industries scored in connection with the questions raised about recreational programs and facilities that they provide for their employees. In industries 201, 251, 339, 346, 354 and 371 where firms in each pair have considerably different accident frequency rates, the difference between their scores shown in the Table IV-ll also varies markedly. There are two possible explanations for such a high level of significance of the hypothesis. Firstly, while providing recreational facilities and programs apparently do not seem to have anything to do with accident prevention or safety performance of the company, it is important to note that such provision is definitely indicative of a com— pany's interest in its employees. Where a company cares about recreation of its employees, it is logical to deduce that the safety of those employees will not be ignored nor would its importance be underestimated. The second possible explanation would be that recreational programs and facili- ties help to bring employees together which in turn may lead to creation of a friendly relationship among workers. It is not an exaggeration nor irrational to say that a friendly work environment where workers care about each other, will be less conducive to work accidents than will an indifferent or hostile environment. .9! 128 TABLE IV-ll COMPANY SCORES ON PROVIDING RECREATIONAL PROGRAMS AND FACILITIES FOR THEIR EMPLOYEES Firms with Egg Firms with gggg Industry Acpident Fr. Rates Accident Fr. Rptes C°d¢ Score5 Fr. Rate Score5 Fr. Rate 201 6 12.0 1 51.7 202 0 25.2 1 37.0 203 3 18.8 2 61.9 251 6 25.2 2 89.4 262 4 14.5 3 45.1 332 5 15.5 5 69.67 335 2 21.6 2 45.0 339 4 41.7 1 71.1 346 4 16.0 1 65.2 354 6 44.2 4 173.0 371 4 51.6 2 128.0 (Mean = 4) (Mean = 2.1) (Mdn = 4) (Median - 2) 5These scores were determined through seven major questions asked and a value of 1 was given to a "Yes" answer and zero to a "No" answer. See Appendix C fOr the nature of these questions. 129 Hypothesis #5. Firms with low injury records have tried to promote employee's safety interest through their families. The results of testing this hypothesis are incon- clusive. Although differences were not statistically sig- nificant in this study, before any generalizations can be made, further and extensive studies are suggested to ex- amine the effects of safety promotion through employee families. The reason for this need is that a great majority of the firms, as shown in Table IV-12, scored low in this practice not because they thought it was not an effective .means of promoting safety but because they were not familiar with it. Most of the interviewees while saying that they had heard about this for the first time from this writer, contended that it was "a pretty good idea" and they were going to try it. It is interesting to note that the interviewee of the firm with high accident rates in industry 332 claimed that for 1970, the firm has been very successful in reducing accident rates. He mentioned the promotion of safety through worker families as one of the most effective means utilized by the company in its recent safety performance achieve- ment. 130 TABLE IV512 COMPANY SCORES ON PROMOTING SAFETY THROUGH EMPLOYEE '8 FAMILY Industry Firms with Egg Firms with HIGH Code Accident Fr. Rates Accident Fr. Rates 201 1 3 202 3 l 203 O l 251 3 0 265 4 1 332 1 . 5 335 1 1 339 1 1 346 1 1 354 1 1 371 1 1 (Mean = 1.54) (Mean = 1.45) (Mdn = 1) (Mdn = 1) Hypothesis #6. Firms with low work injury experience have established comprehensive safety rules covering all oper— ationsy have made sure that all employees understand them, and have consistently enforced them. It is interesting to note that even though 7 out of the eleven firms with relatively low accident rates a. 131 scored6 higher with respect to safety rules, than did their complement in the matched pair with high accident rates, this hypothesis was not supported statistically and a sig- nificance level of a = 0.05 was not obtained. This is partially due to zero scores of the two firms with low accident rates in industries 339 and 346 and to a lesser extent to the lower score of the firm with better safety records in industry 332. Those firms with zero score did not have any safety rules. Table IV—13 shows the score of companies studied with respect to their safety rules. If those firms with a zero score as well as their complement are eliminated from the analysis and the number cu? matched-pairs is then reduced to 9, a strong signifi- cance level could be obtained by the remaining data to support the hypothesis. This would suggest that with dif- ferent sets of companies in other studies the findings could be markedly different with respect to the effect of safety rules in safety performance. The findings of this study, however, do not show significant support for this hypothesis. 6In determining these scores, a total of 16 ques— tions were formulated and asked. Some of the questions required only Yes or No answer with value of (l) and (0) accordingly and some were determined with a 3 point con- tinuum. See Appendix C for the nature of the questions. Ex :3 :Q 11 En 132 TABLE IV-13 COMPANY SCORES WITH REGARD TO THEIR SAFETY RULES Industry Firms with M Firms with H;_GH Code Accident Fr. Rates Accident Fr. Rates 201 12 7 202 9 9 203 10 8 251 11 4 265 11 6 332 14 \ 10 335 4 7 339 0 7 346 0 7 354 14 4 371 11 5 (Mean = 8.7) (Mean - 6.7) (Mdn = 11) (Mdn ” 7) Hypothesis #7. Averpge number of married employees is higher in the firms with better safety records than in firms with poor safety records. This hypothesis was strongly supported by the find- ings of this study with a significance level of (a = 0.01). This simply indicates that the higher the number of married employees, the better will be safety performance experience 133 of the firm. As it is shown in Table IV-14, where firms differ greatly in the percent of their total employees who are married, their accident frequency rates also vary to a great extent. Why married employees tend to behave safety at work is a function of changes in their social status and role. As a husband or father, the married employee carries different sets of responsibilities which influence his be- havior to be more consistent and dictates to him to stabilize his career. Love and concern for the family reminds the married employee of the unaffordable consequences of getting injured and temporary or permanent unemployment. The married worker is normally more restricted in his geo- graphical movements and when he settles down in a company, it is normally a pre—planned and well-thought move. On the other hand, single workers are normally on the search for another organization which will provide them more prestige and monetary gains plus other important consider- ations. This attitude toward a job held presently by a worker, effects safety behavior of the workers. 134 TABLE IV-14 PERCENT OF THE TOTAL EMPLOYEES WHO‘WERE MARRIED IN THE FIRMS STUDIED Firms with HQH Firms with‘ngH Industry Ac;i::nt Fr. Rates Ac;i::nt Fr. Rates Code Employees Fr. Employees Fr. Married Rates Married Rates 201 85% 12.0 80% 51.7 202 95 25.2 95 37.0 203 95 18.8 90 61.9 251 85 25.2 32 89.4 262 90 14.5 65 45.1 332 95 15.5 91 69.67 335 91 21.6 85 45.0 339 90.2 41.7 35 71.1 346 90 16.0 75 65.2 354 98 44.2 82 173.0 371 95 51.6 90 128.0 (Mean = 83.6%) (Mean = 74.5%) (Mdn = 91%) (Mdn 8 85%) ,Hypothesis #8. Average number of years spent with the compapy is higher for employees in the firms with low work injury records than for employees in the firms with high work accident records. The results of testing this hypothesis are 135 conclusive and a significance level of (a = .05) was found. The evidence strongly leads to belief that length of em- ployees service in a company is positively correlated with low accident rates. Higher length of service means lower turnover rates or a more stable employment situation in a company. Higher turnover rate is associated with a higher number of new employees coming to organization. And number of accident occurring to or caused by new employees account for a great portion of total accidents. This presumably has to do with the lack of experience which was discussed in detail in Chapter III. Thus it is not surprising to ob- serve that in all the companies with good safety records -- with one exception -- average length of employee's service is higher than in companies with relatively high accident rates. Table IV-15 illustrates this point. 136 TABLE IV- 1 5 THE AVERAGE LENGTH OF EMPLOYEES' SERVICE IN THE COMPANIES STUDIED Firms with LOW Firms with HIGH Industry Accident Rates Accident Rates Code Average Length Average Length of Service Fr. Rate of Service Fr. Rate (Years) (Years) 201 16 and over 12.0 1-3 51.7 202 13-15 25.2 13-15 37.0 203 4-6 18.8 1-3 61.9 251 10-12 25.2 . 1-3 89.4 262 13-15 14.5 4-6 45.1 332 7-9 15.5 4-6 69.6 335 13-15 21.6 13-15 45.0 339 16 and over 41.7 16 and over 71.1 346 ' 13-15 16.0 10-12 65.2 354 10-12 44.2 7-9 173.0 371 7-9 51.6 13-15 128.0 (Mdn = 13-15) (Mdn = 7-9) Hypothesis #9. Older companies have lower work injury frequency and severity rates than do relatively newly _gptablished companies. This hypothesis was neither supported nor refuted by this study. The main reason for formulating this hypothesis was that newly established companies normally 137 try to identify their position in the competitive market and are to a certain extent aggressive to get as large a market share as they can. A newly established company's survival, in most cases, depends upon the success of this struggle for finding a foothold in the market. When the company is fighting for a greater cause such as survival, then safety, per se. may not represent an urgent problem for these companies. Another reason for formulation of this hypothesis was that normally, older companies through eXpe- rience have develOped fairly good knowledge of the Opera- tions or machinery and equipment that are hazardous and they may have installed safety devices or have provided specific personnel protective equipment in order to reduce their accident rates, whereas this kind of knowledge nor-. mally would not be available for a newly established firm. Despite the reasons given, the findingswuomwwm m>wuommmm uocwwm HH< um O>wuocmwm m>auommmm o>aumwmz uommmm umcsuaom mum> u>aumwmz uummmm umssmaom muo> Opoo oz |H‘ oz muumsch moumm .um ucmpwoo< mmuwm .Hm ucmpwoo< mch Lugs mauwm 30A sues mapfim ZOHHZEVE Hzmaooao. zo chHommmZH wymmN .mmm3MH>MmHZH wNI>H mamdfi 164 Summary In this chapter some specific background data and statistics such as number of employees in the companies studied, firms' accident frequency and severity rates and firms' membership in safety organization were first reported. The second major section of the chapter dealt with reporting of the results of testing the hypotheses of this study. Finally, additional information of importance to safety performance of companies, collected in the course of study were reported. CHAPTER V SUMMARY, CONCLUSIONS AND RECOMMENDATIONS Summary About every 8 minutes, the following estimates turn to actual work accident statistics in the United States: one fatality, 148 disabling injuries and at least 500 less serious injuries. Occupational accidents annually claim more than 14,000 human lives and cause over 2,000,000 dis- abling injuries. Some industries because of the nature of tasks, amount of risks involved, or degree to which they recognize safety as a problem, contribute more to these figures than do other industries. Not only do different industries demonstrate markedly different work accident rates, however, but existing knowledge further reveals that most of the smaller size (under 500 employees» firms expe- rience much higher accident rates than larger companies within the same industry, though some small size firms have excellent safety performance records. This study was an attempt to investigate empiri- cally those factors that could possibly provide an explan- ation for differences in accident experience and rates of 165 166 similar size firms in the same industry. The literature on, and related to, occupational safety was extensively reviewed in order to identify those possible factors. A total of fifteen factors were selected with respect to this endeavor, these factors were: 1) tOp management's support and involvement in safety; 2) formal educational level of workers; 3) age of the employees; 4) company recreational programs: 5) promoting safety through em- ployee's family: 6) safety rules of the firms: 7) marital status of the workers; 8) worker's length of service in a company; 9) company age; 10) accident record keeping system; 11) first line supervisor's relative span of control; 12) relative age of production machinery and equipment; 13) physical workplace conditions: 14) safety devices and controls on machinery: and 15) safety committees. Then fifteen hypothesis were formulated concerning each factor mentioned above. The next major step in the study was sampling. The sample had to be a number of matched-pairs of industrial concerns, each pair comprising firms of similar size in the same industry, but with one of the two having consid- erably more work injuries than the other. Eleven different industries were randomly selected. Then within each selected industry two firms of similar size but markedly 167 different accident rates were selected. All of the twenty- two firms included in the sample of this study were in the State of Michigan and were spread around in 16 different locations throughout the state. For the collection of data all companies were visited their facilities inspected, and members of management interviewed. Also, some statistical data were obtained from the Michigan Safety Bureau after Specific authorization from the several companies. A questionnaire was develOped as the interview guide to in- sure that the same questions in the same order would be asked of all the interviewees. The interview was highly structured but after all questions listed in the interview guide were discussed, the interviewees were asked to feel free and express their viewPoints not covered in the structured phase of the interview. A tour in the actual workshops and plants constituted the second phase of the visit. For the analysis of data collected and testing of hypotheses the Wilcoxon matched-pairs signed rank test was primarily employed. In some cases where data type was not amenable to evaluation by that test, the sign test was used. In reporting the limitations of this study two factors may be pointed out. As far as accuracy of data is Concerned, the writer had to depend on the data provided 168 by the companies except for his personal inspections. Of course, twenty-two firms is a limited sample. Neverthe- less twenty—two should not be considered a small sample for this type Of investigation. The third chapter dealt with review of the litera- ture with respect to variables included in this study. The present status of occupational accident research was then discussed and appraised. Chapter IV constituted the reporting of study findings and was composed of three major sections. In the first section specific background data and statistics such as number of employees, company accident frequency and severity rates and firms' membership in safety organization were reported. The second section dealt with the results of testing the hypotheses of this study. Major findings of the study can be summarized as follows: 1. This study strongly supported the hypothesis that in the firms with better safety records, tOp management is highly interested and in- volved in the company's safety performance. 2. Formal educational level of the workers in the firms included in this study did not correlate with accident rates of these companies. 3. Employees' age was found to be significantly 169 related to the firms' accident rates. The higher the age of employees on the average, the lower their work accident experience tends to be. Where a company provided recreational programs and facilities for its employees, work acci- dent rates tended to be lower. The discussion noted that where a company cares about the recreation of its employees, their health and safety will hardly be ignored. The hypothesis that firms with better safety records have tried to promote safety through the employee's family was not supported by this study. This is not to say that this technique of promoting safety is insignificant, but most of the firms studied did not use it because they did not have a knowledge of it. Safety rules were not found to be a signifi- cant factor affecting firms' safety records. In the firms with more married employees on the average, work accident rates were consid- erably lower. This study supported the hypothesis that a posi- tive correlation exists between low accident 10. 11. 12. 13. 170 rates and higher average employee length of service with a company. This implies lower turnover rates and a more stable employment situation which, in turn, tends to minimize the number of new employees to be hired. The hypothesis that older companies tend to have better safety records was not confirmed by this study. Among the companies included in this study, those with better accident record keeping systems were found to have considerably lower work accident rates. In the companies where the first-line super- visor on the average has a wider span of control -- namely more men to supervise -- accident rates tend to be higher. Data obtained proved inadequate to test the hypothesis that relative age of the production machinery and equipment is correlated to accident rates. Physical workplace conditions were found to be significantly related to occupational accident rates. Firms with more desirable temperature, ventilation, lighting, noise level, roominess, 14. 15. 171 and cleanliness tend to have considerably lower work accident rates. The hypothesis that better and more numerous safety devices on machinery contribute to attainment of better safety performance records of the companies, was strongly con- firmed by this study. This study did not support the hypothesis that firms with better safety records have estab- lished safety committees through which union and/or employees aid and advise management on matters of employees' safety at work. This does not show that employee participation is not a significant factor. Safety committees are only one form of participation, one which might perhaps be most useful in the larger com- panies where personal contact between manage- ment and employees is difficult. In the firms included in this study other types of partici- pation such as employee Opinion surveys, sugges- tion systems and personal contacts, which en- able all of the workers to participate in safety issues, appear to be as significant as commit- tees if not more so. 172 Some Conclusions Good work accident experience is not a chance occurrence. It results from an organized movement directed toward accident control. It appears from this study that a safety movement must be initiated and led by tOp management of a firm. T0p management's support and involvement was one of the major factors included in this study and was found to be significantly correlated with the work accident rates of the firms. When other significant factors of the study are con- sidered, it appears that they are also direct results of tOp management's interest and support. If the first factor does not exist in a firm, it is hard to believe that most of the other features such as good record keeping, orderly and safe working conditions, recreational programs, safety devices and so on will exist. Let us look at employee length of service with the company, as another example. It does not appear to be closely related to tOp management's support of safety programs. Even in this case there is probably a relationship between these two factors. Workers will be less inclined to stay long with a company where its manage- ment is careless about their safety. This is likely to be particularly true of most of the married and older workers, who want more security at the job, and who commonly have 173 the best safety records. If tOp management does not show its continuous and strong interest in adequate physical conditions of the workshOp, this significant ingredient of good safety performance will hardly be added to the com- pany's Operations. Accident cost figures at least should motivate tOp management of companies toward accident con- trol efforts even where they might otherwise hesitate to devote adequate effort and money to the prevention of un- necessary accidents and the saving of human lives. In most of the companies studied the actual cost figures except for insurance premiums are not known to tOp manage- ment. The Simonds method of cost analysis (now in use in the educational program of the Michigan Bureau of Safety) is another potential motivator and basis for managerial decision making. Small size companies do not necessarily have to suffer from high work accident rates. This study included eleven small size firms with relatively low accident rates, and empirical investigation revealed that whenever these rates were low, tOp management's involvement, interest and support of the firm's occupational safety was higher. Occupational safety is like a wheelbarrow that without constant pushing will not move. One of the prime responsibilities of tOp management must be to provide that 174 needed push. Some Recommendations The following suggestions occur to the writer as a result of the experience of conducting the research, but are not actual findings of the study. Unless facts on occupational safety are collected, integrated and analyzed, is not likely that effective measures can be taken to re- duce or minimize work accidents at the state level, industry level or in a single firm. It is in recognition of this important fact that this writer would recommend enforcement of the provisions of Section 15 of the Michigan Occupational Safety Standard Act (Act 282). While preparing and sub- mitting required reports of work accidents to the State Department of Labor, company managements will necessarily become somewhat acquainted with their overall safety pic- ture. Availability of such tangible information may moti- vate a more serious consideration for safety in the com- panies. At the same time analysis of accident reports will enable the State Department of Labor to direct and concen- trate its occupational safety efforts to the types of accidents which appear to occur more often and also to identify industries and firms within each industry that are in need of more attention. 175 As appears evident from the findings of this and other studies, the young and new employees tend to have considerably more work accidents than older employees with experience. This being the case, it is recommended that safety training be emphasized in the later years of high school so that before a prospective employee starts a career, he is well aware of work hazards. Most of the firms studied claimed that they could not afford an exten- sive safety training program for a newly hired employee before he starts working. Some firms added that they con- ducted such formal safety training programs but upon termin- ation of the training, most of the trainees left the firm, leaving only some cost figures for the company. If safety training takes place in the high schools, then it does not matter where the workers will be ultimately settled for working. Of course, this would be a joint effort and would require COOperation of educational institutions, the State Department of Labor, mass media, business organizations, the State Department of Public Health and any other indi- viduals or organizations interested in the matter. A suggestion is also made for application in govern- ment procurement, particularly cost—plus contracts. In determining the producer's cost his accident experience could be taken into account. When his record of injuries 176 significantly exceeded a reasonable figure for the kind of industry, a calculation of his excess costs resulting from the excessive accidents could be made, using the Simonds method. Then these unnecessary excess costs could be sub— tracted from his cost figures that were to be paid by the government agency. Further investigation is recommended in the following areas. There were fifteen variables or factors included in this study. Certainly there are other factors that might have a considerable impact on occupational accidents of the firms. Two specific variables are recommended for empirical investigation. The first variable is the actual plant location. In the course of this study it was noticed that firms located in small towns and townships where the majority of workers were from the same small, quiet and friendly community and generally were relatives or acquaint- ances, tend to have much lower accident rates than do firms in dissimilar conditions. This observation provides an excellent ground for a hypothesis to be empirically tested in the future. Another variable recommended to be scrutinized is the nature of workforce composition in the firms with re- spect to race, national origin and religion. 177 This writer's Observation in the twenty-two firms studied Opens ground for suggestion that racial composition of the firm's workforce might correlate with safety perform- ance. In particular, those companies employing larger num- bers of negroes revealed a marked tendency toward higher accident rates. Such an observation requires further empirical investigation as well as a major caveat. That is, the obvious coincidence of certain racial and environ- mental factors could easily result in a spurious correla- tion between racial variables and safety factors. It is also recommended that this same study be con- ducted in different states with different sets of industries and also in different cultural environments. If the findings of this and the recommended studies coincide, it will be possible to make more universally applicable generaliza- tions. BIBLIOGRAPHICAL ESSAY Books American Public Health Association, Accident Prevention, The Role of Physicians and Public Health Workers (New York: McGraw-Hill BoOk Company, Inc., 1961). Bird, Frank E., and Germain, George L., Damage Control : A New Horizon in Accident Prevention and Cost Imptovement (American Management Association, 1966). Blake, Ronald P., Industrial Safety (Englewood Cliffs, N.J.: Prentice-Hall, Inc., Third Edition, 1963). Browne, R. C., Health In Industry (London: Edward Arnold Publishers LTD: 1961). Carzo, Rocco, Jr., and Yanouzas, John, Formal Orgpnizations: A Systems Approach (Homewood, Illinois: Richard D. Irwin, Inc. and The Dorsey Press, 1967), pp. 44-48 and 79-96. Carvell, Fred J., Human Relations In Business (New York: The Macmillan Company, 1970). Chapanis, Alphonse, Man-Machine Engineering (Belmont, Ca1if.: Wadsworth Publishing Co., Inc., 1966). Costello, Timothy W., and Zalkind, Sheldon 8., Psychology in Administration - A Research Orientation (Engle- wood Cliffs, N.J.: Prentice-Hall, Inc. 1963). Davis, Keith, Human Relations at Work, The Dynamics of Organizational Behavior (New York: McGraw-Hill Book Co , 1967). DeReamer, Russell, Modern Safety Practices (New York: John Wiley & Sons, Inc., 1958). Farmer, Eric, The Causes of Accidents (London: Sir Isaac Pitman & Sons LTD., 1932). Ginzberg, Eli, The Unemployed (New York: Harper & Row, 1934). 178 179 GoldthrOp, John H. et al., The Affluent WOrker: Industrial Attitude and Behavior (London: Cambridge University Press, 1968). Graicunas, V. A., Relationship in Organization in Gulick Luther and Urwick Lyndall (eds.) "Papers on the Science of Administration," (New York: Institute of Public Administration, 1937), pp. 52-57. Greenberg, A Railing on Your Roof (Haifa, Israel: Publish- ing House of the Student Association, 1969; (A preliminary edition reproduced for personal use, educational and research purposes of Mr. Y. Shafai). Haddon, Williams, Jr. et al; (eds), Accident Research Methods andepproaches (New York: Harper & Row Publishers, 1964). Hands, A. H., Ergonomics in Handly, William (ed.), Indus- trial Safety HandboOk (London: McGraw-Hill, 1969), Hannaford, Earle S., Supervisors Guide to Human Relations (Chicago, Illinois: National Safety Council, 1967). Heinrich, H. W., Industrial Accident Prevention (New York: McGraw—Hill Book Co., Fourth Edition, 1959). Herzberg, Fred, et al., Job Attitudes: A Review of Research and Opinion (Pittsburgh: Psychological Service of Pittsburgh, 1957). ' Judson, Harry and Brown, James, Occupational Accident Prevention (New York: John Wiley & Sons, Inc., 1944), pp. 90-100. King, H. E., An Age Analysis of Some Agricultural Accidents in Haddon, William et al (eds.) Accident Research and Approaches (New York: Harper & Row, 1964), pp. 41-47. Maier, Norman R. F., Psychology In Industry (Boston: The Riverside Press, 1955), Chapters 17-18. 180 March, James, Handbook of Organizations (Chicago: Rand McNally and Co., 1965), pp. 398-399. McFarland, Ross A., The Epidemiology of Industrial Accidents (Harvard: School of Public Health, 1965). McCormick, E. J., Human Factors Engineering (New York: McGraw-Hill BoOk Co., 2nd Edition, 1964). Meggison, Leon C., Personnel - A Behavioral Approach to Administration (Homewood, Illinois: Richard D. Irwin, Inc., 1967), pp. 113-115. Miller, Delbert C., Handbook of Research Design and Social Measurement (New York: David McKay Company, Inc., 1964), pp. 76-78. Pigors, Paul and Myers, Charles A., Personnel Administration, A Point of View and a Method (New York: McGraw- Hill Book Co., Fifth Edition, 1965), pp. 567-594. Righby, Paul H., Conceptual Foundations of Business Research (New YOrk: John Wiley & Sons, Inc., 1965), p. 189. Shils, Edward A. and Janowitz Morris, Cohesion and Disinte- gpation in The Wehrmacht in World War II, in Hill, Walter and Egan, Douglas (eds.) Readings in Organ- ization Theory, A Behavioral Approach (Boston: Allyn and Bacon, Inc., 1968). Siegel, Sidney, Non-Parametric Statistics for the Behavioral Sciences (New York: McGraw-Hill Book Co., Inc., Simon, Ju1ian L., Basic Research in Social Science - The Art of Empirical Investigation (New York: Random House, Co., 1969), p. 252. Simonds, Rollin H. and Grimaldi, John V., Safetpranagement (Homewood, Illinois: Richard D. Irwin, Inc., Revised Edition, 1963). Surry, Jean, Industrial Accident ResearcthA Human Engineer- ing Appraisal (Toronto: University of Toronto, 1969). 181 Tannenbaum, Robert, et al., Leadership and Organizations: A Behavioral Science Approach (New York: McGraw- Venn, Grant, Man, Education and Work (Washington, D.C.: American Council on Education, 1964). Vernon, H. M., Accidents and Their Prevention (London: Cambridge University Press, 1936). Wallis, Allen, and Roberts, Harry, Statistics: A New Approach (Glenco, Illinois: The Free Press, 1956), p. 598. Periodicals and Articles Browning, Rufus C., "Opinion Survey Motivates Safety," Personnel Journal, Vol. 38, No. 10, March 1960, Burns, Robert L., "Participative Safety: A Motivating Factor," Supervision, Vol. XXXII, No. 1, Jan. 1970, pp. 3-4 and 14. Bursill, A. E., "The Restriction of Peripheral Vision During Exposure to Hot and Humid Conditions," Qparterly JOurnal of Experimental Psychology, Vol. 10, August 1958), pp. 113-129. Caleo, Robert L., "The Scientific Approach to Safety Administration," Administrative Management, Vol. XXIV, No. 10, October 1963, pp. 46-50. Crawford, Paul L., "Accident Prevention Through Scientific Selection," Personnel JOurnal, Vol. 44, No. 10, Nov. 1965, pp. 560-62. Feree, C. E. and Rand G., "Good working Conditions for Eyes," Personnel JOurnal, No. 15, 1937, pp. 333- 346. Forsgren, Roderick A., "A Model of Supportive Work Condi- tions Through Safety Management," Personnel JOurnal, Vol. 48, No. 5, May 1969, pp. 351-58. 182 Graig, Eugene A., "Situational Stress and Safety," Personnel Journal, Vol. 45, No. 5, May 1966, pp. 269-272. Harris, Frank, "The Quantification of an Industrial Employee Survey, 1, Method," Journal of Applied Psychology, Vol. 33, 1949, pp. 103-111. Inlow, Gail M., "Job Satisfaction of Liberal Arts Graduates," JOurnal of Applied Psychology, Vol. 35, 1951, pp. 178-181. Iron Age, "Rise in Plant Accident Costs Spurs Industry Safety Drive," Iron Age, Vol. 188, No. 18, Nov. 2, 1961, p. 41. Keenan, Vernon, et al., "Psychological Climate and Acci- dents in An Automotive Plant," JOurnal of Applied Psychology. Vol. 35, No. 2, 1951, pp. 108-111. Kinnane, John F. and Pable, Martine W., "Family Background and Work Value Orientation," JOurnal of Counselipg Psychologyo Vol. 9, Nov. 4, Winter 1962, pp. 320-25. Kornhauser, A. W., "The Effect of Noise on Office Output," Industrial Psychology. No. 2, 1927, pp. 621-622. Mihlon, Laurence F., "What is Wrong with Safety Management," Factory, Vol. 119, No. 91, Sept. 1961, p. 241. National Safety Council, "50 Years Ago .... What They Were Saying About Safety," National Safety News, Nov- 1969, p. 36. National Safety Council, "12,000,000 Accident Free Hours -- How's it Done?" National Safety News, Vol. 68, No. 2, August 2, 1962, pp. 30-31 and 57. Okrongley, W. D., "Attitude DevelOpment and Accident Prevention," Personnel Journal, Vol. 45, No. 3, March 1966. PP. 169-171. P0pe, William C., "Computers In Safety Management," National Safeterews, May 1970. PP. 49-54. 183 Pyle, Howard, "National Safety Council Position on the PrOposed Occupational Safety and Health Legisla- tion," National Safety News, Dec. 1969, pp. 38-42. Ranson, W. R., "Management's Role in Safety," National Safety News, October 1969, pp. 51-52. Shaw, JOseph R., "Safety in Changing World," National Safety News, October, 1969, pp. 48-50. Simonds, Rollin H., Is Organization "Structure Reflecting New Techniques and Theory"? M.S.U. Business TOpics, Vol. 17, No. 3, Summer 1969, pp. 65-71. Special Report, "The Key to Safety," Personnel Magazine (U.K.) October 1967, pp. 24-25 and 48. Stallcup, Evan, "A Fresh Look at the Safety Program: When Enough is Too Much," Personnel, Vol. 38, 1961, pp. 26-37. Stubbs, William H., "How a Mill Safety Program Cuts Injuries," Pulp and Paper, Vol. 43, No. 3, March, 1969. pp. 146-150. VanZelst, R. H., "Effect of Age and Experience on Accident Rates," JOurnal of Applied Psychology, Vol. 38, 1954, pp. 313-317. "Ergonomics — New Angle on Employee Health and Safety," Management Review, Vol. 56, No. 12, Dec. 1967. PP. 47—50. Public Documents and Government Publications Chaney, L. W., and Hanna, H. S., The Safety Movement in the Iron and Steel Industry, 1907-1917 (U.S. Bureau of Labor Statistics, Report No. 34). Michigan Department of Labor, Annual Rpport Fiscal Year 1965-66, Vol. 1, 1966. Michigan Department of Labor, Annual Report Fiscal Year 184 Michigan Department of Labor, Annual Rsport FiScal Year 1968-69, Vol. 4, 1961. Michigan Department of Labor, Michigan Work Injuries COOperative Survey 1968. Michigan State, Occupational Safety Standard Act, Act 282, 1967. Simonds, Rollin H., Estimating Costs of Industrial Acci- dents, U. S. Department of Labor (Washington D.C.: U.S. Government Printing Office 1955). United States Bureau of the Budget, Standard Industrial Classification Manual, 1967. United States Department of Labor, Controllinngoise Hazards, Safety In Industry Bulletin No. 27 (Washington D.C.: U. S. Government Printing Office, 1959). United States Department of Labor, Bureau of Labor Standards, Safety Subjects, Bulletin No. 67, Revised 1956 (Washington, D.C.: U. S. Government Printing Office, 1958). Other Sources AFL—CIO Industrial Union Department, DanggrmoMen at WOrk (Washington Industrial Union Department, 1969). Greenwoods, M. and Woods, H. M., A Rsport on the Incidence of Industrial Accidents Upon Individuals with Special Reference to Multiple Accidents (London: British Industrial Fatigue Board, Report No. 4, 1919. International Labour Office, Accident Prevention - A Worker's Education Manual (Geneva: "La Tribune De Geneve," 1961). International Occupational Safety and Health Information Center, Ergonomics of Machine Guarding (CIS Infor— mation Sheet 10 Geneva 1964). 185 JOnes, Don P., Human Factor - Occupational Safety, A Report to the Labour Safety Council of Ontario, Department of Labour, (No date). Michigan Manufacturers and Financial Record, The Directory of Michigan Manufacturers (Detroit: Manufacturing Publishing Co., 1969). National Safety Council, Accident Facts, Chicago, Illinois, 1965. National Safety Council, Accident Facts, Chicago, Illinois 1966. National Safety Council, Accident Facts, Chicago, Illinois 1967. National Safety Council, Accident Facts, Chicago, Illinois 1968. National Safety Council, Accident Facts, Chicago, Illinois 1969. National Safety Council, Accident Facts, Chicago, Illinois 1970. Newbold, E. M.. A Contribution to the Study of the Human Factor in Causation of Accidents (London: British Industrial Health Research Board, Report No. 34, 1926). Osborne, E. E., et al., Two Contributions to the Study of Accident Causation (London: Industrial Fatigue Research Board Report No. 19, 1922. Pikulinski, Jerome R., Protective Labor Legislation In Michigan a Comparative Study (Institute of Labor and Industrial Relations - The University of Michigan - Wayne State University, Nov. 1968). Roehe, John M., Safety and The Foreman (New York: National Foreman's Institute, Inc., 1951). Sands, Paul E., Accident Prevention and Governmental Control in the Construction Industry In Michigan and Ohio, An unpublished doctoral dissertation, Michigan State University, 1964. 186 Surry, Jean, An Annotated Bibliography_For Industrial Accident Research and Related Fields (Labour Safety Council of Ontario, April 1969). United States of America Standard Institute, USA Standard Method of Recording and Measuring Work Injury Experience 2 16.1-1967. 1='———.—*—'—“‘J APPENDICES APPENDIX A Letter Sent to the Director of Bureau of Safety Regulation Michigan State Department of Labor MICHIGAN STATE UNIVERSITY EAST LANSING . MICHIGAN 48823 COLLEGE OF BUSINESS DEPARTMENT OF MANAGEMENT ' EPPLBY CENTER June 16, 1970 Mr. Marshall Fiordelis Director, Bureau of Safety and Regulation Michigan Department of Labor 300 E. Michigan Avenue Lansing, Michigan 48913 Dear Mr. Fiordelis: One of our doctoral candidates, Mr. Yaghoub Shafai from Persia, expects to write his dissertation in the field of industrial safety under my direc- tion. His current plans call for selecting eleven pairs of companies, each pair being composed of two concerns similar in size in the same industry but with one having much higher frequency and severity rates than the other. Then he would secure permission from the companies to gather in- formation that might shed light on why the accident experience in one was better than in the other. We thought someone in your organization might be able to suggest such com- panies. If you have the actual injury rates, this would be helpful, but not necessary. It will be his job with my assistance to obtain permission from the companies. He will probably need more than twenty names to start because some will likely not find it convenient to let him get the informa- tion. Incidentally, he will not be interested in the "touchy" data such as sales, profits, wages, etc. ' Mr. Shafai will contact the Bureau. If it is not practical for your office to give him names of suitable companies, he will probably have to select at random and hope to get companies with markedly differing injury rates. If you are able to give him some suitable names, we shall appreciate it. If you have other suggestions to offer, please feel free to do so. Sincerely, /’ a - I: ,_ . 5: {JC [i/(CI/f/M/é47f‘Kfl-{CL Rollin Simonds Professor of Management cc: Mr. Yaghoub Shafai Professor Penfield Professor Kruger 187 APPENDIX B Letter Sent to the Companies as the Initial and First Contact MICHIGAN STATE UNIVERSITY EAST LANSING . MICHIGAN 4882} COLLEGE OF BUSINESS DEPARTMENT Ol’ MANAGEMENT 0 EPPLEY CENTER Date Company Name (Company President) I am writing to you in connection with a research study to be conducted at Michigan State University. The purpose of the research is to find out what are some of the key factors that cause apparently similar business concerns to have markedly different accident eXperience. Higher work injury frequency and severity rates, of course, mean both employee hardship and loss of profit. All business concerns would naturally prefer to keep such control of their Operations as to minimize accidents. It is posSible, however, that we may not appreciate the part some factors play in this. In hOpe of shedding further light on this, I have selected eleven pairs of industrial concerns, each pair comprising firms of similar size in the same industry, but with one of the two having considerably more injuries than the other. Yours is one of these selected companies. I hOpe to secure permission to visit these companies and compare them in terms of many factors to see if some correlate with high injury frequency and others with low. I shall not be concerned with costs or profits Or wages and shall not reveal the names of the companies studied, although I shall make a summary of my findings available to the companies COOperating. I would appreciate it very much if you would permit me to visit your company and talk with a few of you peOple to get the information I need. I shall conduct this study as a Ph.D. dissertation under the supervision of Professor Rollin Simonds of the Graduate School of Business, who is one of the inter- nationally known eXperts in the field of safety. I shall telephone your office in a few days hOpefully to receive your permis— sion and to answer any questions you may have in connection with the study. Sinc ely 6 Y ghoub Shafai Doctoral Candidate 188 DATE 1. 1. 10. ll. 12. APPENDIX C Questionnaire Used as Interview Guide CODE NO. INTERVIEW GUIDE GENERAL INFORMATION A. -Company Name of Company .When established Type of ownership Address Telephone number Distance from Lansing Type of Business Industry Average number of employees Is company member of any safety organization? Yes No If yes, names of such organization Name of insurance company or actuary Agent B. Interviewee Name of the interviewee Title and position of the interviewee Number of years in this position Number of years spent with the company 189 10. 11. 190 CODE NO. Is he specifically responsible for company's safety? Yes No Does he have regular education and/or training in safety? Yes No Age of the interviewee: 18-23 24-28 29-34 35-40 41-46 47 and over 18 he a member of any safety organization? Yes No If yes, name of such Organization Do you attend any safety meetings and conferences? Yes No If yes, specify which meetings and conferences? If no, do you have a desire to be a member of a safety organization? Yes No Top Management Yes (1) No (0) Does he attend any safety committee meetings in the company? Does he chair any of these meetings Does he regularly receive safety reports? Does he personally conduct any safety audit or inspection? Is he a member of any safety organization? Does he regularly attend any safety meetings or conferences outside the company? Does he emphasize plans for achieving certain safety objectives? Does he actively participate in execution of safety plans? Does he hold review and analysis sessions in order to compare the results of carrying out safety plans with projected objectives? 191 CODE NO. Is safety figures, reports, achievements included on the agenda of board meetings? Employees Average number of employees Male Female Average number of years of regular education? 1-3 4-6 7-9 10-12 13 and over Average age of the employees: 18-23 24-28 29-34 35-40 41-46 46 and over Average number of married employees % of total emp. Male Female Number of employees with higher regular education than average? Number of employees with lower regular education than average? Average number of years spent with company: 1-3 4-6 7-9 10-12 13-15 16 and over Nature of racial composition? Black % or No. White % or No. Other % or No. (Specify) . Nature of religious composition: Catholic % or No. Protestant % or No. Other % or No. (Specify) 192 CODE N09 10. Nature of Nationality origin composition: % or No. %L_______ or No. %L_______ or No. %L________or No. 11. Nature of birth place and residence composition: Local: % or No. non-Local: % or No. 12. Do you have handicapped employees? Yes No No. E. ngervisors (first line) 1. Average no. of the supervisors: Total Male Female 2. Average no. of employees under direction of each supervisor: 1-5 6-10 11-15 16-20 21-25 26-30______ 31-35______ 36-40_______ 41 and over 3. Average age of first line supervisors: 18-22______.23-27______ 28-32___ 33-27__ 38-42___ 43-47____ 48-52____ 53 and over 4. Average number of years spent with the company? 1-3 4-6 7-9 10-12 13-15 16-18 19-21______ 22 and over______ F. Machinery and Eguipment 1. What % of production is automated: 0-10%L______ 10-20%L_____, 2 1-30%___ 31-40%___ 41-50%___ 51-60%___ 61-70%L_____ 71-80%______ 81-90%______ 91-100%L______ 2. Last time that major production machinery were replaced by newer ones was: (check how many years ago) 1-3 4-6 7-9 10-12 13-15 over 15 193 CODE NO. Do you provide at no cost, safety equipment to the employees? Yes No SAFETY Safety Organization Do you have a separate organizational unit responsible for safety of the company? Yes No If yes, how many people are full-time in this unit? Where is this unit located in the firm's organization structure compared to other functions? To whom does it report directly? Pres. Vice-Pres. Plant Mgr. I Personnel Mgr. Other If answer is No for Question 1 above, which organizational unit is basically responsible for company safety? Why has this unit been given safety responsibility §gfety Policigpjang Propgdures Do you have written safety policies? Yes No Who has initiated these policies? Top management Safety Dept. Personnel Dept. Other Have there been any changes in these policies in the past 5 years? Yes No DO you have an established communication channel for reporting, analysis of these reports and decision making on the subject matters pertaining to safety? Yes No Are employees or their representatives consulted in making these policies and procedures? Yes (1) No(0) e "‘-"W 194 CODE NO. C. Accident Record Keeping 1. Do you have a written accident record keeping procedures? Yes (1) No. (0) Does this record keeping system show cause place time cost reason people involved in the accidents? 1 2 3 4 5 6 (each Yes-(1) No=(0) ) Is any periodical report made from these records? Yes (1) N0 (0) If yes, how often? monthly bi-monthly quarterly semi-annually annually 1 2 3 4 5 Who receives these reports? 1 2 3 4 5 Why? Saf ty R l e u es Yes (1) No (0) Do you have written safety rules for all operations? Were employees or their representatives, involved in making these rules? Were first line supervisors involved in making these rules? Have there been any changes in these rules in the past 5 years? What was the reason for these changes? 1 2 3 How do you make sure that all employees have complete knowledge of these rules? 1 g_, 3 Who is responsible for enforcement of these rules? 1 2 3 Are penalties resulting from violation of these rules publicized? Yes No 10. 11. 12. 13. 14. 15. 16. 195 CODE NO. If yes, how? Penalties for violation of rules varies from to Are these rules adopted from another source or organization? Yes No If yes, what is the name of that source? What kind of organization is this source? Business educational governmental other (specify If rules are not adopted and are made specifically for the company, was job analysis used as a basis for making these rules? Yes (1) N°(0) Do you have any knowledge of safety rules in other firms in this industry? Yes (1) No (0) If yes, how would you compare yours with theirs? better (3) same (2) other (specify) (1) Hgfety Programs and Practices Safety training a. Do you have safety training programs? Yes No b. Who is responsible for administration of these programs? c. Do you have safety training for all new employees before they start work? Yes No d. Do you have specific training programs for individual employees, who need personalized training in order to overcome their individual deficiencies? Yes No 196 CODE NO. DO you seek assistance from other organizations for safety training? Yes No If yes, from which organizations: federal insurance companys government state local educational institutions national safety council other (specify) unions In which one of the following cases do you think employees need a safety training and in which cases actually do you provide such a training? Training Training Needed Conducted New employees Changed job Expanded/Different responsibilities New machinery New equipment New process Changed layout or physical condition Occurrence of an accident Do you have special safety training programs for first line supervisors? Yes No Where does your safety training program take place? Yes 1H3 In the company By company personnel By other than company personnel By both Outside the Company Both in and outside the company j. 197 CODE NO. Which of the following training techniques are used for safety training? Most Effective Techniques Technique Actually used Lecture Conference Lecture-Conference Case study Film Role playing 2. Safety Committee Do you have established a safety committee(s)? Yes (1) N0 (0).— If yes, in which of the following levels? 1) Central Safety Committee--Consist of top management with department heads? Yes (1) No (0) If yes, how often 2) Departmental safety committee--consisting of department heads, supervisors and the foremen? Yes (1) No (0) If yes, how often? 3) Foremen's safety committee--inc1ude all men? Yes (1) N0 (0) .____ If yes, how often? 4) Union-Management committee--consist of top management or his representative(s), union member(s), and employee representative(s)? Yes (1) No (0) If yes, how often? When was this committee established? Work accident experience in the company, after safety committee(s) were established has: considerably improved improved little no change worsened considerably worsened ' 311'???" 198 3. Safety Inspections CODE NO. a. Do you have regularly conducted safety inspections? Yes No b. If yes, whose responsibility is it to make inspections? c. How often are these inspections conducted? d. Who receives reports of the results of these inspections? e. Which one of the following ratings explain best the effect of safety inspections in preventing work injury in your company? very effective no somewhat effective effect at all negative effect Safety Promotion and Propaganda a. Do you practice promoting employees' safety interest? Yes No b. Which of the following are most or least emphasized in this ctice? (in order to influence the actions of employees?) pra 1) 2) 3) 4) 5) 6) 7) Fear of personal injury Fear of economic loss Desire for reward Desire for leadership Desire to excel and be outstanding Protection of others Creating a favorable impression 4 It” 8. 199 CODE NO. c. Which one of the following methods do you use least or most in romotin safet interest: p g y least Most 1) Safety meetings--genera1 and "stand-up" 2) Safety contest 3) Demonstrations 4) Displays and exhibits 5) Posters, signs, slogans 6) Publications Manuals rule books payroll inserts Periodicals Handouts 7) Promoting through employee families OTHER PROGRAMS Recreational Facilities and Programs Yes(1) No (0) Have you provided recreational facilities for employees? Are employee families authorized to use these recreational facilities? Is there athletic teams in the company? Does the company encourage these teams to par- ticipate in games and competitions with other teams? Have company teams ever won a championship position? DO you believe that friendly relationships among workers will help to create a favorable work environment in which employees may be more safety conscious? Do you have company sponsored social gatherings for employees, management and their families? If yes, how often and what is the nature of these gatherings? 10. l 200 CODE NO. ~ What % of employees and their families utilize recreational facilities and participate in these programs? .Employees: under 25% 32:18:; m...— 25% 25% 50% 50% 75% over 75% 75% over 75% What have you done to create this friendly relationship among employees? Medical Provisions Do you have the following medical personnel and facilities in the company? Doctor Yes(l)_____ Nurse _ Hospital _____ Dispensary ‘____ Ambulance Do you have first aid Have you provided first aid training for: all employees first line supervisors other (specify) No(0) facilities? Full- time Yes (1) Yes (1) Part- time YeS'(1) No (0) No (0) Not in-com- pany but Number arranged No (0) Do you have regular medical examinations for employees? If yes, how often? Where? COMPANY'S SPECIFIC WORK INJURY INFORMATION AND STATISTICS Frequency and Severity Rates Are you familiar with work injury frequency and severity rates? Yes NO Do you know how they are figured out? Yes No 91:11.5!" 10. 201 CODE NO. Do you know your company's frequency and severity rates? Yes No Do you know industry's average frequency and severity rates? Yes No Total hours worked in 1968 1969 A Total number of disabling injuries in 1968 1969 Z Total number of compensable injuries in 1968 1969 Reported to the government in 1968 1969 Over seven days 1968 1969 3 Total days lost in 1968 1969 " Total days lost (compensable only) 1968 1969 Number of fatalities (if any) in 1968) 1969 Which year do you consider as the best for the company's safety record? If 1969 safety record is different than best year, what are the factors that this difference can be attributed to? Causes of Acgidents How would you rate the following with regard to their effect in causing unsafe employee acts which result in work accidents in your firm? ( ) a. Physical inadequacies (includes poor eyesight, defective hearing, muscular weakness, heart, circulatory, or other organic weakness.) ( ) b. Mental inadequacies (includes slow mental reaction, lack of coordination and nervous instability.) ( ) c. Faulty attitude (includes indifference, inattention, indolence, arrogance, recklessness, hostility.) V. 10. 202 CODE NO. ( ) d. Lack of knowledge or skill (includes ignorance of correct methods, faulty work habits and insufficient experience.) How would you rate the following unsafe actions with regards to their frequency of happening in work accident cases of your company? very very high high low low 8. Making safety devices inoperative b. Using unsafe equipment c. Unsafe loading, mixing, or placing d. Unsafe position or posture e. Failure to use safe clothing or personal protective devices f. Operating equipment at unsafe speeds g. Distracting, teasing, or abusing other workers OBSERVATIONS (to be filled out immediately upon leaving company) Physical Plant Layout and Conditions Passage ways were: Obstructed 1 2 3 4 5 Clear Passage ways were: Unmarked 1 2 3 4 5 Clearly marked Overall plant layout was: cluttered l 2 3 4 5 clean & rOomy P l ftl :“°t 1 id roper p aces or 00 s were provided 2 3 4 5 prov de Tools were: not pro erly 1 2 4 were pro erl placeg 3 5 places y Machinery were: ungarded 1 2 3 4 5 properly guarded Working surfaces were: oily, slippery 1 2 3 4 5 clean and not clear & dry Working surfaces were: uneven 1 2 3 4 5 even Overall lighting of the plant was: insufficient 1 2 3 4 5 sufficient Temperature was: uncomfortable 1 2 3 4 5 just right Ventilation was: inadequate 1 2 3 4 5 adequate ._ _:'.~"' 11. 12. 13. 14. 15. 16. 17. 203 CODE NO. Noise in the plant was: high 1 2 3 4 5 low Sanitary facilities were: poor 1 2 3 4 5 very clean Visibility in the working area was: poor 1 2 3 4 5 very good Suitable containers for were not were spilled and leaking Oil: placed 1 2 3 4 5 placed p under the barrels containing lubricating oil in engine rooms. Emergency exit signs were: not easily 1 2 3 4 5 seen Fire extinguishing equip- not ment was: available 1 2 3 4 5 _ not Fire alarm system was: installed 1 2 3 4 5 Interest in Safety and this Project out- standing fair Degree of their cooperation was: were eaSIly seen available L installed very poor poor Company's interest in safety is: Company's safety practices seemed: APPENDIX D Letter of Authorization Obtained from Companies Mr. Marshall Fiordelis Michigan State Department of Labor Bureau of Safety Regulation Lansing, Michigan '_.‘.I~‘IV Dear Mr. Fiordelis: We are participating with Michigan State University and Mr. Yaghoub Shafai on a safety research project in industrial firms. We hereby authorize him to have access to our safety figures and records. Sincerely, (Company) 204 APPENDIX E Statistical Tests Used For The Analysis of Data The Sign Test The sign test is one of the non-parametric methods which is used to test the significance of the difference between two means in a paired experiment. It is particularly useful for research in which quantitative measurement is impossible or infeasible, but in which it is possible to rank with respect to each other the two members of the pair.1 The null hypothesis tested by this method is that: P(XA7XB) = P(XA< X3) = 35 Where XA is the score of one member of the pair and X3 is the score of the other member in a matched-pair. The above equation implies that mean difference of two scores of the matched-pair is zero. In the sign test, as implied by its name, only the sign of the difference between the paired variates is used and attention is focussed on the direction of the differences noting whether the sign of differ- ence is plus or minus. If the difference between scores of matched- pair is zero thereby that pair is dropped from the analysis and N (the number of pairs) is reduced. The number of fewer signs (X) is determined by counting. Then, by referring to a table, the probability level is found for respective number of observations with differing scores. __7 1 Sidney Siegel, Non-parametric Statistics for the thavioral Sciences (New York: MCGraw-Hill Book Company, 1956), pp. 68-75. 205 7'11:va 1206 The sign test may be useful, if many of the differences are ties.2 If the number of occurrences of the less numerous of the two signs is (S) and the total number of signs -- that is, the number of pairs of observations less the number in which the difference was zero -- is N, then by using the following formula ZS + l K=T-/fi' '. KL" the value of K is determined. By subsequent referent to a related table the probability level can be determined. The Wilcoxon Matched-pairs Signed-Ranks Test The sign test discussed takes into consideration the direction of differences between mean scores of two members in a matched-pair but does not show the magnitude of difference. The Wilcoxon matched- pairs signed-ranks test not only utilizes the direction of differences, but it also gives more weight to a pair showing a large difference between the two members than to a pair showing a small difference. It has been suggested that the Wilcoxon test is a most useful test for the behavioral scientists.3 With the data of behavioral nature, the researcher can identify the member of a pair which is greater than the other member namely the direction of difference be- tween the two. He can also rank the differences in order of absolute size. The procedure for using this test as follows: 1. List the score of each member in a pair. 2. .Determine (di) the difference between scores of the two members. 2Allen Wallis and Harry V. Roberts, Statistics - A New Approach (Glenco, Illinois: The Free Press 1956), p. 598. 3Siegel, op. cit., p. 75. 207 Rank the resulting differences in order of size, disregarding sign. Affix the sign of the original difference to the corre- sponding rank. Obtain T, the value of smaller of the sums of the like- signed ranks. Then number of di's (differences with a Sign) "N" is determined by counting. If the two scores of any pair are equal, and the difference is thus zero, these pairs are dropped from analysis. If the difference between two scores of the two or more pairs are the same size, the rank assigned will be the average of the ranks which would have been assigned if the d's had differed slightly. If the number of pairs is 25 or less -- as it was in this study, a table of critical values of T is given for various sizes on N. For "N" larger than 25 that table cannot be used and the following formuls should be employed. T - «T T " N N” z = T = 4 /N®I)L2N+1) 24 where HT = Mban.and 9T = standard deviation. "lllllllllllllll