For this post I will investigate whether the size of the company’s staff affected how likely accidents were to occur. While theoretically it should not, I chose this question as I considered that the larger companies may have had more formalised rules and money to invest in safety devices, which may have improved their safety records. Additionally, in a period before many safety devices were required to be installed by law, when the hours that employees could be made to work were not regulated and when the companies’ managements were becoming more professional, this question goes right to the heart of whether larger, more impersonal companies are better or worse to work for.
I will use two files from the year 1884 to answer this question. The first is the Board of Trade’s accident returns that detailed the number of railway workers killed and injured in the course of their duties.  The second was a return of the numbers of people the railways employed on the 31st March. From these files I have extracted data relating to thirty-one British railways which employed over 1000 people. Twenty-one were English and Welsh, four were Scottish and six were Irish.
From thirty-one railways, employing 351,889 individuals in 1884, there were 508 fatalities (0.14%) and 2,242 (0.64%) injuries. However, before analysing whether company size affected safety, I noticed that the country in which individuals worked in seemed to be a factor in accident rates:-
It is evident that working for a Scottish railway company was rather more dangerous than working elsewhere. In 1884 the four Scottish companies killed 0.24% of their staff and injured 0.80%. This was while on English and Welsh and Irish railways 0.14% of the staff died, and 0.64% and 0.19% were injured respectively. Indeed, this suggests, in 1884 at least, that Irish railways were the safest to work for. The exact cause for Scotland’s higher accident rate is unknown at present, and only more research will give a definitive answer. However, one possibility, which ran through my head, is that the harsher weather of Scotland made working conditions more treacherous.
Turning to the relationship between size and accident rates, there seems to have been no correlation the fatality rate and staff body size. Of the eleven companies that employed over 10,000 people, the largest proportion of fatalities was attributed to the North British Railway, who killed thirty-five of its 13,896 employees (0.25%). The safest railway was the Great Western, who only suffered thirty-four deaths amongst a staff of 39,547 (0.09%). However, a great range of proportions is also found in the case of the smaller railways. Of the ten companies that employed fewer than 2,500 people, the Waterford and Limerick Railway killed four of its 1503 workers (0.27%), the heist proportion of fatalities of any of the thirty-one companies. However, the Metropolitan railway did not kill any of its 1685 staff. Therefore, because of the range of figures within the samples, for both large and small companies, it suggests that other factors were important in determining how many deaths occurred, for example the technologies the companies were using, the training that staff received and the decisions managers made.
However, when examining injuries it is found that the size of the company’s staffs did have a loose correlation with the number of accidents that occurred. Of the eleven companies employing above 10,000 people, the Lancashire and Yorkshire had the highest injury rate with 284 of its 20,962 employees (1.36%) suffering some minor or major accident. However, the safest company in this regard was the Midland Railway, where only 121 of its 43,699 (0.28%) employees were injured. Amongst the ten companies employing fewer than 2,500 workers, the worst to be employed in was Somerset and Dorset where seven employees out of 1089 were injured (0.64%). The best was the Dublin, Wicklow and Wexford Railway, where only two of the 1089 staff members were injured (0.18%). Indeed, the average injury rate for the largest 11 employers was 0.69%, whereas for the smallest ten it was 0.33%. The (exceedingly) loose correlation is shown in the scatter graph below.
Given the very weak nature of the correlation I would not like to venture a suggestion without further study. Indeed, the injury rates should, theoretically, be subject to the same forces as the death rates.
Therefore, in my next post I will examine how the intensity of the different companies’ networks affected their accident rates and whether companies that moved more passengers and goods in 1884 harmed more of their workers.
 House of Commons Parliamentary Papers [HCPP], Return of Accidents and Casualties as Reported to the Board of Trade By the Several Railway Companies in the United Kingdom, During the Year 31st December 1884, p.17
 1884 (242) Railways (number of persons employed). Return of the number of persons employed by each of the railway companies of the United Kingdom on 31 March 1884 (classified according to the nature of the work performed by them); &c.