How Injury Incidence Is Associated with Business Cycle?

為什麼職災、事故件數與發生率高低會和景氣循環有顯著的正相關？ 

職災率高低和景氣循環相關的現象很早就被觀察與察覺

例如： 

 

最簡單的解釋是-

The occurrence of workplace injuries follows a Poisson distribution with the rate of occurrence, defined as number of injuries over a period of time increasing with growth in labor force or population. This could also be affected by economic upturn owing to prolonging working hours or additional workers during those times.

景氣好-工作量大/工時長/工作人數多，自然觀察到的工傷與事故件數就多

 

為了避免以上分母(工時/工人數)差異造成的錯覺與誤會，最簡單的方法就是讓分母大小固定，統一用每百萬工時的發生率來進行比較

To avoid bias due to population size or the number of hours worked, the injury incidence rate should be stated as the ratio of accidents or injuries occurring per million worker-hours, thereby not inflated by business cycle. To determine the underlying causal relationship between these factors, the hypothesis that incidences of workplace injuries are associated with business cycles must first be justified.

 

近一步探討Why之前，要先確認職災率真的跟景氣循環有關

H0: The incidence of workplace injury is associated with business cycle.

 

挑選了製造、營造、批發零售與住宿餐飲等四大行業，以下是其敘述統計

  

 

 

 

用台灣勞保職災給付資料的驗證結果各行業別職災給付高低(Y1是傷病給付率、Y2是殘廢與死亡給付率)是否與景氣指標(產值PV與生產指數PI) 

PS: 就研究分析方法而言，自變數與因變數兩者都取差分，以解決非穩態時間序列間虛假相關的疑慮

  

 

結果顯示：
四大行業的傷病給付率和兩個景氣指標都有很顯著相關；殘廢與死亡給率，只有製造與營造業顯著(批發零售與餐飲住宿業的職災，通常較為輕微，死亡與殘廢的狀況少，數據少，檢定例小)

 

再看各行業的職災率是否月份的差異

結果顯示四大行業的傷病給付率有很顯著的月份差異-二月份期間的傷病給付率低於其他月份，而殘廢與死亡給率則無月份的顯著差異

 

 

 

小結論

輕微職災的發生率，很明顯跟景氣循環成正相關，景氣好，職災率高；景氣不好，職災率

重大職災的發生率，在製造與營造業可以觀察到比較顯著的關連；在服務業(批發零售與餐飲住宿)則關連不顯著

 

 

針對此一現象學者提出來的理論(=猜想)架構

第一組解釋 勞工組成(Labor Composition)

景氣好企業採用大量菜鳥(新人八字輕, LC1)

資深員工在景氣好跳槽(經驗業務沒有傳承交接, LC2)

景氣好企業採用約聘人員(比新人更糟, LC3)

How labor composition contribute to the incidence of workplace injury

The first set of explanatory variable refers to the composition of workforce, including inexperienced workers, experienced workers and temporary workers. As alluded to in Kossoris (1938) and Robinson (1988), new or inexperienced workers are particularly injury prone as they are less familiar with the working environment and operating (Butani, 1988; Keyserling, 1983). Their presence in working force is directly affected by the economic cycles. That is, increasing with economic upturns and declining during recess or depression as they are subject to first wave of layoff when a company decides to lower their expenses. Consequently, changes in the number of inexperienced workers could intensify the association between business cycles and incidences of workplace injury.

Meanwhile, during economic upturns, an outflow of experienced workers is also more likely to occur, as these workers seek jobs with a higher compensation or are recruited to better jobs (Boeri, 1996; Montgomery, 1999; Saks and Wozniak, 2011). Workers’ experience is a crucial part of human capital that is reported in several studies to be correlated with injury incidence (Chi et al., 2005; Fabiano et al., 2010; Leung et al., 2014; Oh and Shin, 2003). Thus the outflow of experienced workers and turnover of staff also contribute to the association between business cycle and workplace injury.

Reliance on temporary workers is an excellent instrument for employers to adjust the capacity of their workforce to cope with fluctuations in product demand (de Graaf-Zijl and Berkhout, 2007; Jahn and Bentzne, 2012). In addition to their unfamiliarity to the environment, temporary workers are more vulnerable to injuries because they lack a sense of belonging. They might also receive less amount of safety training. All of these could result in a greater likelihood and severity of injury than permanent workers had they been instructed to do the same activities

Consequently, business cycles accompanied with variations of labor composition and turnover, including the flow of inexperienced workers, experienced workers and temporary workers, contribute to variations in the rate of workplace injuries. Asfaw et al. (2011) and Fernández-Muñiz et al.(2016)concluded that labor composition and employment both contribute to this association. Considering the causal inference between business cycle and incidence of workplace injury, the first set of hypotheses regarding how injury incidence is associated with business cycle is proposed as follows:

H1-1: Labor composition is associated with the incidence of workplace injuries.

H1-2: Changes in the labor composition occur prior to the changes in injury incidence. 

實證結果顯示-只有批發零售殘廢死亡率(Y2)和新人增加(LC1)有關

整理來說，所謂人員組成與人力資源，這個說法無法有效解釋為何職災率會和景氣循環有關

第二組解釋 工作條件與狀況(Working Conditions)

正常工時長短(WCd1)

加班工時長短(WCd2)

加班對正常工時的比例(WCd3)

How working conditions contribute to the incidence of workplace injury

The term working conditions, broadly defined, is related to workload and production pace. During economic boom periods, time required to complete a task might decrease under company’s policy. This in turn, forces workers to work overtime and/or to speed up the pace of operations. Working in jobs with overtime schedules or long work hours has been found to be associated with higher injury rates (Dembe et al., 2005; Dembe et al., 2007). In contrast, an economic downturn could lead to a reduction in operations, resulting in a decline of working hours or a slow working pace. Increased overtime and longer working hours during an economic upturn could also lead to greater fatigue, thereby undermining employees’ safety awareness and health with an increase in adverse events and errors (Baker, 1985 ; Fay and Medoff, 1985; Sokejima and Kagamimori, 1998; Olds and Clarke, 2010).

In addition, increased demand during a boom time might lead to an accelerated pace of working, thereby increasing the likelihood of both workers’ missing operations and occurrence of injuries (Davies et al., 2009). Asfaw et al. (2011) also discussed other issues that might contribute to the increase of injury incidence during boom times, including inadequate time for training, insufficient resting and breaks as well as an ignorance of safety rules by workers. However, it is difficult to isolate the impacts of the individual factors as they are generally intertwined. We therefore follow Asfaw et al. (2011) and refer them jointly as working conditions (WCd). The second set of hypotheses is proposed as follows:

H2-1: Working condition is associated with the incidence of workplace injuries.

H2-2: Changes in working conditions occur prior to that of injury incidence.

實證結果顯示-加班對正常工時的比例(WCd3)對輕微職災率有顯著的關聯

加班時數相對於正常工時的增加率要大，職災率越高

第三組解釋 勞資關係(Industrial Relations)

失業率(失業率高，勞工怕沒工作，不敢通報職災IR1)

加入公會的比例(加入工會有保障，敢通報職災、爭取權益IR2)

勞資爭議件數(IR3)

How industrial relations contribute to the incidence of workplace injury

The third set of explanatory variable concerns industrial and labor relations, which are crucial to determining the nature and extent of employers’ rights and obligations towards employees. Studies have shown that this factor plays an important role in organizations to influence workers’ safety (Zacharatos et al., 2005). As Leigh (1985) observed, in times of economic depression with high unemployment rate, workers may want to reduce their risk of being laid off.  They may decide not to report minor injuries, believing that probability of dismissals is high (Boone and van Ours, 2006; Boone et al., 2011). This biases the number of reported workplace injuries downward during times of recession. Workers’ reporting behavior and bargaining power can be thus considered as part of the labor and industrial relations. Davies et al. (2009) used firm size and manual employment as a proxy for workers’ bargaining power. They found that reported minor injuries are related to workers’ bargaining power.

Apart from the issue of reporting behavior and consideration of dismissal, a labor union is another important part of industrial organization that exercises bargain right on behalf of workers. Economou and Theodossiou (2015) showed that increasing union density is associated with a decrease in the number of both fatal and non-fatal work injuries. Industrial labor relations and the bargaining power are also affected by supply-demand condition in the labor market. During an economic upturn, job vacancies increase, and unemployment rates shrink. Therefore employment and industrial relations should be favorable to the labor. Employers are forced to pay more attention to employee welfare as well as safety in the workplace, leading to a decrease in the incidence of workplace injury. This interesting countercyclical possibility has been predicted by Boone and van Ours (2006). From the incentive perspective, real earnings and unemployment benefit of workers may influence the behavior of injury compensation claims in the workplace (Fortin et al., 1996; Kniesner and Leeth, 1989; Tikhonova et al., 2012; Wooden, 1989). Workplace injury insurance claims may not exclusively include only the industrial injuries and accidents, but also could include a whole range of social insurance-based disability programs (Benitez-Silva te al., 2010).

In short, external economic conditions can greatly affect workers’ injury reporting and claiming behavior. We hypothesize that the net impact of industrial relations on injury incidence would be significant by the following two hypotheses:

H3-1: Industrial relations are associated with the incidence of workplace injury

 

H3-2: Changes in industrial relations occur prior to the injury incidence. 

實證結果顯示-餐飲住宿業才有顯著關聯-失業率越高，(製造業)失業人口反而專而從事餐飲住宿業，餐飲住宿業就業人口大增=>職災率高，勞資爭議件數也多

搭配縱斷面的時間序列先後關係檢定，提出以下整體的解釋