企业风险管理论文指导——企业矛盾体系中的灵活组织策略

The next quality of HROs is related to culture and flexible structures but is probably distinct from them: the concept that decision making in HROs is always devolved as far as possible to the people dealing directly with situations and problems. The simplest manifestation is perhaps the veto that anyone on a flight deck can issue on a landing or take-off, irrespective of rank. It is similar to the practice in modern manufacturing plants that anyone on a production can stop the line, on the basis that anyone could notice a significant problem. 

 

It is perhaps worth spelling out why this is important. First, devolving decision making should give a kind of consistency that’s rooted in what happens on ground. Decisions coming from high in a management structure, precisely because they are out of touch with   http://www.ukassignment.org/fxgllw/  day-to-day reality they sometimes seem more whimsical. Devolution should also avoid the problem that when senior people make decisions they tend to be more public and more difficult to reverse. Devolution helps avoid problems of knowing who’s entitled to make what decision in what circumstance. And it should produce collegial authority where individuals are happy to defer to other individuals with more expertise, rather than emphasising formal authority. 

This time the critique is that claiming someone with a veto enjoys devolved decision making is an exaggeration. A veto is not a decision. Also, as we saw in the lecture on failure models, local decision making can contribute to disaster as well as help avoid it. Snook’s model of practical drift incorporated the idea that people ‘on the ground’ make locally sensible, but globally problematic, adaptations to the rules and procedures that are there to protect the system. The HRO people would perhaps say that devolved decision making has to be combined with heedfulness, but it is not clear that it will always be possible for local decision makers to be fully heedful of the consequences of their actions.

Another quality of HROs is the attention they pay to learning. The difficulty for them is that if they involve large hazards learning by trial and error is prohibitively expensive. Therefore they have to learn from imagining potential risks, from simulating what might happen and what might go wrong, from minor incidents that could have been more serious in different circumstances, and so on. Whenever the organization is not engaged in normal production activities it needs to be engaged in rehearsal or training, and to be learning from it.

有一个态度的元素在HROS中，它宣称的态度是，只要有一个错误，它应该被视为一个机会学习，而不是一个机会，指责。组织，强调，怪只会导致人来掩盖自己的错误，而不是诊断或沟通。因此，对人力资源外包原则的支持者描述的“安全区”的错误和故障都可以谈，没有责怪 - 例如，无记名投票和匿名报告系统。一位业内人士这样做很好，得到了良好的声誉，是航空，它的多样性不归的报告系统。

There is an attitudinal element as well. In HROs, it is claimed, the attitude is that whenever there is an error it should be treated as an opportunity for learning, not an opportunity for blaming. Organizations that emphasise blame simply lead people to cover up their errors, not to diagnose them or communicate about them. So proponents of HRO principles describe ‘safe areas’ where error and failure can be discussed without blame – for example secret ballots and anonymous reporting systems. An industry that’s got a reputation for doing this well is aviation, with its multiplicity of non-attributable reporting systems.

However there are again some problems. First, reports of errors are often too poor to be a good basis for learning. People often don’t have time on the ground to write reports well, they are often not experienced in doing so, and they are often ambiguous about what inferences to make. Second, ‘no blame’ systems are not always good for sorting problems out. After the Ladbroke Grove train disaster in the UK it was said at the inquiry that there had been a no-blame culture in operation, and the consequence had been that signals passed at danger were attributed to ‘driver error’. Because there was no blame, however, drivers simply went along with this, even though it was a simplistic diagnosis. If drivers had been blamed they might have contested this diagnosis, and the organization might have realised that there was a systemic problem with trains passing through red lights. Blame is also functional in society in the sense that it reaffirms what we believe to be wrong. You could not imagine not blaming a lecturer for turning up late, because it’s important that lecturers know that punctuality is important. 

 

Also, the military context of most HRO studies is misleading. The military in peacetime is usually rehearsing, so in a sense it can afford a lot of lessons-learned activity. At war it is unlikely to enjoy this luxury. And in fact the high reliability that has been observed in the military is mostly during peacetime. Studies of the military in wartime might well show much less reliability. Incidents of friendly fire, for example, seem to demonstrate that wartime reliability can be a big problem.

The last quality of HROs that we look at is redundancy. This is not particularly an attribute of HROs, however, as redundancy is a ubiquitous way of attaining reliability. You can take a device that is only 90% reliable in any context and raise reliability to 99% by putting a second device in parallel that can take over if the first fails. 

 

This kind of redundancy can be extended from physical devices to people – to obtain social rather than technical redundancy. One person can check the work of another, for example. To some extent, having ‘many eyes’ on a carrier flight deck can compensate for the lack of time by attaining high levels of reliability.

 

But there is a further kind of redundancy, sometimes called ‘conceptual’ slack, where the redundancy is in ideas or concepts rather than resources. It can be good to have redundancy in theories about how the world works, for example, in order to avoid the blindspots that are inevitable in any single theory. Capable managers often entertain multiple theories about situations, and don’t make up their minds too soon. This is a problem if you have to make decisions quickly of course, so is less applicable in tightly coupled systems where events can develop rapidly. But as a general principle it looks plausible.

Finally, most HRO studies seem to have concentrated on a narrow range of operations. The archetypal HRO is American, military and has planes in it. Not all HRO contexts have all these characteristics, but most have at least some of them.

 

The American context might mean that high reliability is seen because American society can spend a lot of its resources on reliability and safety (an estimate is 15%) and because it is preoccupied with safety. The military context might mean that we see operations that are not ‘producing’, because the researchers only had access to them in peacetime. The military context is also one in which ‘culture’ is prominent and one in which individuals can be controlled to a greater extent than in civilian life. The aviation context also means a typically wealthier industry, with more resources to devote to reliability. And the nuclear context is perhaps similar.