Dr Penny Burns, Chair, Talking Infrastructure

Economists speak of the ‘natural level of unemployment’, arguing that even in the best run and most vibrant of economies there would still be unemployment. It is called ‘frictional’ unemployment and refers to those people temporarily ‘between’ jobs and it is necessary if the economy is to adapt to changing circumstances and opportunities.

Is backlog maintenance similar? Can we say that in the best run of agencies there would still be backlog for ‘frictional’ reasons? Say, because unexpected breakdowns required parts to be flow in from overseas, or because it was more efficient to deal with a number of related maintenance problems at the same time so that the early ones are held over? Maybe there could be unexpected demands on a system such that it could not be released for overhaul and maintenance?

Backlog maintenance may be the grease that makes the entire maintenance machine run more efficiently, the tolerance we need to get on with the job. The cost of ridding ourselves of this backlog is then the cost of large inventories of specialised parts, the cost of inefficient small scale attention, and the cost of duplicate capacity. So maybe some maintenance backlog is desirable.

However ‘desirable’ is generally not the word that most immediately springs to mind when we think about maintenance backlogs. Instead, we more normally define a maintenance backlog as that maintenance that ‘should’ have been done but wasn’t (generally because of insufficient funding) implying both that it is a problem and the solution is more funding.

It is worth asking whether, with this definition we could ever spend enough on maintenance to rid ourselves of a backlog, or whether the ‘shoulds’ would not automatically increase to fill any potential gap.  

How is backlog maintenance measured?

Do you ask the maintenance managers or facility managers to tell you what their backlog is? If so, the chances are that this is interpreted as "What would you spend more money on if you had it?"

If you think that it is not interpreted this way, I urge you to check. Consider, when justifying their maintenance budget, who does not ensure that there are enough items on it to utilise any potential windfall? How many estimate their likely maintenance requirements to be less than their expected budget allocation?

Without a requirement for all items to be cost justified, there is no real upper limit to what is, in effect, a wish list. It is clearly not worth documenting all the things that could be done if the chance of funding them is immeasurably small. For this reason ‘same as last year’ is always a good bet which may account for some of the constancy observed in practice. If your measurement system is those items scheduled but not carried out, the problem still exists – only now you must ask yourself what justification there was for scheduling it in the first place. 

Testing for credibility

One test of a backlog item is does the scheduled cost of repair or maintenance, or the assessed risk, increase for having been deferred? If not, then it should not be a backlog item – it shouldn’t have been on the schedule either! 

Cost justification reduces backlog

Many backlog lists would fall if the expenditures had to be cost justified, or if those proposing the item had to justify it in terms of its impact on service delivery. One large government department carried extensive backlog lists for years, then it was corporatised. With it new commercial attitudes, the list of backlog items was reconsidered and found to be not needed. What changed? Not the asset, just the management attitude. 

 For more, check out Dr Penny Burns' story, “The renewal gap is so yesterday!” on the Talking Infrastructure blog.