Wednesday, May 19, 2010

Autonomy and administrative load - the need for simplification

Yesterday's post, Case studies in public administration, took a very long time to write and I have continued musing over the issues.

When I first joined what I now think of as the old public service, my cohort used to complain that the class 11 (section head) kept the most interesting work and passed the rest to the class 9 who did likewise and redistributed the rest. There really was no management as such in the sense of consciously organising resources to get the best results.

When I did my Masters in Economics at ANU, I was part of the first intake in a new course specifically designed to meet public service requirements. I found the course very interesting because, among other reasons, I hadn't done any formal public administration. However, while the course included a range of conceptual material, there was again no management material as such.

All this got me interested in management as a discipline and I started reading books on the subject, including especially Peter Drucker's work. The end result was that when I did get into a position to "manage" and especially after I became a section head, I had formed views and lots of things that I wanted to try.

I now found an odd thing. The very things that I had complained about in the old system actually gave me very considerable freedom to do new things. So long as the work got processed, I could pretty much do what I liked. This included organising the work in such a way as to free up resources for longer term thinking. All this was, quite simply, fun.

Between then and now a couple of things have happened that have diminished real management autonomy. I just want to point to a few. In doing so, I have to be a little cautious because I really don't know the detail applying in all jurisdictions.

The first is the rise of what we can call administrative load. Managers today have to do a lot more administrative things than I had to do when I was a manager. They also have to do them in increasingly systematised and rigid ways.

Consider time sheets and flex time. Just as I became a manager, the old system of time books was replaced by flex time and time sheets. Initially this new system was quite flexible and took very little time to administer. I knew what my people were doing anyway, so simply organised things to be as flexible as possible. This included sometimes bending rules to take individual needs into account.

Today's computer based time systems are far more rigid and complicated. Management by exception is not possible, while there are many more rules - core time, maximum hours, banked time etc. Staff have to learn the rules and enter their time on line, managers have to check and approve. Checks exist to ensure that all this is done.

Say that you have five staff working for you. Checking and approving time sheets now takes between 25 minutes and 50 minutes per week, taking into account the inevitable problems that can arise. This may not sound like a lot, but it is dead time.

This is replicated across systems. It is now the manager's responsibility not just to approve leave, but to handle the processing. Whereas I used to just sign the form and send it off to HR, modern managers have to handle all the HR work themselves. In doing so, they also face more complicated leave rules than used to be the case.

Or consider recruitment. In developing a new area, I grew my staff from 4 to 37 over two years. That's quite a big recruitment load, bigger still if you consider that there was turnover as well during the period. I don't know that I could do this today, or at least not without taking a lot of time away from my primary role.

The recruitment dance has become quite stylised.

Position descriptions have become more complex. It used to take me perhaps two hours to revise or create a one page position description.

In a recent case where I was asked to do this, it took me a day. The multi-page position description then had to go for formal evaluation. So we have increased input and elapsed time.

In advertising positions, I used to prepare a short notice for the Government Gazette and send it to the HR people. If we were advertising outside it was a little more complex because I had to prepare the ad and then send it off, again to the HR people.

Today with computer based systems, and accepting that rules and structures vary between jurisdictions, there is a far more complex process. Without going into details, it took me over half a day input time just to arrange for the position that I am talking about to be advertised in the way we wanted.

The application process has become far more complex and stylised. Short applications have been replaced by far longer applications that must address selection criteria in detail. It is not unusual for managers now to essentially give staff a day of working time to prepare their application.

Job applications serve two purposes.

In the first instance, they are used to exclude candidates. If, for example, you want to interview 7 but have 70 applications, you have to get get rid of 63. More complicated applications don't help here. Worse still, the capacity to meet the formal application requirements becomes an exclusion factor in its own right. This works against candidates from outside.

In the second instance, the application guides questioning during interview. Again, you don't need long and complex applications.

Interviews themselves have become more stylised because of the rules. When I was first interviewing, the focus was on trying to find the best person. Questions of procedural fairness were there, but they were secondary. Of course interviewers are still concerned with getting the best person, but they have now to meet far more rules.

Once the interview has been completed, the consequent checks are far more complicated. I didn't have to worry about police reports, for example.

I am not saying that these things are necessarily wrong in their own right. I am saying that every extra hour a manager has to spend on administration is an hour that not available for that manager's primary role.

I spoke in my post, as I have done before, about the rise of activity based controls. Here I want to introduce a new point.

Cascading performance agreements with their activity focus create a fundamental conflict from the manager's perspective.

The reality is, as it was when I first became a manager, that a fair proportion of work is reactive, A problem comes up, you have to sort it. This holds at all management levels, but cascades down. My boss's problem becomes mine.

This creates an irreconcilable conflict with the standard performance agreement, for that deals with the defined. In theory, you can handle this by building in an allowance for the reactive. In practice, the total time associated with the defined actually exceeds real time available.

All this creates cynicism, especially when combined with project management approaches.

Activity x appears in my performance agreement. I may know that this is not a real activity or project to begin with, just something that appeals up the line. Alternatively, I may think that it is a real project, but I know that it holds up only so long as supported up the line. Then again, it may be real and important, but hard to deliver because reactive demands constantly conflict.

Let me try to illustrate all this with two examples.

The first involved a research task that had been around for some time. I was asked to do something on it. Everybody said that it was important, but nothing had happened. When I looked at it, I thought that there was something there, but previous approaches had been symbolic rather than real. So it kept slipping.

I did a position paper looking at options. This was not welcome: I had gone the wrong way; we just needed to get something done to get it off the list. I had no idea what to do next, so just let it slide. Two years later I believe that it is still on the area's to-do list.

The second is a major project, not one that I have mentioned before. I became project manager. As project manager, I dropped into to do mode. For a number of reasons, nothing happened. All the effort was wasted.

In this case, this is not a criticism of the hierarchy, simply an observation. Once it was defined, given formal status and announced by the minister, once it appeared in my performance agreement, then I had to act. I went to project meetings - this was part of a bigger complex of projects - and reported. I had to field questions on delay. Yet I was missing a fundamental point.

Because the approach was new, because many inter-related things had to be worked through, there was a fundamental conflict between my position and that of the Division Head.

She was constantly changing her position to work out the best overall approach, taking into account a broader range of factors. By contrast, I was meant to get the currently defined thing through.

Had I been more sensitive, had communication been better, I could have managed this. I might have been able to say something like, well if that's your position, this is the way we might handle this. Instead, I tried to deliver on what I had been tasked to do.

I may seem to have come a long way from my starting point. However, I guess in finishing I would make this point. If we are to improve the effectiveness of public administration in this country, we need to do at least two things:

  • We need to increase the autonomy and flexibility of managers
  • We need to reduce the administrative load on managers.

No comments: