Now I am now moving onto the case study. It is our Business Statistics Innovation Program (BSIP). It finished in June 2005. It was aimed at achieving multiple outcomes as described in Table 2.

I will soon tell you a little bit about what we actually achieved through this project. What we were trying to do was adapt new technology and new methodologies to achieve desirable business outcomes. But to be successful we also had to move away from an organisational structure that was based largely on silos. ABS collection activities were largely vertically organised around particular statistical collections or groups of collections and that stopped us getting full advantage of new technologies, new methodologies where economies of scales matter as well as reducing the coherence of statistics across the silos. The following diagram illustrates what we did (Figure 1):

We changed from a vertically structured organisation that I just talked about to a more horizontal structure to provide greater functional specialisation and economies of scale. Starting on the left hand side of the diagram, we have the Economic Statistics Data Centre which undertakes all the data collection activities, all the front end processing activities and so forth. There is only a single data centre whereas there are multiple Business Statistic Centres (BSCs), shown in the middle part of the diagram. They are based around groups of similar business collections. For example we have one Business Statistic Centre (BSC) for all our sub-annual business collections, we have another BSC for all our agriculture collections. The third type of organisational unit is the National Statistic Centres (NSCs). There are multiple National Statistic Centres and they are based on particular statistics themes. So we have a labour NSC, for example, and its role is to get away from being solely based on internally focused labour collection based activity that it was responsible for to having a much greater alignment with the users and producers of labour statistics and to take an interest in all data sources that could be used for labour statistics.

BSIP was a significant program, taking a bit over three years to establish. But we went about it in a structured way. Firstly we consulted very broadly with staff, very early when we were first thinking of making changes. And we actually changed plans as a result of those consultations. I personally got involved in some of the consultations because I wanted to demonstrate my own interest in this major change program as well as hearing first hand the reactions of staff. As I mentioned above, we exploited the available technology. It was new technology, but not leading edge.

A key objective of this whole program was to exploit new administrative data sources. The key source we had was taxation data and with the introduction of the GST all of a sudden we had available to us all the information on the business activity statements. So in effect we were getting a business Census every quarter. People may not have liked GST but statisticians did because they had this wonderful new data source that they could now use to produce new statistics or reduce the collection load we place on businesses. And BSIP was also an opportunity to introduce greater consistency across our statistical collections. They each had their own life and developed the way they did things independently. This may have been justified when looked at in isolation but they were different so we didn’t have the degree of coherence that we would have liked across our statistical sources. Re-skilling staff was an important part of the ‘how’ and very early in the change program we identified the training needs. And consistent with the theme of this chapter, good project management was also very important.

I would now like to make a few more detailed comments on the involvement of staff. Before we actually started the program we developed a set of what we call strategic scenarios and they were the basis of the discussions we had with staff and unions. The key thing was not only that we did talk to staff about the scenarios but we weren’t just lecturing, we were also listening. We actually made changes as a result of the feedback and that helped get greater ownership of what we were doing. Once BSIP was broken down into specific projects we also provided scope for staff to have an influence on exactly what was going to happen within each project, particularly on job design aspects. And there was continuous commitment from senior staff for the whole program. The end result was that, in terms of staff reaction, we moved from staff being very sceptical at first (understandably because there was concern about things like job tenure, what sort of work they were going to be doing, whether they were going to be made redundant, would they have the skills to do it) to reluctant commitment. By the end of the three years we really had most staff very committed and enthusiastic about the whole program and looking for ways to actually do new things that we hadn’t actually thought of.

With respect to governance, we set up a BSIP management board. I will mention one thing that I think is quite important: we included an outside expert on the program management board. We thought it was important to have someone who was going to have a different perspective and that actually proved to be very useful. We also as I mentioned before used a chap named John Smyrk who did a review at the very beginning of the proposed arrangements and also mid way through the program just to give us some insights on some things that we might do a little bit better. The BSIP program was broken up into a series of projects and line managers took responsibilities for particular projects. Because of the size of BSIP we set up an implementation coordination team that provided secretariat and monitoring services and a range of other support activities.

In terms of achievements, I guess the one that we are most pleased about was that we more than achieved our productivity goals. We were planning to reduce from 1,029 staff employed on business statistics to 895 in fact we ended up with 856. And other good things happened whilst BSIP costs stayed within budget. The organisational change was completed but business continuity was important. In the early phases of the project we managers asked about measures of success and business continuity was at the top of the list. We did not want to implement the changes in a way that meant that particular statistics were going to be delayed or not produced for a particular period of time, or that there were serious discontinuities in time series. We managed to do this and being able to achieve this goal was very important to the success of the project.

There were a number of lessons learnt. First of all, the need to be clear and focused at the start about the outcomes being sought. A lot of the projects fail because the outcomes or goals that were being sought were not clear. This was a big program, involving a massive change. We had a number of projects that contributed to the program but they all needed to be inter-related. So having all that set out clearly and setting up the governance arrangements to allow it to happen and manage issues is important.

Keeping the amount of change manageable is another lesson. We actually cut back on some of our plans when we thought that it was going to be too much to achieve at one time. Communication is very important. We used our intranet to get messages, particularly from me, out to staff about what was happening and how progress was made. But there is nothing like face-to-face communication and we encourage the senior people who were involved in BSIP to get out and about as much as they could and let people know what was happening. This is a three-year program, and staff do not want to hear ‘nothing’ about it for those three years. People want to actually know what is happening as the project progresses. And just repeating what I said a few moments ago, address the skills development process early in the program. It is very important to involve staff in working out exactly what the skill gaps are and what you might do to address them. It also helps develop ownership.