I"VE BEEN asked how complex systems differs from other methods of understanding our world and for a change I thought I’d have a go at describing some of the methods used. I can’t explain such a field in a short article, but perhaps I can illustrate some differences and offer some insights.

Systems practitioners are in short supply (like tradesmen) and most work in areas where mission critical outcomes depend on good analysis and design, such as space craft design and the design of weapons delivery systems. The ideas apply to all complex systems however, usually some familiarity with each of the sub systems is required, thus calling more for ‘generalists’ than ‘specialists’. Because of the specialist nature of many sub-systems, systems people often end up working with a lot of different specialists to try to build a better understanding of whole systems.

Systems practitioners need to consider everything relevant, whether it fits within a scientific framework or not. Science is only one way to understand our world – there are plenty of others that produce relevant information or information that people act upon. This means human behaviour and foibles, disinformation, arbitrariness, madness and corrupt practices could be included – if it’s relevant it’s in.

Our own senses are another example of a knowledge development system – one that we’re all confined to and that colours all of our other knowledge and its acquisition, scientific or otherwise.

As a result of these, and other factors, tracking down possible causes of problems in complex systems can be like a broad ranging detective story, each study unique because every system is unique, with many presenting challenges like missing information and deception as part of the mix. Often we have little solid data to go on, just past practice, patterns noted elsewhere and a clutch of rigorous principles to keep us on reasonably solid ground.

A little background

Definition: A system is any group of interacting elements that produce results that no smaller group of the elements can produce. Examples include the human body, a working operating theatre, a swarm of bees, a government and a country. A pin is not a system, but the Hydro is.

Systems combine to form larger systems. Our Hydro system contributes electricity to a distribution system (Transend) which contributes to our industrial and community systems.

Remembering all of the relationships in complex systems can be almost impossible so quite a few graphical display methods are used to show relationships to aid our understanding. When the results of each relationship can be quantified (which is not often in my experience) then computer models can be built to predict future outcomes.

Systems become complex when the interactions between their parts make predicting the outcome of whole system activity very difficult or impossible. This creates serious problems for our governments whose artificial portfolio boundaries often mean that system parts are managed without understanding, knowing or caring about, the effects on the whole system.

For example, an improperly thought through shift from ‘free’ education to ‘user pays’ can have deleterious impacts on housing by failing to produce enough skilled trades people to build or maintain homes and their various infrastructures. Other areas similarly impacted could include health and engineering. Therefore in attempting to ‘improve’ education, we could damage a whole bunch of other desirable capabilities. Such damage is a fairly typical result of linear thinking… that is thinking that fails to take into account the various interactions and implications of system change.

When these kinds of problems occur, the ‘inspired’ leaders who stimulated their creation are usually not only resistant to learning, they are often actively hostile towards any suggestion that their initiatives could have been responsible for the problems.

Students of systems learn pretty early that managing a part of a system usually does not improve system performance. In the case of the human body, just taking up aerobic exercise to increase fitness levels might do little other than bring on a heart attack (or similar) if other bodily activities (like nutrition and circulatory system health) are ignored.

Science is just one example of a system for developing knowledge. Because reliable data and accurate measurements are frequently not available to us, we rely quite a bit on logic; the main underpinning of science and maths, to help us understand what’s happening in systems.

Also we usually have to analyse a system ‘on the fly’, that is while it’s still operating, frequently within aggressive time constraints, and come to useful decisions without the benefit of repeated studies and double blinds etc.

We can use our knowledge of systems for a range of purposes including design, correcting system performance, and facilitating systemic collapse (counter-systems).

Example: Cutting federal budgets to deal with homelessness

Kevin Rudd seems pretty determined to keep his tax cut promises and cut government budgets by around $18 billion as part of his attack on inflation. A problem that often emerges with ‘razor gang’ type activities, is that they can destroy important initiatives so it is usually worth exploring what approaches to cost cutting might usefully be employed.

By analysing the systems involved, we might note that there are government programs that appear to be making our problems worse and, since government programs cost money, it seems sensible to consider what would happen if we reduced them.

As an example, I’ll use homelessness - a recent issue that Kevin Rudd has identified as worth addressing.

Early in our analysis, we’d try to identify some root causes of homelessness, this might include high house prices, poverty, insufficient housing, insufficient skilled tradesmen and rural development restrictions. We might assume that the provision of shelter to Australians is something of a priority, without which our society would soon collapse. Examples of systems intended to provide shelter include an economic system that allows people to purchase/rent/find adequate shelter, enough local activity to assure sufficient jobs for all who need income, sufficient land available for homes, training that assures enough competent tradespeople are available, adequate charities and crisis accommodation etc.

Since homelessness is increasing, we might conclude that some/all of the systems to provide shelter are failing to meet their purpose (i.e. provide adequate shelter) in various ways.

Reports show Australia as having low housing affordability due to high prices, high costs and relatively low incomes. Further analysis indicates that Australia’s tax system has been delivering huge incentives for investing in real estate for decades (e.g. negative gearing) and that the subsequent activity in that market is probably pushing up real estate prices.

We might also note that high home values are delivering windfall benefits to owners while leaving those with fewer resources well behind economically… that the gap between rich and poor is growing, in part due to real estate tax breaks.

Because increases in government taxes and charges have pushed up prices in general, they have also reduced household’s discretionary budgets (that part of their income over which they have a free choice) thereby making homes less affordable. In addition, rural development is being hampered by development restrictions (e.g. PAL Act) and a general run down in rural services (e.g. insufficient doctors, poor telecommunications services) that are both conspiring to make rural life unattractive and push people into less affordable and already pressured urban areas.

By drawing all of these items into a connected map (a simple relationship map is provided as an example)…

image

… we can more easily understand the implications of the various interactions that exist, and some of the likely root causes of the matter under consideration, in this case homelessness.

When we review what’s happening using systems ideas, we often end up with very awkward questions to answer, questions that frequently scare the linear thinkers (thinking restricted to one path, as opposed to multiple paths that exist in complex systems). Because linear thinkers are caught up in their restrictions, they almost always respond in the same way to challenges and ‘threats’ to their point of view.

Example: Hospitals are for…?

In reporting that the CEO of the LGH (Launceston General Hospital) was leaving for better climes and a better job, the Examiner reported “The State Government committed…to reduce bureaucratic red tape that was stifling decision making and return responsibility for day-to-day operations to the major public hospital.”

This statement leads to questions like…’If bureaucratic red tape is, in fact, stifling decision making, why doesn’t the government stop using it? Who is in charge here?’

Even more indicative, later in the same article it stated ‘A copy of minutes…stated that “a decrease in (LGH) performance would be accepted as long as we decrease expenditure’.
This indicates that the purpose of providing adequate health care for all, is not the purpose of the administration. Furthermore it seems that the administration is authorised to reduce health services as long as budgets are reduced.

Is it appropriate for bureaucrats to have such powers?

Coupled with information about the other budget priorities of the government we can soon reach additional conclusions. If budgets for other favoured initiatives were easy to find (e.g. Elwick, Hawthorn) then we could reach some candidate conclusions about the overall priorities of the state government. (Remember, these are just examples)

Such candidate conclusions are based on evidence presented, not on rhetoric or spin. The evidence is produced by state government behaviour, whereas their rhetoric may have no connection whatsoever to their behaviour.

Thus targets for budget cuts might include real estate investment tax incentives (worth quite a lot of money) and other government activities that are diverting money and attention away from taxpayer valued priorities. For example with smartcard technology, we could eliminate whole layers of valueless BAS reporting by collecting the information electronically thereby saving huge business costs and probably lower prices.

To give a sense of scale, it was reported that business compliance with Australian government requirements, cost $86 billion in 2006.

One way of assessing government activities is to examine what percentage of budget goes to activities that do not support the core purpose of the organisation, like administration and PR. Such activities are not mission critical to service delivery and consequently we can afford to reduce them if they get out of proportion.

By understanding how much of our money is used to deliver service, and how much is used for other things, we can better prioritise our budgets to deliver the services that we do need.

Experience tells us that it is usually worthwhile because the pictures that we (literally) get, reveal far more than the restricted views of a few people, and frequently suggest powerful interventions that are too often ignored.


Mike Bolan
http://www.abetteraustralia.com
Mike is a complex systems consultant, change facilitator and executive and management coach.

1) http://www.theage.com.au/news/opinion/jason-koutsoukis/2008/02/09/1202234230366.html?page=fullpage#contentSwap1
2) Link to example Homelessness map here
3) http://www.abc.net.au/lateline/content/2003/s999018.htm

Mike Bolan

Systems become complex when the interactions between their parts make predicting the outcome of whole system activity very difficult or impossible. This creates serious problems for our governments whose artificial portfolio boundaries often mean that system parts are managed without understanding, knowing or caring about, the effects on the whole system.

One way of assessing government activities is to examine what percentage of budget goes to activities that do not support the core purpose of the organisation, like administration and PR. Such activities are not mission critical to service delivery and consequently we can afford to reduce them if they get out of proportion.

By understanding how much of our money is used to deliver service, and how much is used for other things, we can better prioritise our budgets to deliver the services that we do need.