Forecasting techniques

Using sophisticated computer techniques and relying primarily on numerical data, some companies and many governments and international organisations attempt to model changes in the environment. These models often utilise economic data and attempt to esti-mate future economic variables, such as interest rates and the external value of currencies.

There are many private consultancy companies that specialise in developing such models for government and commercial clients. However, as environmental stability is very much a phenomenon of the past, modelling of this type has been subject to considerable ‘bad press’.

Such models find environmental flux and discontinuity difficult, if not impossible, to pre-dict. For example, models of this nature could not possibly have predicted the terrorist attack on New York in 2001 (9/11) and the catastrophic effect this had on US and European airlines. Nor did they predict the 2008 credit crunch.

Figure 1.9 Step-by-step approach to forecasting Selection of environmental variables

Selection of sources of information

Evaluation of forecasting techniques

Integration of forecasting outcomes

Monitoring and evaluation of the critical aspects

Far less expensive to develop, and often just as accurate, are time series and judgemental models. Time series models attempt to identify trends in variables based on historical data or cyclical factors and extrapolate them into the future. For example, a simple time series model may look at the population of a country at five-year intervals over the past hundred years then use this evidence to predict future demographic changes. This method does not, however, allow for environmental discontinuity where the ‘rules’ of the past no longer apply.

A slightly more sophisticated model may add additional variables, such as likely changes in birth rates and predictions concerning the migration of people, which may have a bearing on the population of the country in question. The resulting demographic forecasts (see Chapter 5) may prove useful for strategic planners in government and some organisations.

Minicase 1.4 also highlights the potential problems of demographic change.

Judgemental models are those based upon the informed opinion of people in the relevant field. For example, sales-force personnel may be asked to estimate likely future trends in sales potential, taking into consideration all likely variables. Their experience ‘on the ground’ may prove invaluable and lead to more accurate forecasts than sophisticated model-ling techniques could achieve.

Table 1.5 illustrates a United Nations (UN) prediction of world population growth to the year 2150. This prediction combines knowledge of historic trends, including more recent signs of a reduction in the rate of world population growth, with educated guesswork. It shows population growth only increasing slightly after about 2050. Whereas the UN esti-mate that an extra one billion (1,000 million) people will inhabit the world between 1999 and 2013 (just 14 years), they believe that it will take 129 years for an extra billion people to live between the years 2054 to 2183. Will it be proven correct?

How can anyone predict population growth in the next millennium using ‘scientific’

means? Well, demographers (those who study population) are luckier than many, as trends in population are partially predictable over reasonably long time periods. For example, if world population growth slows shortly (as they predict), this will mean that there will be fewer young adults in a generation’s time, and therefore fewer mothers and fathers to have children.

The UN believes that because world population growth is slowing now, it will continue to do so indefinitely. What they cannot predict is an event or events that cause current trends

Minicase 1.4 China’s ageing population

China has been emerging from under-development and widespread poverty to become a major economic superpower with rapid economic growth. It has for a generation enforced quite strictly a one-child policy in order to control population growth in what is the world’s most populous country. The impact of this is to create an aging population: the 4-2-1 problem where each child often has two parents and four grandpar-ents. Not a problem in many respects but what will this mean in time to come? China remains a relatively young country – median age is about 30 – but rapidly aging. The dependency ratio, that is the number of old and young as a proportion of the number of working-age adults, is currently low – there aren’t so many children any more. However, it is now increasing and will by 2050 be as high as in 1975. The big difference is that whereas most dependants in 1975 were children (about 85+ percent) with only about 15 percent of dependants being old adults, in 2050 about 65–70 percent of dependents will be old-aged adults. The trend has begun. In 2015, almost without doubt now, almost 50 percent of all dependants will be older people – by 2020 it will be over 50 percent (unless the one child family policy is rescinded).

Questions

1 How do the needs of older people differ from those of the young?

2 What is the likely impact of these changes on families, organisations and government?

to radically alter. For example, there may be dramatic new technological inventions that vastly increase longevity and/or fertility rates or, conversely, there may be new viruses that threaten the lives of billions. However, as suggested, some important demographic factors can be predicted with reasonable accuracy and the predictions are vital to certain business sectors. For example, we know that in most of the Western world, with the possible excep-tion of the USA, the populaexcep-tion is ageing. We can make what may be reasonably accurate estimates of the number, even the proportion, of the total population which will be of pen-sion age in the year 2050 – for the simple reason that these people already exist. So, whereas less than 10 percent of the population of the UK in 2010 is of pension age (over 65 years old), it is predicted that that figure will rise to 13.5 percent by 2050. The accuracy of this prediction is highly important for all of us and especially for government and the pension business. It will also have highly significant implications for the healthcare industry; see Chapter 5 for further discussion on patterns of population change.

Another common, rather creative, method of generating ideas and forecasts is widely known as ‘brainstorming’. A number of informed people are encouraged to generate ideas and forecasts in a group setting. It can usefully be employed to estimate future trends in technology development, for example. Many of these ideas may appear fanciful but techno-logical developments often do lead to ‘fanciful’ outcomes! Such techniques can generate useful judgemental ideas about potential future events.

The Delphi method of forecasting is a more systematic technique than brainstorming.

This method attempts to gain consensus among a group of people, such as a senior strategic management group. For example, a company senior management team may meet and aim to forecast their likely competitive position in five years’ time. They will discuss all relevant vari-ables and start to agree on as many points and issues as possible in an attempt to develop the most likely and most widely held view. This can then be used in the strategy process.

It is interesting to note some of the predictions made in the past from reputable sources which turned out to be somewhat misguided. The Quarterly Review (1825), an English jour-nal, asked ‘what could be more palpably absurd than the prospect of locomotives travelling twice as fast as stagecoaches?’ In a similar vein, Henry Ford’s lawyer advised in 1922 that

‘the horse is here to stay, but the automobile is only a novelty – a fad’, while an editorial of the Boston Post in 1865 wrote, ‘well informed people know that it is impossible to transmit the voice over wires and were that it were possible to do so, it would be of no practical value’.

Table 1.5

Year Population

1000 310 million

1250 400 million

1500 500 million

1750 790 million

1900 1,650 million

1950 2,520 million

2000 6,060 million

2050 8,910 million (est.)

2100 9,461 million (est.)

2150 9,750 million (est.)

Source: adapted version of ‘Table 1. World Population, Year O to near Stabilization’ as shown on www.un.org/esa/population/publications/sixbillion/sixbilpart1.pdf from ‘Human Development Report 1999’ by United Nations (1999). By permission of Oxford University Press, Inc.

The growth of world population

Finally, the Chairman of IBM in 1943 wrote, ‘I think there is a market for about five com-puters’. Refer to section 9.7‘Future trends?’for further discussion of ‘Futurology’.

Scenario development recognises judgemental and non-quantitative information such as changing fashions. Scenarios are ‘pictures’ or ‘stories’ of what might be the case some time in the future. They draw upon both subjective and more objective data. Hence, a company may develop two or three likely scenarios for some future date and take these into consideration in their planning process. They may develop contingency plans to cope with each scenario should it arise. The multinational oil giant, Shell, has made extensive use of scenario ‘plan-ning’. An example of this approach, a scenario that hydrogen fuel cells may rapidly replace traditional fossil fuelled engines for road transport, is explored in Minicase 1.5.

Finally, a number of organisations and consultancies have developed ‘political risk’ rat-ings for countries around the world. These take into consideration the stability and predictability of nations and their governments, terrorist activity, and advise commercial organisations and governments on the risks involved in overseas investment.