‘The future will be like the present, only different’ said American baseballer Yogi Berra. Was he making a prediction, a forecast or a projection? These three words (along with many others) all refer to estimates about what the future holds. So how do you as a writer choose between them? What are the differences – and do they matter?
The good news is most readers will be as unsure as you – and probably won’t care which you use. But there are contexts – science, economics, meteorology to name a few – where people may attach more importance to your choice of words. And the bad news is that even in those contexts there’s no real agreement about what the different words mean.
Let’s start with the easiest one – projection. It differs from the others in that it’s as much about the present as the future. A projection is a value based on extrapolation from known variables such as quantity and behaviour. So if we know the population of a country is one million, and we know the current population growth rate is 10% per year, we can project that the population in five years would be 1,610,000. But this is not the same thing as predicting this is what the population will be. Projections just illustrate the outcome if current circumstances and trends continue in line with assumptions.
Forecasting is a bit trickier, and there are quite a few definitions being used. These include:
- A set of possible futures that include probability estimates of occurring.
- A prediction that refers to levels or magnitudes.
- A projection where the assumptions represent expectations of actual future events.
- And, statistically, an estimate about values beyond a particular dataset.
There’s not a lot of common ground here, but it seems that in general forecast has nuances of stating that something which cannot yet be observed will occur or will have a certain value, with a fairly high level of probability.
What about prediction? Once again there are a few not very helpful definitions:
- An estimate of a specific event in the future which will either occur or not occur.
- Statistically, an estimate about not-yet-identified values within a particular dataset.
- A more general term for forecast.
The key difference in nuance seems to be that predictions are more specific and more definite than forecasts. To return to the earlier example, if we say that the population is highly likely to be in the range 1,500,00 to 1,700,00 in five years, we are forecasting; if we say that it will be 1,600,000 we are predicting. And it’s possible to have both a forecast and a prediction – for example, that population growth will slow down (the prediction) by somewhere between 1% and 5% (the forecast), in which case we could be right in our prediction but wrong in our forecast.
Do you need to worry about this? Probably not much. There seems to be a definition to suit any usage, and as for the general reader you can prognosticate, foretell, envisage and prophesy that they will not care which word you choose.
