1.6 What story does this picture tell?
As to the propriety and justness of representing sums of money, and time, by parts of space, tho’ very readily agreed to by most men, yet a few seem to apprehend that there may possibly be some deception in it, of which they are not aware …
(William Playfair (1786) The Commercial and Political Atlas, London)
The political economist William Playfair, who developed many of the graphical representations familiar today, was well aware of the visual impact of graphical presentations, and of the impressions they can create. In his book The Commercial and Political Atlas, published in London in.1786, Playfair published the critical graphic shown in Figure 28. Writing on Playfair’s contribution to graphic design, Edward Tufte commented: ‘Accompanied by [Play fair’s] polemic against the “ruinous folly” of the British government policy of financing its colonial wars through debt, [this graphic] is surely the first skyrocketing government debt chart, beginning, the now 200-year history of such displays’ (Tufte, E. (1983) The Visual Display of Quantitative Information, Graphics Press, Connecticut, p. 65). The way Playfair has drawn the graph, using a tall and narrow shape and by not adjusting the money figures for inflation, emphasises the rapid growth of the British national debt during the eighteenth century.
But Playfair also produced an alternative version a few pages later. Shown in Figure 29, the graph shows the interest on the national debt plotted against time.
Playfair has here taken inflation into account and plotted the cost of the debt in ‘real terms’. To lessen the impact further, he has chosen a different format for the graph: a broad horizontal scale for the time and a relatively short vertical scale for the debt. Now the situation does not look quite so bad, although if you look carefully, the graphs show that the debt and the interest on it increased by about five times from 1739 to 1784: a period during which Britain was involved in wars with Spain, France, and America.
But authors of graphs do not always make their embedded conventions explicit, preferring to rely on the immediate visual impact of the graphic to encourage readers to skip over the details and jump to conclusions. Here is an example. A graph used by a political party on one of a series of publicity postcards, entitled ‘A Better Health Service’, is shown in Figure 30. It shows how the number of nurses and midwives in the UK rose from 440,000 to 500,000 between 1978 and 1987, while the number of doctors and dentists rose from 81,000 to 95,000.
Activity 12: Skyrocketing growth?
What visual impression does the graph in Figure 30 give about the rise in numbers of nurses and midwives, and doctors and dentists over the period in question? What message do you think the graph is being used to convey?
Now look at Figure 30 more carefully and try to read out the actual state of affairs. What methods have been used to create an impression of ‘skyrocketing’ numbers of medical staff? How might you redraw the graph to modify the visual effect?
If you are to be influenced by the shape of the graphs, then clearly the numbers of doctors, dentists, nurses and mid wives rocketed over the period. Not only that, but both graphs are a lot steeper in 1987 than they were in 1978. In other words, the rate at which the numbers were increasing was itself going up. There will not just be a steady growth of doctors and nurses, it seems to suggest, there will be increasingly more and more as time goes on. A better health service indeed-the graphs seem to speak for themselves. And this phenomenal growth, people are clearly encouraged to think, has been achieved as a result of the support of the political party who produced the publicity postcards.
If you look carefully, however, you will see that a number of graphical devices have been used to create this overall impression. The first has been to use a relatively short horizontal axis, and to have a tall, narrow shape to emphasize the growth. Notice also, that unequal periods of time have been given equal space. Further, the graphs have been drawn using axes where the vertical scales do not start from zero. And, although it seems you are encouraged to compare the numbers of doctors and nurses because both graphs have been drawn in the same space-and perhaps even to see the numbers converging as the two graphs come closer and closer-in fact, the vertical scales of the two graphs are quite different.
Figure 31 shows the data redrawn. You can see that the effect is rather less dramatic.