A cat, a hat and a simple measure of gobbledygook: How readable is your writing?

Jonathan Fullwood.

Sometimes the obvious questions are the hardest to answer. In this post I ask how much of what the Bank and the financial industry in general  write can actually be read by a broad audience. Based on my findings, I suggest that both must try harder if claims of accessibility are to be meaningful.

Nine months, 236 words

Among children’s books, The Cat in the Hat is an enduring favourite. Since 1957 at least 10 million copies of the tale have been printed. With children never seeming to tire of the easy rhymes, flamboyant central character and unlikely plot, the next generation will almost certainly read the book to their children, too.

The book’s text, by Theodor Geisel (Dr Seuss to you and me), at first appears to be straightforward. But perhaps this is a result of the great thought that went into it. Geisel claimed to have worked on the book for nine months. His aim was clear. To write an exciting book that would encourage six and seven year old first grade children to read.

Challenged by a publisher’s education division to limit the vocabulary to no more than 225 words, Geisel eventually picked 223 words from a list of 348 first grade target words, plus 13 simple words that were not on the list. The flowing, rhyming prose and Geisel’s trademark illustrations were an immediate hit with his target audience.

Why is this important?

We live in the information age. Both public and private institutions now make a wealth of information freely available via their websites. But does this really make these bodies accessible?

Are documents that are made available written for their target audience? Is time taken to ensure that the vocabulary is suitable? Can those wishing to read the prose actually do so?

If we are able to answer these questions, then we will be closer to answering a question close to our own hearts: Is the financial industry doing all it can to be accessible when it writes public documents?

How can we measure readability?

Let us be clear. The notion of readability is not easily summarised in a blog post. William DuBay’s 2004 paper on the principles of readability, for example, runs to 72 pages. In the paper he notes that by the 1980s there were 200 formulae and over 1000 papers relating to the subject!

Despite the breadth of the area, DuBay concluded that simple metrics can help in preparing readable documents. By starting with a set of texts for which required reading levels have been tested, such metrics can be built by regression using explanatory factors such as sentence length, paragraph length and syllables per word.

Table 1 summarises some simple measures of readability, including McLaughlin’s ‘Simple Measure of Gobbledygook’.

Table 1: Measures of linguistic complexity

Measure	Based on
Coleman-Liau complexity	Sentence length, word length
ARI complexity	Sentence length, word length
SMOG grade index	Polysyllable* words per sentence
Gunning fog index	Number of polysyllables, sentence length
Flesch−Kincaid grade level	Sentence length, average syllables per word
Coleman correct completion	Sentence length, number of one syllable words

* Words with three or more syllables

What is a good target level?

Being able to estimate the level of education needed to read a given text is useful. But which grade level should one target? It may be that one has a very specific audience in mind. In this case it is easy to judge whether the writing is of an appropriate level.

Often however, the target audience will be poorly defined or very diverse. This is a problem since it is not obvious at what grade level the average member of the public reads.

It turns out that there have been some studies in this area. Figure 1 shows the reading grade level of US adults as identified in a 1992 survey. If writing for a general audience it seems prudent to assume a reading level of grade eight or nine. This corresponds to a reading age of 13-15 years.

Figure 1: Literacy grade level of adults (% of population)

Source: US national adult literacy survey 1992

Some examples

Let us assume that a grade eight or nine level is a good starting point for a text aimed at a wide audience. We can then see whether different types of writing fall at or below this target.

Figure 2 shows Flesch−Kincaid (F-K) reading grade levels for some well-known forms of writing. Recall that this measure uses sentence length and the number of syllables per word to assess readability.

Figure 2: Flesch−Kincaid reading grade level of excerpts from different sources

From the chart we can see that tabloid newspapers are pitched at a grade-level consistent with our assumption. This appears consistent with the fact that the tabloids top the circulation tables. Political speeches also appear to use simple language and short sentences.

What about the financial industry?

We looked at a selection of reports and some terms and conditions from four UK banks. We also considered recent speeches and a selection of public reports from the Bank of England.

The chart shows that the world of finance still has some work to make its writing truly accessible. Private sector bank output tends to require a reading level of grade 12 and Bank of England writing requires a reading level of grade 14. The result is similar for the other metrics listed in Table 1.

Why? Those writing in the financial industry tend to use long words. They put those long words in long sentences. And those long sentences in long paragraphs. The longest sentence we found in our sample of Bank of England documents contained 77 words. This is more than three times the maximum sentence length suggested by the Government Digital Service.

The good news is that, in my experience, improvements in readability could be made with relatively little effort.

Such effort would surely be welcomed. As recently as May 2016 the Bank’s Chief Economist spoke of a great divide between banks and the public. Mr Haldane pointed out that effective communication must play a role in rebuilding trust in the financial sector.  He suggested for example that banks might publish an annual report for customers, cleansed of the complexity in the main report aimed at professional investors.

The Plain English Campaign, which cites ‘financial gobbledygook’ as a prime example of jargon, would also be happy to see clearer and more concise writing.

Dumbing down?

Financial writers may argue that it is not fair to compare their work with the content of a tabloid newspaper. They will perhaps say that the content of their reports is technical and not to be ‘dumbed down’.

I would advise them to read Richard Feynman’s lectures in physics. The Nobel Prize winner’s lecture that introduces quantum mechanics and Heisenberg’s uncertainty principle has a reading grade level of less than eight.

But what if writers argue that in order to be interesting, more complex language may be necessary? Ultimately output should be enjoyable to read as well as accurate.

Figure 3, like Figure 2 shows F-K reading grade levels, but this time for famous authors. At least one book by each of the writers appears in the Guardian’s list of the best 100 novels written in English. Only Hardy’s writing produces an F-K grade level above ten. Most of the authors’ writing, based on this simple metric, is accessible to young teenagers.

Figure 3: Flesch−Kincaid reading grade level of excerpts from novels by renowned authors

Source: Project Gutenberg and personal ebooks

Conclusion

Readability is important. Evidence suggests that many writers don’t give sufficient consideration to this fact. In particular, there appears to be plenty of scope for those in the financial sector to improve the accessibility of their work.

Of course, readability metrics are only part of the picture and should not be followed slavishly. Clarity, style, and layout clearly also play a role in determining readability. Nonetheless, why not activate the simple readability measures in your word processor?

Finally, in case you were wondering this blog post has an F-K reading grade level of 8.7. Thankfully, it did not take nine months to write!

Jonathan Fullwood works in the Bank’s Advanced Analytics Division.

If you want to get in touch, please email us at bankunderground@bankofengland.co.uk. You are also welcome to leave a comment below. Comments are moderated and will not appear until they have been approved.

Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.