Editor’s note: Plain language advocates and consultants, like many other communication consultants, make claims for the effectiveness of what they do without necessarily offering any evidence. This was the first of the papers published in Communication News between 1993 and 1996 that looked critically at the claims and the evidence.
Originally published in Communication News 6(5) September/October 1993.

The plain English movement abounds in apocryphal tales about savings made due to plain English. And because people want to believe that plain English works, there is a tendency not to examine closely the claims the movement makes. Under such circumstances, a number of unsubstantiated stories have circulated—stories that have become ever more authoritative as they appear in ever more authoritative publications. For example, claims have been made that the British and Victorian governments saved millions of dollars as a result of plain English. We have been checking the veracity of these claims and have found either that the empirical evidence showed that any savings were attributable to factors other than plain English, or that there was no supporting empirical evidence at all.

In contrast, we have published a number of studies and drawn attention to the work of others which casts doubt on the claim that using plain English increases efficiency, effectiveness and equity (Penman, 1993)1Penman, R. (1993) Unspeakable acts and other deeds: A critique of plain legal language. Information Design Journal 7 (2) 121-131.. Imagine, therefore, our surprise at seeing our own research cited as evidence for the efficacy of plain English, and as the basis for the creation of another apocryphal tale.

An apocryphal tale in the making

David Kelly and Chris Balmford, who offer a plain English service at solicitors Phillips Fox, have written a report for the Life Insurance Federation of Australia (LIFA) in which they state that rewriting in a clearer and more communicative way results in greater effectiveness and efficiency in insurance documents (Kelly & Balmford 1993).2Kelly, D. & Balmford, C. (1993) Simply Better: The case for clearer Life office documents. Life Insurance Federation of Australia.  In support of their statement they cited our work on information design and management in the insurance industry, undertaken for Capita Financial Group (Fisher & Sless, 1989).3Fisher, P. & Sless, D. (1990) Improving information management in the insurance industry: A case study of Capita Financial Group. CRIA Occasional Paper 10. Communication Research Institute of Australia, Canberra. Later published as Sless, David, and P Fisher. “Improving Information Management in the Insurance Industry.Information Design Journal 6, no. 2 (1990): 103–29.

Chris Balmford elaborated their view in an article in Panorama, the Ansett in-flight magazine:

There is a lot of money to be saved by using plain English documents. The Capita Financial Group in Australia now reportedly saves more than $530,000 each year by using forms that can be completed with less chance of error thus saving thousands of hours of staff time in follow-up work.

Balmford has gone on to repeat this statement in a number of media interviews. And so a new story is born.

The events leading up to the Capita success are of some interest here. As Phil Fisher and I acknowledged in the Capita paper, the methods that proved so successful in Capita had been pioneered in our ground-breaking study in the Australian Taxation Office which led to the successful introduction of the new Tax Form S, firstly in WA in 1985, and then throughout Australia the following year.

The Taxation project

The ATO, in order to improve its tax forms, set in motion a Taxation project, managed by ATO staff member Phil Fisher. The project began with a great deal of confidence in the plain English solution—until two plain English versions of the form were tested with ordinary taxpayers. The results were damning:

On the criteria of delivering better quality information…neither form is any more than a marginal advance over the old form. Our qualitative assessment…is that there are only marginally fewer errors and omissions in the information supplied by taxpayers on the test forms than was the case when the old form was tested (Lenehan et al 1985).4Lenehan, Lynton and Bloom (1985) Market testing of two draft designs for a new tax form ‘S’: A qualitative/ quantitative report prepared for The Australian Taxation Office, Canberra.

The plain English experts had given it their best shot and failed. At the time, the results were a shock. Since then a great deal of evidence has accumulated consistent with this finding, suggesting that applying plain English principles will result in only marginal improvements. Faced with an obvious failure, the Taxation project turned to our Institute. Our approach is based on communication research and detailed observation of the dialogue that goes on between forms and people. For example, normal form-filling behaviour follows what David Frohlich later described as ‘the principle of least effort’ (Frohlich 1986):5Frohlich, D. M. (1986) On the organisation of form-filling behaviour. Information Design Journal, 5 (1) 43–59.

I will only read as much as is absolutely necessary in order to complete this form.

This is why the plain English version had failed. It did not matter how clear the language or layout was, the form-fillers simply didn’t read most of it. If the Institute was to improve the form, we had to alter form-filling behaviour—restructure the dialogue between form and form filler. We did this by introducing what we later called the directed or algorithmic form. Pioneered in the UK (Waller, 1984), this type of form changed form-filling behaviour from one of topic scanning—where the form-filler decides which questions to answer on the basis of a minimal reading of captions or headings—to one in which the form-filler can only fill in the form by making a series of decisions that can only be made by reading the questions. This, and not plain English, was the single most important reason for the improvement in form-filling behaviour among taxpayers. It was also an important factor which led to the massive improvements in form-filling by Capita agents, though this was not the main reason for the project’s success, as I shall explain shortly.

Behind the Capita success

There was very little that could be described as plain English on the Capita forms. These forms were specialised technical documents used in the writing of insurance contracts. For example on one of the forms, agents were asked, under the heading of ‘Investment linked roll-over’, to select from the following options: Managed, Equity, Resource, International, Property, Fixed, or Cash, and then to give a percentage of each—hardly plain English. Even in the parts which were meant to be read by clients, such as the duty of disclosure, the wording was straight out of the regulations and known to be unintelligible to most readers of insurance contracts (Penman, 1990).6Penman, R. (1990) Summary Report to: NRMA Insurance Ltd, The Insurance Contracts Act 1984 & the Duty of Disclosure: A dilemma arising from consumer testing. Canberra, Communication Research Institute of Australia.

In short, the Capita forms were not in plain English, and any claim that the savings were attributable to plain English is false. The main reason for the improvement with the Capita form was the change in form-filling behaviour. On its own this change would not have resulted in success, because many other factors affect the eventual performance of a document in a large complex administrative environment. The overwhelming reason for Capita’s success was the methodology that was used to control the development of the new forms. This involved a carefully thought-out development and testing program, which took into account all the internal stakeholders with an interest in controlling the document. This process was managed by Phil Fisher, who was now employed by Capita to manage their forms unit following his successful work in Taxation. Thus Capita’s success was due to communication research, design methods, testing, project planning and successful negotiation. Plain English was nowhere in sight. Anyone trying to match Capita’s success using plain English is in for a disappointment.

Misrepresentation

Misrepresenting research results is a serious matter, which is generally regarded by researchers as unprincipled and beneath contempt. In the research community there are a variety of ways of dealing with misrepresentation. In this case, where the enthusiastic but illegitimate celebration of good quantitative results is being used to promote a commercial service which distorts the methodology which led to those results, the old non-plain Latin adage caveat emptor applies. Perhaps tellingly, the promotion of plain English services uses advertising techniques reminiscent of miracle medicine cures: testimonials by ‘satisfied’ clients, and juxtaposition of before-and-after the cure examples. Perhaps, also, there is something profoundly ironic about promoting a plain English service in this way—the English is plain but the story is much more complex.