Next

Previous

Houston

Home

Feedback

banner for Houston PLAIN conference

Experiences in Writing, Testing, and Rewriting Regulations in Plain Language

Presentation by Phil Knight and Jacquie Harrison
Facilitator Joe Kimble
Reported by Debra Bulmer

This session was actually more about statutes, than regulations, but Joe Kimble's prediction clearly applied to both: governments and others are short sighted in not springing for the cost of testing. Joe's own presentation on Saturday told us more about why he believes that testing documents is the only way to know that they are in fact doing what we think they are doing AND what we want them to do.

Phil Knight and Jacquie Harrison have conducted the only documented reader-focused studies to evaluate the quality of legislation. Phil tested two versions of South Africa's Human Rights Commission Act, 1995. Jacquie tested three versions of the new provisions of New Zealand's tax legislation dealing with exempt income. Their reports were published in 1996 and 1999 respectively and will guide other governments who are forward thinking enough (and provide adequate funds) to test their own legislation to see if the intended audience can understand it.

Phil's presentation reinforced how important it is to identify what we hope to learn from testing. In order to answer this question, we also have to know what our documents are supposed to be doing. Grabbing our attention with a sample of writing involving a story about ambiguous sex and a mailbox (or maybe it was the story that was ambiguous), Phil sent us back to the drawing board with the message that some testing is always better than no testing.

Phil also demonstrated an unofficial electronic version of South Africa's competition legislation. His informal testing of that version on an informed audience of lawyers and accountants was designed to evaluate general concepts and approaches. He soon learned where his design was confusing people during their attempts to locate information in the legislation. Phil's demonstration helped us all see what might lie ahead for electronic access to legislation-based government programs, perhaps even officially, some time in the not too distant future. We can now visualize what it would be like to link from the legislation to the regulations and even to the form that we could fill in on-line and then email to the appropriate government agency.

Jacquie told us how Joe Kimble had accessed her report through the language-l listserve, highlighting once again how valuable the list is as a networking device. She then described how Parliamentary Counsel had been skeptical of the potential to simplify the New Zealand tax laws. A Canadian, Elizabeth McAra, was parachuted in from the Alberta Justice Ministry and proceeded to try to show the local drafters that plain language and complex tax concepts could indeed coexist in the same legislative document.

With a meager budget of $5000, Jacquie and Elizabeth conducted a reader-based test of the original version, one simplified rewrite and another simplified rewrite that used questions as headings (Robert Gentle: that would be headlines, to you). While more professional uncertainty was displayed by the group using version two, Jacquie reported higher comprehension rates overall for the two simplified versions. The data was unfortunately inconclusive as to the effectiveness of using headings that were questions, as opposed to statements. The good news is that the degree of cooperation between the Inland Revenue Department and the Parliamentary Counsel Office has improved as a result of this effort made by the Department to introduce plain language concepts into New Zealand's statutes.

Later an informal presentation by Guy Grenon from Human Resources and Development Canada described his government's plan to test part of a plain language rewrite of the Employment Insurance Act. Thanks to Jacquie's and Phil's work, Canada can now design its test knowing more about what questions to ask and what methodologies to use, in order to learn whether the reader has a better understanding of the plain language version or the original version. We all know which version we suspect will be the more effective, but won't it be nice to be able to run some numbers - if only to put a smile on Joe's face?