The Fédération Internationale des Ingénieurs-Conseils (FIDIC) first published its Client/Consultant Model Services Agreement (the White Book) in 1990. This standard-form contract between the client and the consulting engineer was drafted after editing and revising the document it was intended to replace, namely FIDIC's International General Rules of Agreement (IGRA). The IGRA had been published in three versions for different kinds of consultancy services (IGRA 1979 for Design and Supervision of Construction of Works, IGRA 1979 for pre-investment studies and IGRA 1980 for project management). In contrast, the White Book is suitable for all the kinds of consultancy services encompassed by the three former agreements (see the foreword to the 2006 edition of the White Book). Regarding the extent to which the White Book is used in practice, the Danish International Development Agency (2004, p. 6) states, although without providing evidence, that the White Book, along with the World Bank document “Standard Request for Proposal”, is the main internationally recognised document for contracting of consulting services.
After its initial publication in 1990, revised editions of the White Book were published in 1991, 1998 and 2006. This paper reports on a study in which the four editions were examined by quantifying the extent of changes. The aim of what was, in effect, a counting exercise was to track the changes made to the contract, and then to identify some of the underlying causes of those changes. The construction sector has seen many important developments since the White Book was first published, and it is interesting to see whether the changes identified in the White Book can be related to these developments. Also, analysing these changes against the backdrop of important developments that have taken place in the construction industry will lead to some conclusions about the responsiveness of the drafting committee to these developments.
1. 2 Scope of research: the general conditions of the White Book
The subject matter of the present work is only the general conditions, i.e. the standard clauses of the White Book. These were called “Standard Conditions” in the editions of 1990 and 1991. The other parts of the agreement (the particular conditions, the four appendices and any letter of offer and acceptance, see Clause 1.1. of the 2006 edition), have not been considered in the quantification.
In this paper, we use the structure (and especially the clause numbering) of the fourth edition 2006 as the main point of reference. This structure differs from that of previous editions. In the 2006 edition, the general conditions are divided into eight sections: general provisions; the client; the consultant; commencement, completion, variation and termination; payment; liabilities; insurance; and disputes and arbitration. For the purpose of this research, the two sections dealing with the client and the consultant as well as the two sections covering insurance and liability are treated as one. Hence, we have only six sections (Section 3.1).
1. 3 The FIDIC drafting committees
The four editions of FIDIC's White Book were prepared by two different drafting committees. The editions of 1990, 1991 and 1998 were prepared by FIDIC's client/consultant relationships committee ( FIDIC, 2001 . Acknowledgements Section; see also FIDIC, 1998, p. 5 ). In contrast, the 2006 edition was drafted under the responsibility of FIDIC's (2006, p. 5; and the Acknowledgements Section of the 2006 edition of the White Book) Contracts Committee.
The primary method used in this piece of work is a specific type of quantitative content analysis. In this method, researchers establish a set of categories and then count the number of instances that fall into each category ( Silverman, 2001, p. 123 ). Since the data can be accurately measured it is a quantitative research method. This method is particularly useful for the analysis of documents, which are regularly updated such as standard-form contracts. One example of research work using this method has been reported by Hilburn and Hughes (2005) , who looked into the changes in the code of conduct of the royal institute of British architects over different editions. Minor and major changes could be identified. It appeared that some of these changes were responses to regulatory pressures instigated by the British Government to deregulate the profession of architects. For example, it appeared that some clauses of the code were omitted from one edition to the next in order to comply with statutory requirements. Elsewhere, Bunni quantified text to ascertain the length of sentences in the FIDIC's Conditions of Contract for Electrical and Mechanical Works 1987 ( Bunni, 1987, pp. 178, 194, 195 ) and the number of sub-clauses in FIDIC's Red Book 1999 ( Bunni, 2005, pp. 503, 504 ). He used the data to draw conclusions about the clarity of the standard-form contracts.
In the present research it was anticipated that the counting of words of the General Conditions of the White Book, over the four editions, would reveal trends towards or away from certain topics. With an understanding of these trends an attempt can be made to look for the motivation behind the changes introduced.
Regarding the counting itself, the following rules have been applied in order to be systematic about the process:
Everything outside the General Conditions has been excluded.
Marginal words and other headings have not been counted, as according to Clause 1.2.1 of the 2006 edition of the White Book these ought not to be considered while interpreting the clauses. Only the text of the clauses has been considered for the counting of words.
Hyphenated words, for example “[…] vice-versa […]” in Clause 1.2.2 of the 2006 edition, have been counted as one word.
Words with a slash, for example “[…] and/or […]” in Clause 4.4.1(b) of the 2006 edition, have been counted as two words.
Terms such as “[…] Appendix 1 […]” (Clause 1.1.1 of the 2006 edition) are counted as two words.
Some of the clauses have text containing alternate names for various appendices in square brackets, for example “[…] Appendix 1 (Scope of Services) […]” in Clause 1.1.3 of the 2006 edition. The words in such square brackets have not been counted.
Figures such as “[…] 56 […]” (Clause 4.6.1 of the 2006 edition) are counted as one word.As mentioned above, this quantification provides a basis for examining why particular clauses of the White Book were amended. For this purpose the clauses have been examined, in the context of prevailing developments within the construction sector that may have led to the changes.
A full word count of the different editions shows that the total number of words has grown significantly, the 2006 edition having about 34 per cent more words than the 1990 edition (Figure 1). There was no significant increase in the number of words in 1991 but the 1998 and 2006 editions of the White Book had considerably more words than the previous ones.
Figure 2 shows the number of words in the different sections of the General Conditions. When compared to the first edition of 1990, the 1991 edition had almost the same number of words under most headings, there being no difference at all or only a trivial difference of one or two words. The only significant change was in the commencement, completion, variation and termination section where there were 14 more words in 1991. This was due to the addition of a second sentence in Clause 29 which says that if the contract is terminated the provisions regarding the “limit of compensation and indemnity” (Clause 18) remain in force. The comparison of 1991 with 1998 reveals that significant changes were made only to the general provisions and the disputes sections. All other sections were retained as before. In 2006, the number of words has gone up considerably in all sections. The details of these changes will be described and discussed in the following sections of this paper.
3. 2 Changes in individual clauses
Figure 3 tracks the changes in the individual clauses. The types of changes that have been made to each clause in successive editions can be seen.
The 1990, 1991 and 1998 editions each have 44 clauses, in the same format, so comparing them is straightforward. In Figure 3, these three editions are displayed in the first...
