Investigating presentational change in U.K. annual reports: a longitudinal perspective.

To enable comparison to Beattie and Jones (1992b), a similar checklist to that adopted by Beattie and Jones was used in Part 2. Specifically, data were collected on the following: graph usage (key financial variable graphs and non-key financial variable graphs), aspects of graph design, and details concerning the incidence and degree of measurement distortion and the specific causes of any distortion. For the non-key financial graphs, all the graph titles were recorded and subsequently grouped into appropriate categories (e.g., social responsibility, net assets, product information, etc.).

A graph discrepancy index (GDI) was used to calculate measurement distortion. Following Beattie and Jones (1992b), this was defined as

GDI = (a/b - 1) x 100%

in which a = the percentage change (in centimeters to one decimal place) depicted in the graph (i.e., height of last column less height of first column/height of first column) and b = the percentage change in the data over the same period. A value of zero indicates no distortion, a positive value indicates the exaggeration of the trend (i.e., data distorted in the company's favor if the trend is favorable), and a negative value indicates an understatement of the trend (i.e., data distorted to the company's disadvantage if the trend is unfavorable). (7)

RESULTS

The presentation of results is in three parts. First, Tables 2 and 3 compare Lee's (1994) sample, Beattie and Jones's (1992b) 1989 sample, and the 2004 sample. In assessing the trends over time in these two tables, differences in the composition of the samples must be borne in mind. The 1988 sample in Lee's study is based on only 25 of the largest U.K. companies, whereas the 1989 full sample in Beattie and Jones (one year later) is based on 240 companies drawn from the top 500 (approximately one third) of all listed U.K. companies. The 2004 full sample is based on 100 companies from the top 500 U.K. listed companies. In addition, to facilitate closer comparison with Lee, who focused on large companies, a restricted "large company sample" of 25 companies was drawn from both the Beattie and Jones' (1992a, 1992b) sample and the 2004 sample. Second, Table 4 provides a detailed analysis of the content of the 1989 and 2004 samples. Finally, Tables 5 to 12 focus on graph use, comparing the results from the 2004 sample to those from Beattie and Jones's 1989 sample. As the sample sizes vary over time in all tables, the percentage figures offer the key basis for comparison.

Comparison of Annual Report Structure and Form: 1965 to 2004

In Table 2, a comparison among Lee (1994), Beattie and Jones (1992b), and the 2004 sample is provided. This comparison enables an evaluation to be made of changes in the structure and form of annual report content over a period spanning 30 years. Panel A shows results for the full Beattie and Jones and 2004 samples, whereas Panel B shows the large company subsets. This subset, although less statistically representative of the population, more closely equates to that used by Lee. We first compare panel A and panel C (full samples). Several clear trends are apparent. There is an increase in the size of the annual report from 26 pages in 1965 to 75 pages in 2004. This threefold increase reflects a rise in both regulatory and voluntary material: although regulatory information rose at a rate of 186% (from 15 pages to 43 pages), voluntary information rose at a marginally greater rate of 190% (from 11 pages to 32 pages). The increase was not, however, uniform over time. Although the regulatory page count witnessed a dip between 1978 and 1989 and then rose sharply between 1989 and 2004 at a rate of 115%, the voluntary page count rose consistently with the highest rise at a rate of 71% between the 1978 and 1989 period. This finding was despite the countervailing trend of successive regulatory capture, whereby voluntary items, such as corporate governance and statement of directors' remuneration reports, have become mandated over time. The proportion of voluntary page count to regulatory page count was stable at 42% at the beginning and at the end of the period studied, although during the 1980s more than half the page count volume was voluntary in nature. Our data thus contradict D&S's (2007) finding that voluntary material was only 17% in 2002. This difference is probably accounted for by their rather restrictive definition of voluntary information. (8) Nevertheless, the rising trend of voluntary material representing an overall majority in the reports between 1965 (16%) and 1988 (72%) was reversed from then on, so that by 2004 only 17% of the companies disclosed more voluntary information than regulatory information. This result, together with that of the proportion of voluntary page count to regulatory page count, confirms the higher variation (standard deviation) in reporting practice in relation to voluntary material.

Two important presentational formats for the non financial information were narrative and pictorial. The narrative information increased from 8 pages in 1965 to 38 pages in 2004 (a rise of 375%). In particular, the period from 1978 to 1988 saw the narrative pages rise threefold from 6 pages to 19 pages. However, the percentage allocated to pictorial material (mainly pictures) has not risen so fast over the entire time period (the percentage fluctuates over the 30 years, doubling from 1965 to 2004). The percentage of companies presenting their financial statements at the back of the annual report has risen markedly over time and constitutes 100% of the companies studied by 2004.

When Panels B and C are compared (the large company subsample), we find the same trends as with the full sample comparison. However, these trends are magnified. The total page count increased by 265% rather than 188%. The regulatory pages and voluntary pages increased by 253% and 281%, respectively. Narrative and pictorial information also rose markedly. However, the rise in narrative information is the most impressive. It increased by 625%, whereas for the full sample it was 375%. These larger companies, being in the public spotlight, are likely to be in the vanguard of financial reporting and are under special pressures to be accountable. Where they lead, others are likely to follow.

In Section 3 of Table 2, information is provided on high-level design features of the annual report. There was a major increase in the use of prominent corporate logos on the front cover of the annual report from 28% in 1965 to 78% in 2004. The table also shows an increase in the use of external design consultants from 12% in 1965 to 72% in 2004. Although practices of the large companies (Panel B) were broadly not dissimilar to those of the total sample (Panel A), in 1989 a smaller proportion of the large firms relied on logos or external consultants. Overall, these design features reinforce the general tendency for presentation to become more important.

Comparison of Annual Report Content: 1989 and 2004

Table 3 investigates in detail the content of annual reports in 1989 and 2004. Nineteen generic sections other than the financial statements and related notes (the financial accounts) are identified. Two of these sections were mandatory in both sample years: auditors' report and directors' report. Two other sections were mandated by 2004: statement of directors" responsibilities and corporate governance. In addition, the remuneration report, though introduced since 1989, was mandated only in 2002. The sections are shown in the order in which they commonly appear in annual reports. For each section, tour key attributes are reported: (a) whether a particular section was present or absent, (b) whether the section was presented before or after the financial accounts, (c) the number (and percentage) of companies that included pictures, graphs, charts, and tables in each section, and (d) the mean number of pages devoted to each section.

The popularity of several voluntary sections has remained relatively static. Table of contents, financial highlights, chairman's statement, advisors, board of directors, and historical record have all remained popular sections. However, the inclusion of several voluntary sections has changed substantially over time. The inclusion of a combined operating and financial review has grown from 2% to 22%. This reflects the introduction of a recommended operating and financial review statement in 1993. However, it is interesting to note that, of the two component parts of the operating and financial review, the operating review has become less popular than the financial review. Although the inclusion of a separate operating review has fallen (from 51% to 36%), the inclusion of a separate financial review has risen (from 9% to 67%). Companies, in general, preferred to report separate operating and financial reviews rather than combine them. In 2004, it was expected that the operating and financial review would become mandatory (Department of Trade and Industry, 2002), and the overall rise reported was in anticipation of legislation. (9)