Sample Selection and Data Sources

The initial sample consists of all FTSE 350 companies listed in the London Stock Exchange (LSE), for the three-year period between 2008 and 2010. Conducting this study in the above specifications is fascinating for the following reasons. First, as already mentioned in the introductory chapter, using a sample period between 2008 and 2010 the author attempts to address regulatory concerns about firms having low financial reporting quality and ineffective external audit processes. During the 2008 to 2010 period, the UK financial reporting guardian, the FRRP, had raised concerns regarding the quality of financial reporting due to the belief that firms are more inclined to manipulate revenues after the recession. In parallel, the UK House of Lords criticises the role of Britain’s Big 4 auditors during the global financial crisis and recommended restrictions on the auditors’ provision of non-audit services to the FTSE 350 firms and an enhanced role for audit committees to monitor the auditor-management relationship aspects, one of which is auditor remuneration. Second, in contrast to the US rules-based accounting system where the

175

majority of similar studies took place, the UK has a principles-based accounting system which focuses on the substance of the principle rather than its form. Third, in contrast to the US mandatory corporate governance system in which the vast majority of the relevant studies were conducted, the UK follows the ‘comply or explain’ approach whereby audit committees are voluntary. Fourth, compared to the US, in the UK “there is greater variation in outside director representation on boards” (Peasnell et al. 2005, p.1313). Finally, FTSE 350 firms are the largest UK listed firms by market capitalization. The previously discussed calls to restrict auditors from providing non-audit services to FTSE 350 firms and to enhance the role of the audit committees of these firms make FTSE 350 the ideal sample to examine44. Moreover, using FTSE 350 index ensures higher availability of data and similar level of governance recommendations. The largest 350 firms have a higher level of recommendations to comply with, since, for instance, they are required to establish audit committees with at least three independent non-executive directors instead of the two required for smaller firms.

Corporate governance data are collected manually from annual reports. Financial and accounting data are obtained from DataStream. Firms in the insurance, financial and utilities industries are excluded for two reasons. First, with respect to the first empirical model (financial reporting quality), firms in the mentioned industries have revenues and accruals which are different from other firms (Stubben 2010). Second, regarding the second empirical model (auditor remuneration), the excluded industries differ from other industries in terms of their regulatory environment (Zaman et al. 2011) and characteristics. For instance, in the UK, firms in the utilities industry are governed by their own relative service regulatory bodies:

44 _{It is worth noting that all of the FTSE 350 firms examined in the auditor remuneration model are}

audited by one of the Big Four auditors who were criticized by the House of Lords on their role during the financial crisis.

176

Office for the Regulation of Electricity and Gas, Office of Water Services, Office of Communication etc. Moreover, financial and utility institutions are characterised by relatively large assets, but they entail less audit effort and testing than firms with extensive receivables and inventory (Hay et al. 2006b).

The two empirical models in this study have different sample sizes. For the sample of the first empirical model tackling the impact of audit committee and board characteristics on financial reporting quality, this study further excludes industries consisting of less than six firms to minimize the possibility of biased estimates in calculating discretionary revenues and accruals. Moreover, in order to account for outliers, variables are winsorized at the bottom and top 2 percent. The final samples of the first and second empirical models consist of 662 and 619 observations respectively. Panel A and Panel B in Table 5.1 present the sample selection procedures of the financial reporting quality models and the auditor remuneration models respectively. The sample size of the second empirical study (619) is less than that of the first empirical study (662), despite the exclusion of less than six firm industries in the latter, because of the large number of missing audit fees and non- audit service fees data on DataStream.

Table 5.2, Panel A and Panel B, present the distribution of sample firms by industry and year for the first and second empirical models respectively. The majority of firm observations in both of the empirical models are from the industrials and consumer services industries. Firms in these industries represent about 58% of the total sample of each of the two empirical studies. In Table 5.2, Panel A, the number of sample firms in the telecommunications industry in 2008 is zero due to the exclusion of industries having less than six firms. Graphical representations of sample firms’ distribution by industry and year for each of the financial reporting

177

quality sample and the auditor remuneration sample are presented in Figure 5.1 and Figure 5.2 respectively. Industries in both of the samples are roughly equally distributed over the three years except for the telecommunications industry which is excluded in the year 2008 of the first sample (Figure 5.1).

Finally, with respect to the pre-financial crisis sample examined in the additional analysis, the author follows the same sampling procedures mentioned above for both financial reporting and auditor remuneration models. Given that the mandatory adoption of the International Financial Reporting Standards (IFRS) by the UK took place on January 1, 2005, the pre-crisis sample period is chosen to be for the three years from 2005 to 2007 to avoid the effect of the change in standards on accounting quality.

178

Table 5.1 Sample Selection Procedures

2008 2009 2010 Total

Sample

Total firms in FTSE 350 at year end 358 355 356 1069

Companies in financial & insurance

industries (ICB 8000) -113 -112 -116 -341 Companies in utilities industry (ICB 7000) -10 -9 -9 -28

Companies with missing corporate

governance and financial values -6 -5 -3 -14 Industries having less than 6 firms -13 -7 -4 -24

Total sample 216 222 224 662

Panel B: Auditor remuneration models

2008 2009 2010 Total

Sample

Total firms in FTSE 350 at year end 358 355 356 1069

Companies in financial & insurance

industries (ICB 8000) -113 -112 -116 -341 Companies in utilities industry (ICB 7000) -10 -9 -9 -28

Companies with missing corporate

governance and financial values _{-29 -21 -31 -81}

179

Table 5.2 Distribution of Sample Firms by Industry and Year

ICB code Industry 2008 2009 2010 Total

0001 Oil and Gas 18 19 20 57

1000 Basic materials 20 18 26 64 2000 Industrials 69 65 59 195 3000 Consumer goods 23 26 25 74 4000 Health care 8 9 8 25 5000 Consumer services 65 63 64 192 6000 Telecommunications 0 6 6 10 9000 Technology 13 16 16 45 Total 216 222 224 662

Panel B: Auditor remuneration models

ICB code Industry 2008 2009 2010 Total

0001 Oil and Gas 15 18 15 48

1000 Basic materials 20 20 21 61 2000 Industrials 67 66 60 193 3000 Consumer goods 24 27 24 75 4000 Health care 8 9 8 25 5000 Consumer services 60 55 53 168 6000 Telecommunications 3 4 4 11 9000 Technology 9 14 15 38 Total 206 213 200 619

180