Audit and earnings management in Spanish SMEs

Huguet, David; López Gandía, Juan Article Audit and earnings management in Spanish SMEs BRQ Business Research Quarterly Provided in Cooperation with: Asociación Científica de Economía y Dirección de Empresas (ACEDE), Madrid Suggested Citation: Huguet, David; López Gandía, Juan (2016) : Audit and earnings management in Spanish SMEs, BRQ Business Research Quarterly, ISSN 2340-9436, Elsevier España, Barcelona, Vol. 19, Iss. 3, pp. 171-187, https://doi.org/10.1016/j.brq.2015.12.001 This Version is available at: https://hdl.handle.net/10419/206345 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ BRQ Business Research Quarterly (2016) 19, 171---187 www.elsevier.es/brq BRQ Business Research Quarterly ARTICLE Audit and earnings management in Spanish SMEs夽,夽夽 David Huguet∗, Juan L. Gandía Department of Accounting, Faculty of Economics, Edifici Departmental Oriental, Avda. dels Tarongers, s/n, 46071 Valencia, Spain Received 11 December 2014; accepted 1 December 2015 Available online 18 February 2016 JEL CLASSIFICATION M40; M42 KEYWORDS Auditing; Earnings management; Earnings quality; SMEs; Spain Abstract Evidence about the relation between earnings management and voluntary audits is scarce, and there is no research about the effectiveness of mandatory audits to improve earnings quality. Using a sample of Spanish SMEs, where some companies are mandatorily audited and some are exempt from audit, we examine if audits, either mandatory or voluntary, help to improve accounting quality by constraining earnings management. We also examine differences between voluntary and mandatory audits, as well as the role of Big 4 and Middle-Tier auditors. After controlling for other characteristics that affect earnings management, we find that audited companies have lower absolute discretionary accruals, but do not find significant differences among auditors. Voluntary audits also restrain earnings management, but in a lesser extent than mandatory audits. When we use signed accruals, audits are only effective against income-increasing behaviours, what is explained by the auditor conservatism. Additional analyses support the results obtained. © 2016 ACEDE. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 夽The authors gratefully acknowledge the financial contribution of the Spanish Ministry of Economy and Competitiveness (research project ECO2013-48208-P). 夽夽 We are grateful to the participants at 36th EAA Congress (2013), to the participants at IX Accounting Research Symposium (2013), to the participants at the 9th Workshop on European Financial Reporting and to the comments of Alexander Brüggen, Nadine Funcke, Ann Vanstraelen, Caren Schelleman, Annelies Renders, Patrick Vorst and Jonas Hesse. ∗Corresponding author. E-mail addresses: [email protected] (D. Huguet), [email protected] (J.L. Gandía). Introduction Literature about earnings management and accounting quality is extensive (García Osma et al., 2005; Dechow et al., 2010). A stream of research that has been deeply studied is the relationship between auditing and earnings management, because it is expected that audits work as a constraint to managerial discretion in reporting earnings and help to improve the reliability and the quality of the financial information. The papers that have studied this relationship have focused on the differential value among auditors to deter earnings management activities, depending on specific dimensions of auditors. However, there is a lack of empirical research in two issues: (i) whether audits, http://dx.doi.org/10.1016/j.brq.2015.12.001 2340-9436/© 2016 ACEDE. Published by Elsevier España, S.L.U. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 172 D. Huguet, J.L. Gandía regardless of the characteristics of the auditor, actually constrain earnings management and improve earnings quality; and (ii) whether there are differences between voluntary and mandatory audits. First, papers commonly examine the differences between auditors with different characteristics (Becker et al., 1998; Balsam et al., 2003; Chung et al., 2005; Carey and Simnett, 2006; Cano, 2007; Basioudis et al., 2008). However, they do not study if audits, regardless the auditors’ characteristics, have a differential value compared to the non-audit case. The fact that most of the literature examines settings where audits are mandatory, such as large private or listed companies, involves that these papers cannot value the differences between unaudited and audited companies. Secondly, the few papers that study audits per se generally show that audited companies have a lower cost of debt than the unaudited ones (Kim et al., 2011; Minnis, 2011), so audited financial statements are perceived to be more reliable and thus seem to provide higher quality information. However, with the exceptions of Minnis (2011), Ojala et al. (2011) and Dedman and Kausar (2012), there is a lack of empirical evidence examining whether the audited companies actually provide higher quality information. Minnis (2011) and Dedman and Kausar (2012) examine the effects of voluntary audits on accounting quality, but there is no research about the effects of mandatory audits. This is important because accounting quality can be affected in a different way depending on whether audits are voluntary or mandatory. On the one hand, voluntarily audited companies may be willing to send a signal about the quality of the accounting information, and Minnis (2011) and Dedman and Kausar (2012) show that audits improve accounting quality. However, we can expect a ‘‘label’’ effect for the voluntary audit, i.e. companies only choose to be audited to increase their perceived accounting quality (Daske et al., 2013; Koren et al., 2014). On the other hand, mandatory audits are assumed to ensure a minimum quality of the financial information (Ruiz and Gómez, 2008), thus companies that shun the audit requirement would have lower accounting quality. If these differences are not observed, mandatory audits would fail to achieve their basic aim. SMEs are a natural setting to test the effect of audits on accounting quality. First at all, it is worth noting their relevance in the economy in both the EU and the US (Allee and Yohn, 2009; Wymenga et al., 2012). In Spain, this importance is even higher (EC, 2012). Secondly, the SMEs setting allows us compare audited and unaudited firms, a comparison that is not possible among public and big private companies because all of them are mandatorily audited. Moreover, although there are papers that analyze the relation between audit quality and earnings management in the private setting (Cano, 2007; Van Tendeloo and Vanstraelen, 2008), the value of audits for the smaller of them, however, is not as obvious, because their stakeholders may rely more in alternative information sources (Berger and Udell, 2006; Gill de Albornoz and Illueca, 2007), so the role of auditors may be partially different. The Spanish case may shed light to this limited value, because of the lower tradition in the use of accounting information compared to common-law countries with a longer history of auditing, such as the UK and the USA. Moreover, similar to most EU countries, Spain requires audits for companies that exceed a certain size. The Spanish Statutory Audit Thresholds (SAT) are lower than those generally applied in the EU, so we can test if audits have a different effect depending on their character (voluntary or mandatory) in a relatively homogeneous sample, i.e. in a sample that only includes small and medium companies. Furthermore, because of the more limited usefulness of financial information for SMEs, mandatory audits are often considered a potential source of administrative burdens, so the EC is considering the possibility of revising the requirement for mandatory audits for these companies (EC, 2010). Finally, the audit market for SMEs also gives us the opportunity to test the role of Middle-Tier auditors (Boone et al., 2010; Sundgren and Svanström, 2013). Therefore, using a sample of Spanish SMEs, we examine if audits are a deterrent to earnings management, measured through the signed and absolute values of discretionary accruals, and test whether this effect is driven by a real commitment with accounting quality among voluntarily audited companies, or with a minimum accounting quality ensured by mandatory audits. We also test differences between voluntary and mandatory audits. Moreover, we examine if audit quality, proxied by a three-level classification (Big 4, Middle-Tier and small auditors), means differences on the level of earnings management. Since papers about audit choice have serious endogeneity problems, we use a fixed-effects approach instead of OLS estimations to partially mitigate them (Kim et al., 2011; Lennox et al., 2012) We find that audited companies have a lower level of absolute discretionary accruals than the non-audited ones; voluntary audits also restrain earnings management, but in a lesser extent than mandatory audits. These results suggest that although both mandatory and voluntary audits improve earnings quality by restricting the magnitude of accruals, the lower visibility and litigation risks faced by auditors in the voluntary setting encourage them to be less restrictive. When we examine separately the signed discretionary accruals, we do not find a significant effect of audits on negative accruals, what may be due to the auditor conservatism, for which auditors are not effective against earnings management behaviours when companies have incentives to manage downward. On the other hand, we do not find that significant differences for companies audited by Big 4 and Middle-Tier auditors. Additional analyses support the results obtained. The paper contributes to the literature about auditing and the quality of financial information in the following ways: first at all, it extends literature about the audit/nonaudit discussion. Although previous papers have studied if auditors, playing an information role, help to improve the credibility of the financial statements, there is a lack of empirical evidence examining whether the audited financial information is actually of higher quality. As far as we know, only Minnis (2011), Ojala et al. (2011) and Dedman and Kausar (2012) have examined the effect of voluntary audits on accounting quality. We complement these studies by examining this association in a code-law country, and by considering also both the effect of mandatory audits and Audit and earnings management in Spanish SMEs 173 the differences between voluntary and mandatory audits. As far as we know, this is the first study to test whether voluntary audits have a different effect on earnings quality than mandatory audits. On the other hand, it contributes to the study about the relationship between audit quality and earnings management in private companies (Vander Bauwhede et al., 2003; Cano, 2007; Van Tendeloo and Vanstraelen, 2008). As previous papers show that the relationship between audit quality and earnings management is affected by the level of litigation risk, we examine if this relation is also affected among SMEs. Moreover, we extend the study of audit quality, proxied by the auditor size, by considering a Middle-Tier level of auditors (Boone et al., 2010; Swanquist et al., 2012). Finally, the paper is also relevant for business in ethics, since earnings management is an ethical dilemma for accountants (He and Ho, 2011; He and Yang, 2014), and the ethical perceptions of it may vary between managers and auditors, because of their different motivations (Kaplan, 2001a,b). The rest of the paper is organized as follows: in Section ‘‘Literature review and hypothesis development’’ we review previous literature and develop our research hypotheses; Section ‘‘Empirical study’’ describes the sample and the research design; Section ‘‘Empirical results’’ reports the results of the main analysis; Section ‘‘Additional analyses’’ reports the results of the additional tests; and Section ‘‘Conclusions’’ presents our conclusions and the limitations of the study. Literature review and hypothesis development Audit vs. non-audit and accounting quality Auditing, as an activity consisting in the revision of the financial information, performs a relevant role in guaranteeing the credibility and reliability of the financial information. This assurance is provided by three sub-roles of the audit function (Cano and Sánchez, 2012): (i) the information role, which improves the credibility of accounting information and helps to reduce financing costs (Kim et al., 2011); (ii) the monitoring role, which helps to improve the quality of the accounting information, by reducing the opportunistic behaviour of managers (Dedman and Kausar, 2012); and (iii) the insurance role, which guarantees that users can rely on the audited financial information because of the responsibility auditors assume in case of audit failures (Melumad and Thoman, 1990; Khurana and Raman, 2004; Mansi et al., 2004). Empirical evidence about the information role of audits (Blackwell et al., 1998; Kim et al., 2011; Minnis, 2011; Niemi et al., 2012; Huguet and Gandía, 2014) supports the idea that auditing helps to improve the credibility of the financial statements, i.e. audited information is perceived to be of higher quality that the unaudited one. However, empirical evidence about the monitoring role is scarce. As stated by Dechow et al. (2010), although the basic premise that auditors could mitigate misstatements is straightforward, compelling empirical evidence is limited. Only a few studies (Minnis, 2011; Ojala et al., 2011; Dedman and Kausar, 2012) examine the effect of voluntary audits on accounting quality, while there is no evidence on the role of mandatory audits. Minnis (2011) examines the role of the auditor in the setting of US private companies, where audits are voluntary, and finds that auditing helps to reduce the cost of debt, because lenders place more weight on audited financial information when setting the interest rate. Furthermore, he finds that this increased credibility of the audited information is due to the increase of the actual accounting quality, because accruals from audited financial statements are better predictors of future cash flows and thus are more informative. On the other hand, Dedman and Kausar (2012) examine the effects that the change from mandatory to voluntary audits had on the credit ratings for UK private companies and find that those firms that decided to be voluntarily audited, even though reported lower average profits, obtained upgrades to their ratings, i.e. their audited information was perceived as being of higher quality. Moreover, they also find that voluntarily audited companies report more conservative financial statements. Taking into account what has been previously stated, and in line with the monitoring role, we expect that auditors oversee the relationship between the company and its stakeholders, by checking the accounting and ensuring that financial statements have been properly prepared, thus as audits would be a constraint to earnings management, audited SMEs should have lower discretionary accruals than the non-audited ones: H1a. Audited SMEs report a significantly lower level of earnings management than non-audited SMEs. Nevertheless, we have to note that voluntary audits may not necessarily involve higher accounting quality, because some firms may choose to be voluntarily audited to increase the perceived quality of their financial statements, without a true commitment on accounting quality. Daske et al. (2013) find that some ‘‘label’’ companies adopt voluntarily IAS/IFRS but do not make material changes to their financial reporting, i.e. they are willing to feign that they report according to IAS/IFRS, when they actually do not. In this line, Koren et al. (2014) examine a sample of Slovenian small firms and find that, among voluntarily audited companies, only those audited by Big 4 auditors report financial information of higher quality. Their results suggest that the rest of voluntarily audited companies choose to be audited only to pretend higher perceived quality. Therefore, it is not clear whether voluntary audits actually constrain earnings management: H1b. Voluntarily audited SMEs report a significantly lower level of earnings management than non-audited SMEs. On the other hand, mandatory audits are expected to guarantee a minimum accounting quality (Ruiz and Gómez, 2008). Moreover, as stated by Dedman and Kausar (2012), there is a concern among audit firms that a potential consequence of audit exemption may be a reduction in the quality of the financial reporting. Therefore, companies that shun the audit requirement would report accounting information 174 D. Huguet, J.L. Gandía of lower quality. However, an alternate view is that some mandatorily audited companies may be only passive compliant with the audit requirement, and thus may choose more permissive auditors. Moreover, among the SMEs, it has been argued that the financial information has a more limited usefulness, compared to the larger companies, because lenders rely more in alternative information sources (Berger and Udell, 2006), so mandatory audits would be a legal requirement, rather than a social need (Navarro and Martínez, 2004), what supports the idea of passive compliant companies. Therefore, we also formulate H1 for mandatory audits: H1c. Mandatorily audited SMEs report a significantly lower level of earnings management than SMEs non-compliant with the audit requirement. Voluntary audits vs. mandatory audits and accounting quality There is scarce research about whether the differences between voluntary and mandatory audits involve differences in the audit outcomes (Lennox and Pittman, 2011; Kim et al., 2011; Huguet and Gandía, 2014, 2015). Lennox and Pittman (2011) examine a sample of UK small companies and find that voluntary audits have a signalling effect that disappears when companies are mandatorily audited, and companies that choose to be voluntarily audited benefit from upgrades in their credit ratings. These results are similar to those obtained by Kim et al. (2011), who find that Korean companies with voluntary audits have greater interest savings than those mandatorily audited. In the Spanish setting, Huguet and Gandía (2014) do not find a significant relationship between voluntary audits and the cost of debt, but find that companies that breach the audit requirement have a higher cost of debt that those mandatorily audited, what suggests an asymmetric effect of voluntary/mandatory audits on the cost of debt. In a more recent paper, Huguet and Gandía (2015) examine differences in audit pricing between voluntary and mandatory audits. Therefore, previous literature generally shows that voluntary and mandatory audits involve differences in their audit outcomes. To date, however, there is no research about whether accounting quality is affected by the audit status. Nevertheless, and considering prior literature, we can expect that differences in earnings quality may arise. Following the reasoning for H1b and H1c, we can expect that voluntarily audited companies have a higher commitment with accounting quality, while some of the mandatorily audited companies may be only passive compliant. It may involve that, although mandatory audits may have a positive effect on earnings quality, the effect of voluntary audits may be even stronger. On a competing view, we can expect that voluntarily audited companies may report higher levels of earnings management than mandatorily audited companies, because auditors face lower visibility and litigation risks in the voluntary setting, and thus may be more permissive than when performing mandatory audits. Therefore, we formulate our second Hypothesis in null form: H2. There are no differences in the level of earnings management between voluntarily audited SMEs and the mandatorily audited ones. Audit quality and accounting quality Most of previous literature examines the effect of specific characteristics of audits, proxies for audit quality, on accounting quality and earnings management, because it is considered that high quality audits are a constraint to earnings management and an element that improves accounting quality (Becker et al., 1998; Balsam et al., 2003; Cano, 2007). Several characteristics of auditors have been used as a measure of audit quality, such as auditor specialization (Balsam et al., 2003), audit and non-audit fees (Basioudis et al., 2008; Carmona and Momparler, 2011), or auditor tenure (Chung et al., 2005; Carey and Simnett, 2006), but auditor size, through the dichotomy Big 4/Rest of auditors, is the most common proxy for audit quality, because large auditors are considered more professionally competent (Becker et al., 1998; Francis et al., 1999) and independent (DeAngelo, 1981). Moreover, it is assumed that large auditors face greater losses than the rest of auditors, not only economic losses derived from the insurance role and the perception that large auditors are ‘‘deep pockets’’ (Khurana and Raman, 2004; Mansi et al., 2004), but also reputational (DeAngelo, 1981). In general, these studies find that Big 4 auditors are associated with higher accounting quality, because Big 4 auditors restrain earnings management more than small auditors (Becker et al., 1998; Francis et al., 1999; Balsam et al., 2003; Cano, 2007; Jara and López, 2007) and are related with a higher level of conditional conservatism (Chung et al., 2003; Francis and Wang, 2008; Cano, 2010). In the Spanish setting, although Navarro and Martínez (2004) do not find a significant effect, later papers on both public and private companies (Jara and López, 2007; Cano, 2007) find that Big 4 auditors are associated to lower levels of earnings management. Although most of previous literature shows a negative association between auditor size and earnings management, later papers go into the relationship between audit quality and earnings management, by looking for factors that affect this relation, reducing the effectiveness of Big 4 auditors to constrain earnings management, such as the auditor conservatism (Kim et al., 2003; Cano, 2010) and the visibility and litigation risk of auditors (Vander Bauwhede et al., 2003; Van Tendeloo and Vanstraelen, 2008). Regarding the auditor conservatism, the auditors’ behaviour is generally conservative, in the sense that they have preference for income-decreasing accounting choices instead of income-increasing ones. This behaviour involves two negative effects for accounting quality: (i) it increases not only the level of conditional conservatism, (i.e. asymmetrical recognitions of good and bad news, which is considered desirable), but also the level of unconditional conservatism (early recognition of losses, independently of the news), which can reduce the quality of accounting information (Cano, 2010); and (ii) when managers have incentives to manage downward, Big 4 auditors are not effective to Audit and earnings management in Spanish SMEs 175 deter earnings management (Kim et al., 2003; Francis and Krishnan, 1999). With regard to the visibility and litigation risk, we stated before that Big 4 auditors have incentives to perform higher quality audits because of the potential losses they face, both economic (derived from the litigation risk) and reputational (derived from their visibility). In settings with lower litigation risk and visibility, these incentives may be reduced or even disappear, thus Big 4 auditors would not be more effective than the rest of auditors in constraining earnings management (Vander Bauwhede et al., 2003; Van Tendeloo and Vanstraelen, 2008). On the other hand, previous studies do not test the role of Middle-Tier auditors, in spite of being considered to provide a similar quality than Big 4 auditors (Boone et al., 2010; Sundgren and Svanström, 2013). We examine if Big 4 and Middle-Tier auditors provide similar audit quality. Therefore, we formulate our third Hypothesis for both large auditors (Big 4 and Middle-Tier) and separately for Big 4 auditors: H3a. The level of earnings management is significantly lower for SMEs audited by large auditors than for SMEs audited by small auditors. H3b. The level of earnings management is significantly lower for SMEs audited by Big 4 auditors than for SMEs audited by non-Big 4 auditors. Empirical study Sample and descriptive statistics For the selection of the sample we have used SABI, a database that contains financial data from financial statements of Spanish companies submitted to Registro Mercantil (Spanish Company Register). Our sample period covers since 2008---2013. We initially select private companies which have been, for the whole sample period, below at least two out of the three following thresholds: D 6,000,000 Total Assets, D 12,000,000 Turnover and 50 employees. These limits are the upper thresholds established by the Directive, 2013/34/EU to consider a company is small and thus to be exempt from the audit requirement. In practice, however, most of EU members apply lower Statutory Audit Thresholds (SAT). In Spain, private companies are exempt to be audited if they do not exceed two out of these criteria for two consecutive years: total assets of D 2,850,000 (D 2,374,000 until 2007); net turnover of D 5,700,000 (D 4,748,000 until 2007); and (iii) 50 employees. The use of the upper EU limits let us examine the effects of both voluntary and mandatory audits on earnings management, but avoiding an excessive variation in company size within the sample. Therefore, our sample includes small companies below Spanish SAT (and thus ‘‘a priori1’’ exempt 1SABI does not contain information about the reason companies are audited, thus we cannot differentiate between mandatory or voluntary audits when companies are below legal thresholds. from the audit requirement) and the ‘‘small’’ medium companies above Spanish SAT (i.e. required to be audited). Since the selection process excludes the companies that have been above the EU thresholds in some years of the sample period and this could be producing a survivorship bias, we do an additional test by including the companies below the EU thresholds in any of the years of the period (rather than in the whole period). Furthermore, in order to achieve a sample with more homogeneous characteristics, we only include companies that have been audited at least once over the sample period, because companies that have never been audited are probably much smaller than those near SAT, and the accounting information of audited companies and never-audited companies may be not comparable. In an additional test, in order to test if this exclusion affects our results, we include the observations of the never-audited companies. For the same reason, we also exclude the observations of companies that are considered micro-firms under Directive, 2013/34/EU (companies that do not meet two of the following thresholds: (i) D 350,000 in total assets; (ii) D 700,000 in net turnover; and (iii) less than 10 employees). Companies belonging to financial and insurance industries, firms having unlimited liability2and firms with share participation by public entities are also excluded. Finally, we eliminate observations that have no information to calculate accruals and also observations with strange values (negative values for assets, debt or financing expenses). Table 1 Panel A shows that the final sample has 34,562 firm-year observations from 8066 companies, with 9181 observations from non-audited companies (7634 of them are from observations below SAT and 908 are observations from companies that breach the audit requirement), and 25,381 observations from audited companies (4278 below SAT and 20,683 above SAT and thus mandatorily audited because of size). As companies must meet the thresholds for two consecutive years, 1967 observations (1328 from audited companies and 639 from unaudited companies) are not classified in either range because they do not satisfy the criteria to be considered either below or above SAT and thus are excluded from our analysis. We have to remark that the relatively low number of observations below SAT is explained by the exclusion of the never-audited companies. As we have stated before, we do an additional analysis including the never-audited companies to test how results are affected. Table 1 Panel B shows the sample distribution of audited SMEs by auditor choice (Big 4/Middle-Tier/rest of auditors). It should be pointed out the low proportion of companies audited by large auditors, although the proportion of companies that choose to be audited by either a Big 4 or a Middle-Tier auditor is higher among the observations below SAT than among the larger ones. In Table 2 we show the descriptive statistics of the continuous variables. We observe that, on average, discretionary accruals are negative, thus consistent with prior literature (Arnedo et al., 2007; García Lara et al., 2005), earningsdecreasing behaviour is more pervasive among private 2Spanish companies with limited liability are the Sociedad Anónima (SA) and Sociedad Limitada (SL). 176 D. Huguet, J.L. Gandía Table 1 Sample distribution. Panel A: Sample distribution by audit status Below SAT Above SAT Unknown Total Year No audit Audit Total No audit Audit Total No audit Audit Total No audit Audit Total 2008 56 54 110 118 2490 2608 100 355 455 274 2899 3173 2009 1121 487 1608 207 3911 4118 146 175 321 1474 4573 6047 2010 1799 668 2467 159 3327 3486 138 310 448 2096 4305 6401 2011 1574 931 2505 135 3442 3577 92 183 275 1801 4556 6357 2012 1515 985 2500 122 3557 3679 99 151 250 1736 4693 6429 2013 1569 1153 2722 167 3048 3215 64 154 218 1800 4355 6155 Total 7634 4278 11,912 908 19,775 20,683 639 1328 1967 9181 25,381 34,562 Panel B: Sample distribution of audited companies by auditor choice Below SAT Above SAT Unknown Total Year Small M-Tier Big Small M-Tier Big Small M-Tier Big Small M-Tier Big 2008 32 14 8 2083 240 167 306 31 18 2421 285 193 2009 367 68 52 3238 406 267 135 18 22 3740 492 341 2010 496 91 81 2684 362 281 249 38 23 3429 491 385 2011 715 111 105 2761 381 300 147 25 11 3623 517 416 2012 729 135 121 2849 379 329 106 27 18 3684 541 468 2013 858 160 135 2435 350 263 121 18 15 3414 528 413 Total 3197 579 502 16,050 2118 1607 1064 157 107 20,311 2854 2216 (%) (74.73%) (13.53%) (11.73%) (81.16%) (10.71%) (8.13%) (80.12%) (11.82%) (8.06%) (80.02%) (11.24%) (8.73%) firms. On the other hand, on average, unaudited companies have more negative total accruals and higher absolute discretionary accruals. Moreover, audited companies are larger, have higher leverage and less liquidity, and are older. Research model We test the Hypotheses with the following regression models: EMit = ˛ + ˇ1AUDITit + ˇ2VOLit + ˇ3LARGEit + ˇ4VOL LARGEit + ˇ5BIGit + ˇ6VOL BIGit + CONTROL + εit (1) EMit = ˛ + ˇ1LARGEit + ˇ2VOL LARGEit + ˇ3BIGit + ˇ4VOL BIGit + CONTROL + εit (2) The dependent variable in both models is the level of earnings management (EM). Since this variable is not directly observed, we use a proxy based on the level of discretionary accruals (DA). The discretionary accruals models assume that the accruals that are not explained by innate factors, which are a consequence of the company’s activities, are a measure of the level of earnings management. We estimate discretionary accruals3using the Jones Model (1991) and the absolute value of the discretionary accruals is considered the measure of earnings management. Some studies (Hribar and Nichols, 2007; Francis and Wang, 2008; Dedman and Kausar, 2012) use the signed discretionary accruals rather than their absolute value. Following Arnedo et al. (2007), we consider that SMEs may have incentives to manage earnings downward because of tax issues, whereas income increasing behaviours may be less frequent because of the limited usefulness of accounting information for lenders (García Lara et al., 2005; Berger and Udell, 2006; Gill de Albornoz and Illueca, 2007). Therefore, we also regress the models separately for positive and negative accruals. In an additional analysis, we use alternative measures of earnings management. Model [1] includes AUDIT, which equals 1 when companies are audited and 0 otherwise. We have to note, however, that SABI presents problems with the identification of audited companies. When a company is not audited in t, the data is left blank in the database. However, some of these blank data are really missing observations. Therefore, ‘‘blank data’’ may be either unaudited or audited (but missing) observations, thus some audited observations may be erroneously considered non-audited. This limitation can be partially overcome by changing some of these ‘‘blank data’’ observations to ‘‘audited’’ observations (Huguet and Gandía, 2014). Considering that the shortest auditor tenure 3We need at least 6 observations by each industry-year to estimate discretionary accruals. Audit and earnings management in Spanish SMEs 177 Table 2 Descriptive statistics of continuous variables. Panel A: Distributional properties of continuous variables (32,595 observations) Variable Mean Std. dev. 1% 25% 50% 75% 99% +/−DA −0.0072 0.1360 −0.3766 −0.0694 −0.0075 0.0534 0.3824 |DA| 0.0907 0.1016 0.0010 0.0275 0.0617 0.1187 0.4592 TA −0.0368 0.1443 −0.4152 −0.1012 −0.0375 0.0243 0.3614 SIZE 8.6574 0.7361 7.0198 8.2198 8.5560 8.9942 10.9354 LEV 0.5711 0.2383 0.0886 0.3934 0.5807 0.7468 1.1872 GROWTH −0.0308 0.2783 −0.6674 −0.1679 −0.0388 0.0721 1.1205 ROA 0.0137 0.0681 −0.2425 −0.0054 0.0110 0.0387 0.2376 LIQ 1.9826 1.8311 0.1826 1.0319 1.4221 2.1873 11.1468 AGE 22.0518 10.9984 5.0000 14.0000 21.0000 28.0000 54.0000 Panel B: Mean and standard deviation of variables by audit status Non audited Audited Test for mean differences Variable Obs. Mean S. dev. Obs. Mean S. dev. diff t p-Value +/−DA 8542 −0.0058 0.1432 24,053 −0.0077 0.1334 0.0020 1.14 0.127 |DA|8542 0.0935 0.1086 24,053 0.0897 0.0990 0.0038 2.94 0.002 TA 8542 −0.0396 0.1611 24,053 −0.0358 0.1378 −0.0038 −2.07 0.019 SIZE 8542 8.4460 0.7856 24,053 8.7324 0.7026 −0.2865 −31.36 0.000 LEV 8542 0.5663 0.2467 24,053 0.5728 0.2352 −0.0065 −2.17 0.015 GROWTH 8542 0.0212 0.3519 24,053 −0.0492 0.2443 0.0704 20.22 0.000 ROA 8542 0.0079 0.0735 24,053 0.0157 0.0659 −0.0078 −9.16 0.000 LIQ 8542 2.0992 2.0184 24,053 1.9411 1.7580 0.1581 6.86 0.000 AGE 8542 20.9348 10.5710 24,053 22.4485 11.1195 −1.5137 −10.95 0.000 Non audited below SAT Audited below SAT Test for mean differences Variable Obs. Mean S. dev. Obs. Mean S. dev. Diff t p-Value +/−DA 7634 −0.0057 0.1440 4278 −0.0111 0.1402 0.0054 1.98 0.024 |DA|7634 0.0938 0.1094 4278 0.0901 0.1080 0.0036 1.75 0.040 TA 7634 −0.0402 0.1634 4278 −0.0479 0.1491 0.0077 2.53 0.006 SIZE 7634 8.4186 0.7993 4278 8.5446 0.9172 −0.1261 −7.83 0.000 LEV 7634 0.5610 0.2482 4278 0.5426 0.2537 0.0184 3.86 0.000 GROWTH 7634 0.0319 0.3619 4278 0.0261 0.3329 0.0058 0.86 0.195 ROA 7634 0.0070 0.0739 4278 0.0052 0.0745 0.0018 1.31 0.096 LIQ 7634 2.1361 2.0638 4278 2.0788 2.1163 0.0573 1.44 0.075 AGE 7634 20.9925 10.6194 4278 22.4353 11.9685 −1.4427 −6.79 0.000 Non audited above SAT Audited above SAT Test for mean differences Variable Obs. Mean S. dev. Obs. Mean S. dev. Diff t p-Value +/−DA 908 −0.0065 0.1364 19,775 −0.0070 0.1318 0.0005 0.12 0.453 |DA| 908 0.0908 0.1020 19,775 0.0896 0.0970 0.0011 0.35 0.364 TA 908 −0.0343 0.1410 19,775 −0.0332 0.1351 −0.0011 −0.24 0.406 SIZE 908 8.6765 0.6132 19,775 8.7731 0.6397 −0.0966 −4.46 0.000 LEV 908 0.6108 0.2292 19,775 0.5794 0.2305 0.0315 4.02 0.000 GROWTH 908 −0.0684 0.2348 19,775 −0.0655 0.2171 −0.0029 −0.39 0.349 ROA 908 0.0148 0.0693 19,775 0.0180 0.0637 −0.0032 −1.46 0.072 LIQ 908 1.7892 1.5525 19,775 1.9114 1.6689 −0.1222 −2.16 0.015 AGE 908 20.4493 10.1479 19,775 22.4514 10.9274 −2.0020 −5.41 0.000 in Spain is three years, if: (a) a company is audited in t − 1 and t + 1; and (b) is above SAT in t, we consider the company is audited in t. We applied this improvement and 664 observations with blank data were changed to audited observations. Furthermore, we include VOL, which equals 1 when company is voluntarily audited and 0 otherwise. The introduction of this variable let us test the differential effect of voluntary audits over mandatory audits: the effect of mandatory audits is observed from ˇ1, while the sum of ˇ1+ ˇ2captures 178 D. Huguet, J.L. Gandía the effect of voluntary audits. Therefore, we can test simultaneously the differences between audited and unaudited observations (Hypothesis 1), as well as the differences between voluntary and mandatory audits (Hypothesis 3). On the other hand, the auditor choice may affect the level of earnings management among the audited companies, so we include two audit-based variables more: BIG, which equals 1 for companies audited by a Big 4 auditor and 0 otherwise; and LARGE, which equals 1 when companies are audited by large auditors (either Middle-Tier or Big 4 audit firms) and 0 otherwise. We have considered an audit firm is a Middle-Tier auditor when this firm has revenues for audit fees higher than D 9,000,000 in 2010 and 2011.4These firms, although smaller than the Big 4, operate in several regions of the country and have also significant revenues from consulting and tax services. BIG captures the differences between Big 4 and Middle-Tier auditors, while LARGE captures the differences between both types of large audit firms and the rest of auditors. Since the effect of the Big 4 and Middle-Tier auditors may change between the voluntary and the mandatory setting, we also include the interaction terms between these two variables and VOL. Regarding Model [2], we use this model in the sample of audited companies, and thus we exclude AUDIT from it. Results from this model support the results of Model [1] for Hypotheses 2 and 3, but focusing only on the differences between auditors. We include in both models a set of control variables that have been used in previous research, which are defined in Appendix. Company size (SIZE) is measured as the natural logarithm of total assets (Balsam et al., 2003; Kim et al., 2003; Van Tendeloo and Vanstraelen, 2008). Although earlier papers predict a positive association between earnings management and size as a consequence of decreasing-income accruals to avoid political costs (Jensen and Meckling, 1976; Watts and Zimmerman, 1986), recent papers show that company size is positively associated with earnings quality (from whom earnings management is an inverse measure) because of the economies of scale to produce higher quality information (Ashbaugh-Skaife et al., 2007) and a stricter control by the public and government (Arnedo et al., 2007). On the other hand, we measure leverage (LEV) as the ratio of total liabilities to total assets (DeFond and Jiambalvo, 1994; Becker et al., 1998; Reynolds and Francis, 2000). Since managers have incentives to not meet the debt covenants, we expect a positive association between leverage and earnings management (DeFond and Jiambalvo, 1994). Company growth (GROWTH) is measured as the sales growth (Khurana and Raman, 2004; Chen et al., 2008). We expect that companies with problems have more incentives to engage in earnings management, so the association between GROWTH and EM should be negative. Profitability (ROA) is measured as the Return-On-Assets, calculated as the ratio of earnings before interest and taxes to total assets at the beginning of the period (Velury and Jenkins, 2006; Van Tendeloo and Vanstraelen, 2008). We measure liquidity (LIQ) as the ratio of current assets to current 4These data have been collected from the Spanish newspaper Expansión. This newspaper prepares a yearly ranking of the 40---50 top auditors in Spain by total fees and audit fees. liabilities (Butler et al., 2004; Caramanis and Lennox, 2008). N EARN is a dummy that equals 1 if the company has negative earnings and 0 otherwise (Francis et al., 1999; Jara and López, 2007). We expect a positive association between NEARN and EM. Finally, the age of the company (AGE) is measured as its age since its creation (Myers et al., 2003; Chen et al., 2008). The models include year dummies to control for unobserved time-specific effects common to all companies. Previous literature shows that the use of audit-based variables is often affected by endogeneity problems (Kim et al., 2011; Cano and Sánchez, 2012). They may appear because the test variables LARGE and BIG are the result of a corporate decision to choose the type of auditor rather than a random assignment, a problem also present in AUDIT and VOL for observations below SAT considering that these firms decide to be (or not) audited. Therefore, OLS estimations are not proper because they are biased. Some studies try to mitigate the endogeneity problems through a Heckman two-stage approach (Chaney et al., 2004; Mansi et al., 2004; Pittman and Fortin, 2004; Monterrey and Sánchez, 2007). Nevertheless, recent literature (Clatworthy et al., 2009; Larcker and Rusticus, 2010; Lennox et al., 2012) shows that the Heckman results depend on a proper selection of the instrumental variables, are fragile, and can be even more unreliable than the OLS estimation. Kim et al. (2011) and Lennox et al. (2012) state that the use of fixed-effects regressions can mitigate the potential endogeneity issues as long as the unobserved source of endogeneity is timeinvariant. Therefore, we estimate Eqs. (1) and (2) using a firm fixed-effects (FE) regression procedure. However, it is worth noting that the FE estimation is only a partial solution to endogeneity problems, because it does not allow a simultaneous equation approach when the relation between the dependent and the test variable is bidirectional, and because the FE regression only solves the endogeneity problem when the source of endogeneity is fixed over time. Empirical results This section presents the results of the main analysis. First, we compute a correlation matrix (Table 3) to examine potential multicollinearity problems. The highest correlation is 0.9515 between discretionary and total accruals, showing that most of accruals are abnormal according to our model to estimate discretionary accruals. Nevertheless, TA is not used in the regressions, so there are no problems associated with this association. Furthermore, we use TA as the dependent variable in an additional analysis. Other high correlations are among AUDIT and SIZE (0.6088), BIG and LARGE (0.6632), and ROA and NEG EARN (0.5347). However, as all the correlations are below 0.80, we do not expect collinearity problems (Judge et al., 1988; Firth, 1997; Carmona and Momparler, 2011). We then run models [1] and [2] in the samples explained in Section ‘‘Sample and descriptive statistics’’ using absolute and signed discretionary accruals as our measure of earnings quality. Table 4 Panel A shows the results from Model [1], whereas Panel B shows the results from Model [2] in the sub-sample of audited companies. Audit and earnings management in Spanish SMEs 185 Appendix. See Table A1. Table A1 Description of variables. Variable name Description Dependent variable EM Earnings management measure |DA|Absolute discretionary accruals +/− DA Signed discretionary accruals ABWCA Abnormal working capital accruals (DeFond and Park, 2001) ABACC Abnormal accruals (Francis and Wang, 2008) TA Total accruals Test variables AUDIT Dummy that equals 1 if company is audited, 0 otherwise VOL Dummy that equals 1 if company is voluntarily audited, 0 otherwise BIG Dummy that equals 1 if company is audited by a Big 4 auditor, 0 otherwise LARGE VOL LARGE Dummy that equals 1 if company is audited by either a Middle-Tier auditor or a Big 4 auditor, 0 otherwise Interaction term of VOL and LARGE VOL BIG Interaction term of VOL and BIG Control variables SIZE Company size, measured as the natural logarithm of total assets LEV Leverage, measured as the ratio of interest bearing debt to total assets GROWTH Growth of the company, measured as the growth of sales ROA Profitability, measured as the ratio of earnings before interest and taxes to total assets LIQ Level of liquidity, measured as the ratio of current assets to current liabilities NEG EARN Dummy variable that equals 1 if company has negative earnings and 0 otherwise AGE Age of the company DUMyear Year dummies References Allee, K.D., Yohn, T.L., 2009. The demand for financial statements in an unregulated environment: an examination of the production and use of financial statements by privately held small business. Account. Rev. 84, 1---25. Arnedo, L., Lizarraga, F. , Sánchez, S., 2007. Does public/private status affect the level of earnings management in code-law contexts outside the United States? A study based on the Spanish case. Int. J. Account. 42, 305---328. Ashbaugh-Skaife, H., Collins, D., Kinney, W., 2007. The discovery and reporting of internal control deficiencies prior to SOX- mandated audits. J. Account. Econ. 44, 166---192. Balsam, S., Krishnan, J., Yang, J.S., 2003. Auditor industry specialization and earnings quality. Auditing: J. Pract. Theory 22, 71---97. Basioudis, I.G., Papakonstantinou, E., Geiger, M.A., 2008. Audit fees, non-audit fees and auditor going-concern reporting decisions in the United Kingdom. Abacus 44, 284---309. Becker, C.L., DeFond, M.L., Jiambalvo, J., Subramanyam, K.R., 1998. The effect of audit on earnings management. Contemp. Account. Res. 15, 1---24. Berger, A.N., Udell, G.F., 2006. A more complete conceptual framework for SME finance. J. Bank. Financ. 30, 2945---2966. Blackwell, D.W., Noland, T.R., Winters, D.B., 1998. The value of auditor assurance: evidence from loan pricing. J. Account. Res. 36, 57---70. Boone, J.P., Khurana, I.K., Raman, K.K., 2010. Do the Big 4 and the Second-tier firms provide audits of similar quality? J. Account. Public Policy 29, 330---352. Butler, M., Leone, A.J., Willenborg, M., 2004. An empirical analysis of auditor reporting and its association with abnormal accruals. J. Account. Econ. 37, 139---165. Cano, M., 2007. Tama˜ no del auditor y calidad de auditoría en las empresas espa˜ nolas no cotizadas (Auditor size and audit quality in Spanish private companies). Span. J. Account. Financ. 36, 481---507. Cano, M., 2010. Big auditors, private firms and accounting conservatism: Spanish evidence. Eur. Account. Rev. 19, 131---159. Cano, M., Sánchez, S., 2012. The value of audit quality in public and private companies: evidence from Spain. J. Manage. Gov. 16, 683---706. Caramanis, C., Lennox, C., 2008. Audit effort and earnings management. J. Account. Econ. 45, 116---138. Carey, P. , Simnett, R., 2006. Audit partner tenure and audit quality. Account. Rev. 81, 653---676. Carmona, P. , Momparler, A., 2011. Nonaudit services provided by incumbent auditors and earnings management: evidence of auditor independence from an EU country. Span. J. Account. Financ. 40, 587---612. Chaney, P.K., Jeter, K.C., Shivakumar, L., 2004. Self-selection of auditors and pricing in private firms. Account. Rev. 79, 51---72. Chen, C.Y., Lin, C.J., Lin, Y.C., 2008. Audit partner tenure, audit firm tenure, and discretionary accruals: does long auditor tenure impair earnings quality? Contemp. Account. Res. 25, 415---445. Chung, R., Firth, M., Kim, J.B., 2003. Auditor conservatism and reported earnings. Account. Bus. Res. 33, 19---32. Chung, R., Firth, M., Kim, J.B., 2005. Earnings management, surplus free cash flow, and external monitoring. J. Bus. Res. 58, 766---776. Clatworthy, M.A., Makepeace, G.H., Peel, M.J., 2009. Selection bias and the Big Four premium: new evidence using Heckman and matching models. Account. Bus. Res. 39, 139---166. Daske, H., Hail, L., Leuz, C., Verdi, R., 2013. Adopting a label: heterogeneity in the economic consequences around IAS/IFRS adoptions. J. Account. Res. 51, 495---547. DeAngelo, L.E., 1981. Auditor size and audit quality. J. Account. Econ. 3, 183---199. Dechow, P. , Sloan, R.G., Sweeney, A.P., 1995. Detecting earnings management. Account. Rev. 70, 193---225. Dechow, P. , Ge, W., Schrand, C., 2010. Understanding earnings quality: a review of the proxies, their determinants and their consequences. J. Account. Econ. 50, 344---401. 186 D. Huguet, J.L. Gandía Dedman, E., Kausar, A., 2012. The impact of voluntary audit on credit ratings: evidence from UK private firms? Account. Bus. Res. 42, 397---418. DeFond, M., Jiambalvo, J., 1994. Debt covenant violation and manipulation of accruals: accounting choice in troubled companies. J. Account. Econ. 18, 145---176. DeFond, M., Park, C.W., 2001. The reversal of abnormal accruals and the market valuation of earnings surprises. Account. Rev. 76, 375---404. Directive, 2013/34/EU of the European Parliament and the Council of 26 June 2013. On the annual financial statements, consolidated financial statements and related reports of certain types of undertakings, amending Directive 2006/43/EC of the European Parliament and of the Council and repealing Council Directives 78/660/EEC and 83/349/EEC. European Commission, 2010. Green Book. Audit Policy: Lessons from the Crisis. European Commission, 2012. SBA Fact Sheet 2012. Spain. Firth, M., 1997. The provision of non-audit services and the pricing of audit fees. J. Bus. Financ. Account. 24, 511---525. Francis, J.R., Krishnan, J., 1999. Accounting accruals and auditor reporting conservatism. Contemp. Account. Res. 16, 135---165. Francis, J.R., Maydew, E.L., Sparks, H.C., 1999. The role of Big 6 auditors in the credible reporting of accruals. Auditing: J. Pract. Theory 18, 17---34. Francis, J.R., Wang, D., 2008. The joint effect of investor protection and Big 4 audits on earnings quality around the world. Contemp. Account. Res. 25, 157---191. García Lara, J.M., García Osma, B., Mora, A., 2005. The effect of earnings management on the asymmetric timeliness of earnings. J. Bus. Financ. Account. 32, 691---726. García Osma, B., Gill de Albornoz, B., Gisbert, A., 2005. La investigación sobre earnings management (Research on earnings management). Span. J. Account. Financ. 34, 1001---1033. Gill de Albornoz, B., Illueca, M., 2007. La calidad de los ajustes por devengo no afecta al coste de la deuda de las pymes espa˜ nolas (The quality of accruals does not have effect on cost of debt in Spanish SMEs). Investig. Econ. 31, 79---117. He, L., Yang, R., 2014. Does industry regulation matter? New evidence on audit committees and earnings management. J. Bus. Ethics 123, 573---589. He, L., Ho, S., 2011. Monitoring costs, managerial ethics and corporate governance: a modeling approach. J. Bus. Ethics 99, 623---635. Hribar, P. , Nichols, D.C., 2007. The use of unsigned earnings quality measures in tests of earnings management. J. Account. Res. 45, 1017---1053. Huguet, D., Gandía, J.L., 2014. Cost of debt capital and audit in Spanish SMEs. Span. J. Account. Financ. 43, 266---289. Huguet, D., Gandía, J.L., 2015. Audit fees and voluntary audit. In: 38th EAA Congress. Jara, M.A., López, F.J., 2007. Auditoría y discrecionalidad contable en la gran empresa no financiera espa˜ nola (Auditing and earnings management in the large non-financial Spanish firm). Span. J. Account. Financ. 36, 569---594. Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. J. Financ. Econ. 3, 305---360. Jones, J., 1991. Earnings management during import relief investigations. J. Account. Res. 29, 193---228. Judge, G.G., Hill, R.C., Griffiths, W.E., Lutkepohl, H., Lee, T.C., 1988. Introduction to the Theory and Practice of Econometrics. Wiley, New York. Kaplan, S.E., 2001a. Ethically related judgments by observers of earnings management. J. Bus. Ethics 32, 285---298. Kaplan, S.E., 2001b. Further evidence on the ethics of managing earnings: an examination of the ethically related judgments of shareholders and non-shareholders. J. Account. Public Policy 20, 27---44. Khurana, I.K., Raman, K.K., 2004. Litigation risk and the financial reporting credibility of Big 4 versus non-Big 4 audits: evidence from Anglo-American countries. Account. Rev. 79, 473---495. Kim, J.B., Chung, R., Firth, M., 2003. Auditor conservatism, asymmetric monitoring, and earnings management. Contemp. Account. Res. 20, 323---359. Kim, J.B., Simunic, D.A., Stein, M.T., Yi, C.H., 2011. Voluntary audits and the cost of debt capital for privately held firms: Korean evidence. Contemp. Account. Res. 28, 1---40. Koren, J., Kosi, U., Valentincic, A., 2014. Does Financial Statement Audit Reduce the Cost of Debt of Private Firms? Working paper, Available at http://papers.ssrn.com/ sol3/papers.cfm?abstract id=2373987. Larcker, D.F., Rusticus, T.O., 2010. On the use of instrumental variables in accounting research. J. Account. Econ. 49, 186---205. Lennox, C., Francis, J.R., Wang, Z., 2012. Selection models in accounting research. Account. Rev. 87, 589---616. Lennox, C., Pittman, J.A., 2011. Voluntary audits versus mandatory audits. Account. Rev. 86, 1655---1678. Mansi, S.A., Maxwell, W.F., Miller, D.P., 2004. Does auditor quality and tenure matter to investors? Evidence from the bond market. J. Account. Res. 42, 755---793. Melumad, N.D., Thoman, L., 1990. On auditors and the courts in an adverse selection setting. J. Account. Res. 28, 77---120. Minnis, M., 2011. The value of financial statement verification in debt financing: evidence from private U.S. firms. J. Account. Res. 49, 457---506. Monterrey, J., Sánchez, A., 2007. Un estudio empírico de los honorarios del auditor. Cuadernos de Economía y Dirección de la Empresa 10, 81---110. Myers, J.N., Myers, L.A., Omer, T.C., 2003. Exploring the term of the auditor---client relationship and the quality of earnings: a case for mandatory audit rotation? Account. Rev. 78, 779---799. Navarro, J.C., Martínez, I., 2004. Manipulación contable y calidad del auditor. Un estudio empírico de la realidad espa˜ nola (Earnings management and audit quality in Spain: an empirical study). Span. J. Account. Financ. 33, 1025---1061. Niemi, L., Kinnunen, J., Ojala, H., Troberg, P. , 2012. Drivers of voluntary audit in Finland: to be or not to be audited? Account. Bus. Res. 42, 169---196. Ojala, H., Collis, J., Niemi, L., Kinnunen, J., 2011. Voluntary Audit, Earnings Quality and the Capital Structure of Small Companies. EUFIN. Pittman, J.A., Fortin, S., 2004. Auditor choice and the cost of debt capital for newly public firms. J. Account. Econ. 37, 113---136. Reynolds, J.K., Francis, J.R., 2000. Does size matter? The influence of large clients on office-level auditor reporting decisions. J. Account. Econ. 30, 375---400. Ruiz, E., Gómez, N., 2008. Discusión sobre Gobierno corporativo, conflictos de agencia y elección de auditor (Discussion of Corporate Governance, agency conflicts and auditor choice). Span. J. Account. Financ. 37, 136---156. Sundgren, S., Svanström, T. , 2013. Audit office size, audit quality and audit pricing: evidence from small- and medium-sized enterprises. Account. Bus. Res. 43, 31---55. Swanquist, Q., Shipman, J., Whited, R., 2012. Do Second Tier Auditors Provide First Tier Quality? Working Paper. University of Tennessee, Available at http://ssrn.com/abstract=2191071. Van Tendeloo, B., Vanstraelen, A., 2008. Earnings management and audit quality in Europe: evidence from the private client segment market. Eur. Account. Rev. 17, 447---469. Audit and earnings management in Spanish SMEs 187 Vander Bauwhede, H., Willekens, M., Gaeremynck, A., 2003. Audit firm size, public ownership, and firms’ discretionary accruals management. Int. J. Account. 38, 1---22. Velury, U., Jenkins, D.S., 2006. Institutional ownership and the quality of earnings. J. Bus. Res. 59, 1043---1051. Watts, R.L., Zimmerman, J.L., 1986. Positive Accounting Theory. Prentice-Hall. Wymenga, P. , Spanikova, V., Barker, A., Konings, J., Canton, E., 2012. EU SMEs in 2012: at the crossroads. Annual report on small and medium-sized enterprises in the EU, 2011/2012. Report for the European Commission.