Tourist Flow and Earnings Benchmarks

Abstract

This study examines the relationship between meeting earnings benchmarks and regional factors related to the hospitality industry. Our hypothesis assumes that the tourist flow in which a hotel is located influences the hotel’s performance, but it is not the same for all firms. Specifically, firms meeting earnings benchmarks, when it is documented that managers have strong incentives to engage in earnings management strategy, present a different relationship with regional tourist flow. The evidence suggests that regional tourist flow is associated with the corporate performance of firms but that there is an inflexion point for firms meeting earnings benchmarks.

Keywords

hotel management lodging financial performance strategy international

Introduction

The important role of financial information in the global markets is well known. Financial reporting is a key factor in satisfying capital providers’ contracting demands for information on firm performance and the actions of managers. Accounting earnings are one of the most important pieces of financial information that an investor uses to evaluate company performance. The quality of financial reports is associated with the usefulness of financial statements and the relevance of such statements to predicting future cash flow and firm performance.

Prior literature documents that managers are aware of various external targets (e.g., analysts’ forecasts, accounting-based bonus targets) that affects directly the firms’ financial performance and financial position. As the first and most intuitive income target, managers wish to avoid a negative net income (i.e., reach a positive net income) and a decrease in earnings (i.e., reach an earnings increase) in the financial statements (see, e.g., Burgstahler & Dichev, 1997; Degeorge, Patel, & Zeckhauser, 1999).

The benefits of meeting financial benchmarks and the costs of failing to do so are well documented in the literature. Empirical evidence shows that firms that reach earnings targets consistently are rewarded and firms missing earnings targets are penalized (see, e.g., Bartov, Givoly, & Hayn, 2002; Graham, Harvey, & Rajgopal, 2005; Skinner & Sloan, 2002). Agency theory provides an appropriate framework for explaining information asymmetry between managers and shareholders and the different objectives that each group pursues (Jensen & Meckling, 1976).

Most of the influential studies of earnings targets have been conducted on U.S. listed companies (Bartov et al., 2002; Burgstahler & Dichev, 1997; Degeorge et al., 1999; Durtschi & Easton, 2005, 2009; Jacob & Jorgensen, 2007; Kerstein & Rai, 2007; Skinner & Sloan, 2002). More recently, the earnings benchmarks literature has turned its attention toward nonlisted companies and specific industries or countries (Ball & Shivakumar, 2005; Burgstahler, Hail, & Leuz, 2006; Coppens & Peek, 2005; Jara & López Iturriaga, 2011; Parte & Such, 2011; Turner & Guilding, 2011).

There has been surprisingly little research in this area that is focused on the services sector. For example, Barber, Ghiselli, and Deale (2006); Kim and Gu (2005); and Guillet, Kucukusta, and Xiao (2012) analyze the relationship between CEO compensation and firm financial performance using a sample of restaurant firms. Turner and Guilding (2011) document earnings management through the selective capitalization or expensing of asset-related expenditures in the hotel industry. Parte and Such (2011) focus on the determinants of earnings benchmarks in the Spanish hotel industry.

This article evaluates the relationship between meeting earnings benchmarks and regional factors related to the hospitality industry. We focus on a fundamental sector for the Spanish economy. The contribution of the tourism industry to gross domestic product (GDP) and employment is approximately 12% and covers approximately 60% of the Spanish commercial deficit. Because of the importance of economic conditions in a cyclical industry such as the hotel sector, we use a set of macroeconomic factors that are directly related to the industry. Chen (2010) posits that a business expansion (contraction) can strengthen (weaken) corporate earnings and profit, which may in turn improve (worsen) the corporate performance of hotel companies. Specifically, this article examines the relationship between earnings benchmarks and the hotel occupancy rates, the number of arrivals, and the number of visitors staying overnight. To the best of our knowledge, this research is the first to analyze this relationship.

The contribution of this article to the previous literature may be summarized in the following points. This study is motivated by the ongoing debate about earnings quality and the recent studies that focus on economic, social, and cultural elements as possible associated factors (Burgstahler et al., 2006; Leuz, Nanda, & Wysocki, 2003; Monterrey & Sánchez-Segura, 2006). Expanding on the economic factor, which may influence the earnings quality of hotel firms, we use a set of tourist flows that directly affects the industry. That is, when examining the Spanish hotel industry, which is remarkable for its importance, we analyze specific factors that directly affect the business. According to García-Pozo, Campos-Soria, Sánchez-Ollero, and Marchante-Lara (2012), the differential behavior of the labor market in the hospitality industry and the significant structural differences that exist between the primary tourism regions in Spain require that each region be studied individually.

Compared with research on listed or large firms, few studies have addressed the accounting practices of privately held firms. Thomas, Shaw, and Page (2011) are surprised at the limited engagement of scholars in research on small firms in the tourism industry because such companies are commonly referenced in the tourism literature. The Spanish hotel industry is populated with small firms (see DIRCE, 2011). Therefore, this study focuses on a current topic of significant interest, the earnings quality, and the determinants of earnings quality, of nonlisted companies (e.g., Ali, Chen, & Radhakrishman, 2007; Ball & Shivakumar, 2005; Burgstahler et al., 2006; Coppens & Peek, 2005; Jara & López Iturriaga, 2011; Monterrey & Sánchez-Segura, 2006; Wang, 2006).

Several steps have been taken in conducting our analysis. First, as a proxy for earnings benchmarks, this article analyses two critical points—avoiding reporting small losses and avoiding reporting small decreases in earnings—using frequency histograms (Burgstahler & Dichev, 1997). Second, the association between earnings benchmarks and the specific economic hospitality factors driven by all Spanish regions are examined. The analysis is performed at the industry and regional levels by considering the effect of tourist flow and the geographical locations of hotel firms. The methodologies are the Pearson and Spearman correlations, mean differences (the t test for independent samples), median differences (the Wilcoxon rank parametric test), and a logistic model.

The evidence suggests that the environment and economic conditions in which a hotel operates affect the corporate performance of the firm and, therefore, its business strategies. Moreover, the evidence suggests a positive relationship between tourist flow and earnings—but not for all firms. In particular, firms meeting earnings benchmarks where the literature documents that managers have strong incentives to engage in an earnings management strategy show a negative relationship with regional tourist flow.

This article is structured as follows. The second section offers a review of the literature and the hypothesis. The third section describes the sample and the empirical research. The fourth section provides results. The fifth section presents discussions, conclusions, and limitations.

Literature Review

The Topic: Earnings Quality

The Financial Accounting Standards Board (2008) and the International Accounting Standards Board (2008) consider accounting quality information to be relevant and useful for making financial decisions about firms. The primary output of financial reporting entails the use of net income (or earnings) as a measure of performance. However, accounting earnings are subject to a set of estimates and professional judgments that influence how a firm’s profitability is measured for the period in question.

Managers may use the flexibility that is offered in the accounting rules to choose certain accounting treatments in providing their private information. Healy and Wahlen (1999) claim that earnings management occurs when “managers use their judgment in financial reporting to alter financial statements to either mislead certain stakeholders regarding the underlying economic performance of a company or to influence contractual outcomes that depend on reported accounting numbers” (p. 398). Schipper (1989) defines earnings management as “purposeful intervention in the external financial reporting process, with the intent of obtaining some private gain” (p. 92).

As earnings quality is not directly measurable, empirical research uses proxies, such as earnings persistence, accounting accruals, smoothness, timeliness, loss avoidance, and investor responsiveness (see Dechow, Ge, & Schrand, 2010; Francis, LaFond, Olsson, & Schipper, 2004), to gauge earnings quality.¹ The literature shows no consensus regarding the optimal measure of earnings quality. Francis et al. (2004) classify the proxies of earnings quality as accounting-based attributes if they are estimated using accounting data (e.g., earnings persistence, predictability, and smoothness) and market-based attributes if they are estimated through a reliance on both accounting and market data (e.g., value relevance, timeliness, and conservatism).

This article uses earnings benchmarks as a proxy for earnings quality for several reasons. First, the Spanish tourism industry is primarily composed of private companies (only two companies are publicly listed), which constrains the use of market data. Therefore, accounting measures are more adequate. Second, because nonlisted companies reveal less financial information than listed companies, it is more fitting to choose a measure that is available for the entire sample. One of the most popular dimensions examined in the literature is the earnings management hypothesis, which involves smoothing, earnings benchmarks, or discretionary accruals.

According to the survey by Graham et al. (2005): “Several performance benchmarks have been proposed in the literature . . . such as previous years’ or seasonally lagged quarterly earnings, loss avoidance, or analysts’ consensus estimates” (p. 21). Many studies document that managers have strong motivations to research these earnings targets (Bartov et al., 2002; Burgstahler & Chuk, 2012; Burgstahler & Dichev, 1997; Degeorge et al., 1999; Jacob & Jorgensen, 2007; Kerstein & Rai, 2007; Skinner & Sloan, 2002).

In psychology research, Pope and Simonsohn (2011) assert that round numbers are “cognitive reference points.” Rounding numbers in performance scales act as reference points; individuals exert effort to perform barely above rather than barely below such numbers. In the context of loss avoidance, important distinctions exist between reported positive numbers and reported negative numbers. Similar arguments may be made for performance relative to the prior comparable period and relative to analysts’ earnings forecast.

The presumption is that reaching the earnings target influences investors’ expectations of the future earnings and the decision making of a firm. That is, previous studies document that meeting earnings benchmarks is associated with characteristics such as an equity premium in the market, high stock returns, better conditions under which to access financial resources, and better relationships with owners, customers, suppliers, and other actors (see Bartov et al., 2002; Dechow et al., 2010; Graham et al., 2005; Skinner & Sloan, 2002). Additionally, managers have strong incentives to boost earnings in certain situations, such as where managers’ remuneration is based on financial numbers (as reported sales and profits).

For example, regarding executive compensation in the restaurant industry, Barber et al. (2006) suggest a positive relationship between CEO compensation and firm financial performance. Additionally, Guillet et al. (2012) reveal that determinants of equity-based compensation vary by different types of executives, and this compensation is determined by firm performance measures that are in addition to executive-related characteristics, such as tenure. Turner and Guilding (2011) explain 18 distinct earnings management motivations for hotel owners and operators to undertake earnings management through the selective capitalization or expensing of asset-related expenditures. Because hotel management contracts typically reward operators based on a percentage of profit and/or gross revenue, the earnings figure is a key factor to promote the incentive fees (Turner & Guilding, 2010b).

Theoretical Development and Hypothesis

The Spanish hotel industry is dominated by small firms. Only two Spanish companies (Sol Melia and NH Hotels) are listed on an exchange. It is noteworthy that the quality of financial statements for nonlisted companies is a matter of considerable interest in international research (e.g., Ali et al., 2007; Ball & Shivakumar, 2005; Burgstahler et al., 2006; Coppens & Peek, 2005; Jara & López Iturriaga, 2011; Land & Lang, 2002; Wang, 2006).

For example, Land and Lang (2002) compare U.S. cross-listed and non-cross-listed companies and find that the former report less earnings smoothing, more timely loss recognition, and more value relevance than non-cross-listed companies. Using a sample of companies from the United Kingdom, Ball and Shivakumar (2005) note that earnings quality are lower for nonlisted companies than for listed companies, although both types of companies are governed and constrained by identical accounting regulations and audit supervision.

Prior studies posit that contracting and regulatory factors might inspire use of earnings management strategy in private firms (see, e.g., Burgstahler et al., 2006; Coppens & Peek, 2005). Using a sample of 13 European countries, Burgstahler et al. (2006) report that nonlisted companies show lower earnings quality than listed companies. Factors such as different legal systems, relationships between accounting and taxation, and enforcement systems may explain differences between countries. Using a sample of companies from eight European countries, Coppens and Peek (2005) find that tax regulation strongly affects earnings management practices in the absence of capital market pressures.

Agency theory provides an appropriate framework for explaining information asymmetries between managers and shareholders and the different objectives that each group may pursue (Jensen & Meckling, 1976). This information asymmetry creates incentives for corporate managers to engage in dysfunctional behavior that maximizes short-term wealth and private interest at the expense of maximizing long-term value.

Wang (2006) and Ali et al. (2007) find that family firms present higher earnings quality than nonfamily firms and interpret this finding as an indication that family firms generally encounter lower agency costs. More recently, Jara and López Iturriaga (2011) note that earnings management in family-owned firms is reduced when the largest shareholder increased contestability of the control.

With respect to the hotel industry, Turner and Guilding (2010a, 2011) explain that many hotels are governed by a management contract under which hotel owners engage the services of a specialist hotel operator to manage and operate the hotel. The basis of this relationship indicates a potential for conflicting interests between the two contracting parties with respect to the determination of reported profits. Turner and Guilding (2010b, 2011) posit that hotel operators’ incentive fees are typically based on the profits achieved by the hotel that they manage. Therefore, hotel operators are likely to boost reported profit and to increase their fees.

In summary, considerable effort has been devoted to improving the motivations for, and the context that explains differences in accounting quality. Recent academic research suggests that earnings quality is related to accounting standards and to a firm’s investor type, institutional setting, enforcement system, corporate governance, ownership structure, firm characteristics, and economic factors (e.g., Burgstahler et al., 2006; Coppens & Peek, 2005; Jara & López Iturriaga, 2011; Leuz et al., 2003; Monterrey & Sánchez-Segura, 2006).

This study contributes to the recent literature by focusing on the determinants of earnings benchmarks using a sample of hospitality firms. The most intuitive income target are avoiding negative net income (i.e., reaching positive net income) and a decrease in earnings (i.e., reaching an earnings increase) in financial statements (see, e.g., Burgstahler & Dichev, 1997; Degeorge et al., 1999; Graham et al., 2005). Parte and Such (2011), using a sample drawn from the Spanish hotel industry, document a set of accounting variables that determine different behaviors between firms meeting earnings benchmarks and other firms.

Extending these studies, we analyze a set of regional factors that directly affect hotel performance. Haktanir and Harris (2005) and Chen (2010) emphasize the importance of economic conditions in a cyclical industry such as the hotel industry. Haktanir and Harris (2005) contend that the context of a business is one of the primary elements of performance management systems. Performance management systems might vary among firms that operate in different environments and regions. Consequently, this study attempts to explore the relationship between economic and regional factors related to the hospitality industry and meeting earnings targets.

We argue that both the general economic conditions of the country and industry-specific economic conditions might affect earnings and whether earnings benchmarks are met. In fact, García-Pozo et al. (2012) claim that performance management may vary among firms that operate in different environments and regions. Monterrey and Sánchez-Segura (2006) note that regional factors (such as GDP and level of education) are associated with earnings quality (measured as a smoothing earnings ratio).

When examining a specific industry such as the Spanish hospitality industry, we use a set of factors related to the industry—the supply trend and demand trend of tourist flow. We posit that higher profits would be expected with higher measures of demand. However, prior research shows that managers have strong motivations to meet important earnings targets as positive earnings and increases in earnings. Moreover, the literature posits that earnings management is likely to occur when firms barely beat/meet earnings benchmarks. We expect that there will be a positive relationship between the likelihood of reported profits and tourist flow measures but not in all cases (because the earnings distribution shows an inflexion point at approximately zero profits). In this context, we argue that the likelihood of reporting a profit could be negatively related to tourist flow measures.

Three specific variables related to regional tourist flow are examined in this study. For the supply trend, hotel occupancy rates (OCCUP) are examined. In terms of demand trend, the number of arrivals in a region (ARRIV) and the number of visitors staying overnight (NIGHT) are used. As prior research suggests, we consider the location of the hotel and the tourist flow driven by the region as regional economics factor (see, e.g., García-Pozo et al., 2012). We then analyze the relationship between firms that meet earnings benchmarks (Suspect firms) and tourist flow. Thus, we can derive the following alternative hypotheses:

Hypothesis 1: Hotel occupancy rate is negatively associated with earnings benchmarks.

Hypothesis 2: The number of arrivals in a region is negatively associated with earnings benchmarks.

Hypothesis 3: The number of visitors who stay overnight is negatively associated with earnings benchmarks.

Data and Empirical Research

Financial data were collected from the SABI database, which covers a large proportion of Spanish small firms and it is an appropriate database for the study of local economies. The sample selection criteria covered all companies that were classified in Section 551 of the National Classification of Economic Activities (CNAE-2009). During the selection procedure, we attempted to retain a maximum number of observations with the objective of achieving maximum representation of the Spanish hotel industry. The selection procedure resulted in 19,053 observations of Spanish hotels and covers the 2000 to 2008 period. The accounting statement information was completed together with the “Hotel Occupancy Survey” (INE, 2011a) to gather information on tourism trends with respect to supply and demand in the hotel sector.²

First, we examine the hypothesis of avoiding small losses and avoiding small earnings decreases through frequency histograms for all regions in Spain. Using this methodology, Burgstahler and Dichev (1997) reveal a low density of observations in the intervals immediately below the earning of zero profits and a high density of observations in the intervals immediately above the earning of zero profits when compared with what was expected under normal conditions. The earnings distribution shows a discontinuity in terms of near-zero profits and zero changes in results. The evidence might be considered as indicative of earnings management behavior (Burgstahler & Dichev, 1997; Degeorge et al., 1999) because the frequency of firms in the slightly to the right of zero (Suspect firms) is significantly different from the expected frequency.

The frequency histograms represent earnings distribution of net income scaled by assets at the beginning of the year (NI/A_{t − 1}), and the earnings distribution of changes in net income scaled by assets at the beginning of the year (ΔNI/A_{t − 1}) for all regions in Spain. The interval widths are 0.01 for net income and changes in net income.³

Second, the article analyses the association between firms meeting earnings benchmarks (Suspect firms) and specific tourist flow in all regions (hotel occupancy rates, the number of arrivals in each region, and the number of visitors staying overnight). The methodology consists of Pearson and Spearman correlations, mean differences (the p value from the two-sample t test, two-sided), median differences (Wilcoxon rank-sum test, two-sided), and a logistic regression analysis (logit model).

As discussed above, we expect a positive relationship between reporting profits and the measures of demand—but not for all firms. Particularly, firms that present a high probability to engage in earnings management to meet certain earnings benchmarks could not present a positive relationship with tourist flow. Firms meeting earnings benchmarks present a complex history and may be firms that report very small profits and other firms that use accounting discretion to show this level of earnings. An example of the latter might be a firm that uses its accounting discretion to beat the zero point to report a small profit, when it would otherwise report a small loss. We expect a negative association between Suspect firms and hotel occupancy rates (Hypothesis 1), the number of arrivals in a region (Hypothesis 2), and the number of visitors who stay overnight (Hypothesis 3). We argue that regional tourist flow will be negatively associated with firms meeting earnings benchmarks.

In addition, a single variable, COM, is created by combining these three individual variables. The variable COM assumes a value of 1 if a hotel is situated in a region above the national average in terms of tourist flow and a value of zero otherwise. The national average is calculated by the mean of tourist flow in each year. We expect a positive relationship between the regional tourist flow and earnings quality.

Finally, we introduce control variables related to earnings management behavior: Leverage (LEV), Size (SIZE; see, e.g., Healy & Wahlen, 1999), and the Return on Assets of previous year (ROA_1;⁴ see, e.g., Kothari, Leone, & Wasley, 2005; Monterrey & Sánchez-Segura, 2006).

The model is specified as follows:

E a r n i n g s B e n c h m a r k s = f (O C C U P_{k}, A R R I V_{k}, N I G H T_{k}, L E V_{i t}, S I Z E_{i t}, R O A_1_{i t - 1}, Y e a r f i x e d e f f e c t s),

where Earnings Benchmarks is a dummy variable that takes a value of 1 for firms allocated in the first two intervals immediately to zero and a value of 0 otherwise. OCCUP is the hotel occupancy rate in region K in which each company is located; ARRIV is the number of arrivals in region K in which each company is located; NIGHT is the number of visitors with overnight stay in the region K in which each company is located; LEV is calculated as Total Debts/Total Assets; SIZE is the firm’s total assets in year t. We use the logarithm of total assets; and ROA_1 is calculating as Return on assets of firm i in year t − 1.

Results

Descriptive Statistics

Table 1 shows descriptive statistics for the net results for all regions. Hotel companies are profitable during the period from 2000 to 2008 with a mean profitability of 0.029. The results show that 24.8% of the sample reported losses in net earnings. The mean change in earnings is −0.002. The results reveal that 52.8% of the sample reported decreases in net income [(NI_t − NI_t _{− 1})/A_t _{− 1}].

Table 1

Descriptive Statistics for Net Income (NI_t/A_t _{− 1}) and Changes in Net Income (ΔNI_t/A_t _{− 1}) for All Regions in Spain

		Net Income				Change in Net Income
	Total	Mean	Median	Desv.	% Loss	Mean	Median	Desv.	% Decrease
Andalusia	1,960	0.027	0.019	0.092	0.223	−0.001	−0.001	0.084	0.512
Aragón	664	0.034	0.018	0.096	0.253	0.004	0.001	0.101	0.491
Asturias	512	0.015	0.013	0.077	0.264	−0.004	−0.001	0.083	0.539
Balearic Islands	2,464	0.026	0.019	0.085	0.278	−0.003	−0.002	0.085	0.543
Canary Islands	1,176	0.017	0.014	0.101	0.382	−0.005	−0.004	0.097	0.562
Cantabria	96	0.032	0.025	0.051	0.208	−0.001	0.001	0.057	0.458
Castilla-León	944	0.026	0.020	0.080	0.227	−0.001	−0.001	0.084	0.530
Castilla-Mancha	432	0.034	0.022	0.096	0.160	0.001	0.000	0.082	0.493
Catalonia	4,024	0.034	0.024	0.101	0.220	−0.003	−0.001	0.094	0.532
Madrid	928	0.031	0.023	0.124	0.278	−0.004	−0.002	0.114	0.528
Murcia	168	0.005	0.012	0.074	0.292	−0.008	−0.001	0.069	0.548
Valencia	1,368	0.033	0.021	0.088	0.224	−0.002	−0.001	0.090	0.528
Extremadura	288	0.030	0.025	0.083	0.181	−0.003	−0.001	0.079	0.514
Galicia	1,208	0.025	0.018	0.089	0.257	−0.001	−0.001	0.084	0.513
Navarra	320	0.038	0.025	0.102	0.253	−0.001	−0.001	0.104	0.522
Basque Country	304	0.033	0.029	0.105	0.283	−0.008	−0.003	0.089	0.520
Rioja (La)	80	0.035	0.022	0.080	0.188	−0.009	−0.007	0.079	0.600
Total	16,936	0.029	0.020	0.093	0.248	−0.002	−0.001	0.089	0.528

Figure 1 shows the distribution of the sample by region. Catalonia, Andalusia, and the Balearic Island are the most represented regions, constituting 50% of the observations. Moreover, if the observations for Valencia, the Canary Islands, Galicia, Castilla-León, and Madrid are added, then 83% of the entire sample is represented. Holland (2004) notes that the frequency histogram methodology is sensitive to the sample size. To avoid potential bias in the application of the methodology, we combined the regions for which there are fewer observations. This group is called “Remainder of the Regions” and includes the following regions: Aragón, Asturias, Cantabria, Castilla-La Mancha, Extremadura, Pamplona, Murcia, Navarra, and the Basque Country. These regions represent approximately 16.9% of the sample of Spanish hotel companies.

Figure 1

Distribution of the Sample Among All Regions of Spain

Table 2 shows the hotel occupancy rates (OCCUP), the number of arrivals in a region (ARRIV), and the number of visitors staying overnight (NIGHT) by region for the 2001 to 2008 period. The variable COM was constructed using the above variables, and this dummy variable assumes a value of 1 if a hotel is situated in a region above the national average and a value of zero otherwise. Specifically, we observe that the regions that are above average in terms of tourist flows are Andalusia, the Balearic Islands, Catalonia, the Canary Islands, Madrid, and Valencia. Most of these regions occupy significant positions in terms of GDP (see National Statistics Institute, INE).

Table 2

Mean of the Number of Arrivals in Each Region, the Number of Visitors Who Stay Overnight, and the Hotel Occupancy Rates by Region for the 2001 to 2008 Period

	OCCUP	ARRIV	log(ARRIV)	NIGHT	log(NIGHT)	COM
Andalusia	51.74	13,530,140	16.42	40,010,108	17.50	1
Aragón	38.57	2,013,070	14.52	4,240,440	15.26	0
Asturias	35.90	1,289,264	14.07	2,856,741	14.87	0
Balearic Islands	71.14	7,310,610	15.80	49,885,189	17.73	1
Canary Islands	67.76	5,765,972	15.57	41,782,851	17.55	1
Cantabria	41.69	1,017,941	13.83	2,453,044	14.71	0
Castilla-León	36.07	4,028,311	15.21	6,869,823	15.74	0
Castilla-Mancha	32.58	1,969,857	14.49	3,375,521	15.03	0
Catalonia	56.41	12,303,235	16.33	39,438,577	17.49	1
Madrid	51.88	7,239,537	15.80	14,525,469	16.49	1
Murcia	47.55	1,001,070	13.82	2,703,148	14.81	0
Valencia	58.98	6,284,979	15.65	22,869,080	16.95	1
Extremadura	32.37	1,066,860	13.88	1,876,560	14.44	0
Galicia	35.41	3,282,651	15.00	7,204,698	15.79	0
Navarra	38.10	655,287	13.39	1,282,980	14.06	0
Basque Country	47.02	1,791,359	14.40	3,393,541	15.04	0
Rioja (La)	43.27	471,405	13.06	849,286	13.65	0

Note: OCCUP is the hotel occupancy rate, ARRIV is the number of arrivals in each region, NIGHT is the number of visitors staying overnight by region, and COM is a dummy variable that adopts a value of 1 if a hotel is situated in a region above the national average in terms of tourist flow and a value of zero otherwise. When ratios or logarithm transformations are used, economic and accounting variables are comparable between firms (see, e.g., Chen, 2010). According to previous studies, the logarithm of ARRIV and NIGHT are used.

Frequency Histograms

Figure 2 shows the earnings distribution of net income scaled by assets at the beginning of the year (left side) and the earnings distribution of changes in net income scaled by assets at the beginning of the year (right side) for all regions in Spain.⁵ The interval widths are 0.01.

Figure 2

Net Income Distribution (Levels and Changes) Among All Regions in Spain

As Degeorge et al. (1999) explained, frequency histograms show that the following three phenomena that arise when managers misreport earnings: (a) an earnings benchmark creates a gap in earnings distribution barely below zero; (b) for a range of values of latent earnings, a profit barely sufficient to beat/meet the earnings targets is rendered; and (c) the figure earnings will be a sharply discontinuous function of latent earnings.

Turning to Figure 2, the histograms on the left side reveal a discontinuity at the zero point that is statistically significant (p < .05). In contrast, the histograms on the right side show a concentration of observations near the interval (−0.01 and 0.01). The distribution of net income shows a low density of observations in the intervals immediately below zero profits and a high density of observations in the intervals immediately above zero profits. This evidence might indicate earnings management behavior (Burgstahler & Dichev, 1997; Degeorge et al., 1999).⁶ The empirical evidence of the distribution of changes in net income reveals an accumulation of observations in the intervals immediately to the left and right of zero.

The literature considers that factors related to the deflation used to normalized earnings might remap observations toward the zero point, the intervals widths might affect the earnings kink, also the sample used (including the selection process, number of observations, and premanaged distribution of earnings) should be taken into account (see, e.g., Dechow et al., 2003; Degeorge et al., 1999; Durtschi & Easton, 2005, 2009; Holland, 2004). In the section that follows, we control for several factors that may affect the earnings distribution.

Figure 3 presents the earnings distribution of net income and classifies the regions by tourist flow (the number of arrivals in each region, the number of visitors who stay overnight, and the hotel occupancy rates). The first group contains regions where tourist flow is above the mean (Andalusia, the Balearic Islands, Catalonia, the Canary Islands, Madrid, and Valencia), and the second group contains regions where tourist flow is below the mean.

Figure 3

Net Income Distribution (NI_t/A_t _{− 1}) Driven by Regions With High Tourist Flow and Low Tourist Flows

Both frequency histograms show a break at the zero point that is statistically significant (p < .05). Panels A and B present the earnings distribution of firms that are located in regions where tourist flow is above and below the mean, respectively. The break at the zero point is slightly higher for firms that are below the mean than it is for firms that are above the mean. This suggests that a relationship exists between meeting earnings benchmarks and regional tourist flow.

In the next section, the effect of regional tourist flow on Suspect firms is examined by using the following three variables: hotel occupancy rates, the number of arrivals in a region, and the number of visitors staying overnight.

Association Between Earnings Benchmarks and Tourist Flow

Table 3 shows the Pearson coefficients between positions on the earnings distribution. Panel A provides the correlation between Earnings (excluding the two intervals nearest to the zero point to avoid confusing effects) and regional tourist flow. Panel B represents the correlation between Suspect firms and regional tourist flow using the right side of the earnings distribution. Panel C shows the correlation between Suspect firms and regional tourist flow using the two intervals immediately to the right and left of zero.

Table 3

Pearson Correlations Between Different Partitions of Earnings Distribution and Tourist Flow

	Panel A	Panel B	Panel C
OCCUP	0.032***	−0.024***	−0.064***
OCCUP	0.000	0.007	0.000
ARRIV	0.007**	−0.013**	−0.029*
ARRIV	0.044	0.015	0.097
NIGHT	0.024***	−0.002***	−0.018**
NIGHT	0.005	0.004	0.030
COM	0.017**	−0.011**	−0.012**
COM	0.034	0.023	0.031

Note: Panel A provides the correlation between Earnings (excluding the two intervals nearest the zero point to avoid confusing effects) and regional tourist flow. Panel B represents the correlation between Suspect firms and regional tourist flow, using the right side of the earnings distribution. Panel C shows the correlation between Suspect firms and regional tourist flow, using the two intervals immediately to right and left to zero.

,**, ***Represent significance at the level of 1%, 5%, and 1%, respectively.

As expected, Panel A shows that the Profit coefficient is positively associated with OCCUP (p < .01), ARRIV (p < .05), NIGHT (p < .01), and COM (p < .05). The evidence suggests that firms that have high earnings are associated with higher tourist flow and also with regions that have higher tourist flow. Panels B and C show that, in all cases, Suspect firms coefficients are negatively associated with OCCUP, ARRIV, NIGHT, and COM and the coefficients are statically significant. That is, Suspect firms present a negative association with the tourist flow measures using the right side of the histogram and the two first intervals immediately to the right and left of zero, respectively.

Furthermore, Table 4 indicates a positive and significant association among Suspect firms and LEV (p < .01 Spearman and Pearson) and SIZE (p < .01 Spearman and Pearson). That is, firms that meet earnings benchmarks are larger firms with a high leverage ratio. In contrast, a negative association is found between Suspect and ROA_1 of the previous year (p < .01, Spearman and Pearson). See Appendix A for definitions of the variables.

Table 4

Pearson and Spearman Correlations Between Firms Meeting Earnings Benchmarks and Control Factors

	Suspect	LEV	SIZE	ROA_1
Suspect	—	0.111***	0.039***	−0.171***
Suspect		0.000	0.000	0.000
LEV	0.082***	—	−0.131***	−0.339***
LEV	0.000		0.000	0.000
SIZE	0.045***	−0.160***	—	−0.002
SIZE	0.000	0.000		0.976
ROA_1	−0.096***	−0.379***	−0.019**	—
ROA_1	0.000	0.000	0.018

Note: The Pearson coefficients are below the diagonal and the Spearman coefficients are above the diagonal. See Appendix A for definitions of the variables.

,**, ***Represent significance at the level of 10%, 5%, and 1%, respectively.

Table 5 reports the mean differences (the p value from the two-sample t test, two-sided) and the median differences (Wilcoxon rank-sum test, two-sided) between Suspect firms and remaining firms. When examining hotel occupancy rates (OCCUP), we observe that the mean and median differences are statistically significant (p < .01). The results show statically differences in medians looking at the number of arrivals (p < .01). When examining the number of visitors staying overnight (NIGHT), we find marginal differences in means (p < .10). Finally, the mean and median differences are statistically significant for LEV, SIZE, and ROA_1 (p < .01); the variables related to tourist flow and corporate performance for Suspect firms are different (as mean or median) when compared with nonsuspect firms.

Table 5

Test of Mean Differences (t Test) and Median Differences (Wilcoxon Rank-Sum Test)

	OCCUP	ARRIV	NIGHT	LEV	SIZE	ROA_1
Suspect firms	0.522	15.608	16.748	0.629	7.646	0.084
Otherwise	0.532	15.582	16.760	0.550	7.437	0.086
t Mean Pr(T > t)	0.000***	0.149	0.090*	0.000***	0.000***	0.000***
Wilcoxon Pr(T > t)	0.000***	0.009***	0.787	0.000***	0.000***	0.000***

Note: This table tests the mean and median differences between Suspect firms (earnings in the first two intervals immediately to the right of zero) and nonsuspect (earnings outside of this category, referred to as “Otherwise”).

,**, ***Represent significance at the level of 10%, 5%, and 1%, respectively.

Table 6 shows the results of the logistic regression analysis (logit model) using different segmentations. Panel A shows the logit model where the dependent variable adopts a value of 1 if firms report profits and zero if firms present losses (we exclude the first two intervals nearest zero to avoid confusing effects). As expected, the probability of reporting a profit is positively associated with OCCUP (p < .01), ARRIV (p < .10), NIGHT (p < .01), and COM (p < .05). The higher measures of demand indicate higher profits. Panel A also indicates that the probability of reporting a profit is positively associated with ROA_1 of the previous year (p < .01) and negatively associated with LEV (p < .01) and SIZE (p < .01).

Table 6

Results of the Logit Model

		OCCUP		ARRIV		NIGHT		COM
	Coef exp	Coefficient	P > \|z\|	Coefficient	P > \|z\|	Coefficient	P > \|z\|	Coefficient	P > \|z\|
Panel A: Profits firms versus losses firms (excluding intervals near to zero)
const		1.765***	0.000	1.236***	0.008	2.342***	0.000	1.381***	0.000
Tourist flow	+	1.157***	0.000	0.008*	0.090	0.063***	0.006	0.124**	0.020
LEV	?	−1.572***	0.000	−1.561***	0.000	−1.558***	0.000	−1.553***	0.000
SIZE	?	−0.017	0.268	−0.044***	0.004	−0.034**	0.026	−0.036**	0.019
ROA_1	+	8.680***	0.000	8.732***	0.000	8.728***	0.000	8.738***	0.000
Fixed effects		Yes		Yes		Yes		Yes
LR χ²(11)		2579.45		2547.46		2555.16		2552.83
Prob > χ²		0.000		0.000		0.000		0.000
Pseudo R²		0.29		0.29		0.29		0.29
Correctly classified		79.29%		79.55%		79.43%		79.38%
Panel B: Suspect firms versus non-suspect firms
const		−2.601***	0.000	−2.848***	0.000	−2.522***	0.000	−2.903***	0.000
Tourist flow	−	−0.948***	0.000	−0.011	0.708	−0.045**	0.045	−0.156***	0.003
LEV	+	0.988***	0.000	0.992***	0.000	0.994***	0.000	1.002***	0.000
SIZE	?	0.121***	0.000	0.100***	0.000	0.106***	0.000	0.110***	0.000
ROA_1	−	−1.562***	0.000	−1.522***	0.000	−1.530***	0.000	−1.525***	0.000
Fixed effects		Yes		Yes		Yes		Yes
LR χ²(11)		338.63		317.51		321.35		326.07
Prob > χ²		0.000		0.000		0.000		0.000
Pseudo R²		0.1		0.1		0.1		0.1
Correctly classified		85.57%		85.57%		85.57%		85.57%
Panel C: Suspect Firms versus healthy firms (right side of the histogram)
const		−2.466***	0.000	−2.586***	0.000	−2.304***	0.000	−3.008***	0.000
Tourist flow	−	−0.602***	0.006	−0.005**	0.026	−0.023**	0.033	−0.113***	0.005
LEV	+	1.521***	0.000	1.522***	0.000	1.524***	0.000	1.531***	0.000
SIZE	?	0.154***	0.000	0.140***	0.000	0.143***	0.000	0.147***	0.000
ROA_1	−	−5.742***	0.000	−5.723***	0.000	−5.723***	0.000	−5.715***	0.000
Fixed effects		Yes		Yes		Yes		Yes
LR χ²(11)		1029.2		1021.7		1022.6		1025.7
Prob > χ²		0.000		0.000		0.000		0.000
Pseudo R²		0.15		0.15		0.15		0.15
Correctly classified		79.64%		79.64%		79.64%		79.71%
Panel D: Suspect firms versus firm with small losses (firms near to zero)
const		2.098***	0.000	0.468	0.575	1.893***	0.006	1.730***	0.000
Tourist flow	−	−0.862**	0.019	−0.090*	0.096	−0.010*	0.080	−0.008***	0.004
LEV	+	0.634***	0.000	0.638***	0.000	0.651***	0.000	0.650***	0.000
SIZE	?	0.150***	0.000	0.177***	0.000	0.168***	0.000	0.170***	0.000
ROA_1	+	2.950***	0.000	2.950***	0.000	2.955***	0.000	2.954***	0.000
Fixed effects		Yes		Yes		Yes		Yes
LR χ²(11)		81.94		79.15		76.47		76.41
Prob > χ²		0.000		0.000		0.000		0.000
Pseudo R²		0.1		0.1		0.1		0.1
Correctly classified		73.81%		73.84%		73.78%		73.78%

Note: See Appendix A for definitions of the variables.

,**, ***Represent significance at the level of 10%, 5%, and 1%, respectively.

Panel B shows the results of the logistic regression analysis between Suspect firms and non–Suspect firms. Specifically, the dependent variable adopts a value of 1 if the firms are within the first two intervals immediately to the right of zero (Suspect firms) and zero otherwise. Panel B indicates that the probability of meeting earnings benchmarks (that represent Suspect firms) is negatively associated with OCCUP (p < .01), NIGHT (p < .05), and COM (p < .01). In contrast, the coefficient that is associated with ARRIV is not statistically significant. Additionally, the probability of meeting earnings benchmarks is negatively associated with ROA_1 of the previous year (p < .01) and positively associated with LEV (p < .01) and SIZE (p < .01). These results are consistent with the univariate analysis.

Further analysis is required of the right side of the earnings distribution and Suspect firms (Panel C, Table 6), in addition to the left hand side of the earnings distribution (Panel D, Table 6). Panel C shows that the probability of meeting the earnings benchmark (Suspect firms) is negatively associated with OCCUP (p < .01), NIGHT (p < .05), ARRIV (p < .05), and COM (p < .01). The statistical tests suggest that a negative association exists between earnings benchmarks and tourist flow. Moreover, the negative relationship suggests that the positive relationship between earnings and tourist flow shows an inflexion point around Suspect firms (the intervals immediately to the right zero).

Finally, Panel D shows the results of Suspect firms and firms reporting small losses (the two intervals immediately to the right and left of zero). The dependent variable takes a value of 1 if firms are within the first two intervals immediately to the right of zero (Suspect firms) and a zero for firms suffering small losses. Panel D documents that the probability of meeting the earnings benchmarks (firms that report very small profits and other firms that use accounting discretion to show this figure of profits) is negatively associated with OCCUP (p < .05), ARRIV (p < .10), NIGHT (p < .10), and COM (p < .01).

In sum, the statistical tests suggest that the likelihood of reporting small profits are negatively associated with regional tourist flows. One plausible explanation for this correlation may result from the different characteristics of firms that meet earnings benchmarks. The earnings interval that represents firms meeting earnings benchmarks is composed of firms that obtain small profits and firms that are further from profits but who may use accounting discretion to meet the earnings benchmarks; these firms may face negative correlation with regional tourist flow.

Sensitivity Analysis

The frequency histogram methodology is not exempt from criticism and bias. Dechow, Richardson, and Tuna (2003); Holland (2004); and Durtschi and Easton (2005, 2009), among others, believe that the discontinuities in earnings distributions are driven by factors that include deflation, sample selection, and differences between what characterizes profit and loss in different firms. However, authors such as Jacob and Jorgensen (2007), Kerstein and Rai (2007), and Burgstahler and Chuk (2012) reexamine the evidence, consider the arguments provided by previous works (e.g., Durtschi & Easton, 2005, 2009; Holland, 2004), and demonstrate that earnings management remains the only plausible explanation for the extensive body of evidence that reveals discontinuities in earnings distributions (Burgstahler & Chuk, 2012).

This article presents several controls for various factors in the frequency histogram (e.g., the earnings deflator and interval widths) in an examination of the competing arguments. In this section, we construct the histograms under different assumptions.

Durtschi and Easton (2005, 2009) claim that the distribution of net income deflated by market capitalization prices (extensively to other denominators) differs according to the magnitude and sign of the earnings. That is, deflator distorts the underlying distribution of net income. These authors demonstrate that the price per share of loss firms systematically differs from the price per share for firms that report profits of the identical magnitude and provide evidence, therefore, that deflation will affect the shape of the earnings distribution at approximately zero.

To ensure that our deflator does not move observations between zones in the frequency histograms, we use the mean differences for independent samples and the nonparametric Wilcoxon rank (Durtschi & Easton, 2005, 2009). Appendices B and C contain a sensitivity analysis of the deflation. The final columns of Appendices B and C show that the differences in mean and median are generally not significant. To ensure consistent results, frequency histograms are used to represent earnings without a deflator and to use alternative intervals.⁷ The evidence obtained is similar to that reported in this article.

Discussions, Conclusion, and Limitations

This study examines the relationship between meeting earnings benchmarks and regional factors related to the hospitality industry. The tourism industry was chosen because of its relevance to Spain’s economy. In 2010, the tourism sector was responsible for 10.4% of Spanish GDP. Within the tourism industry, the hotel subsector has significant weight, representing nearly 27% of total production. The Annual Survey of Services (INE, 2011b) in Spain has shown that this subsector contributes 33% of the total staff costs and 22% of the personnel employed in the “tourism” industry. Thus, this situation underlines the need for quality financial reports of hotels and highlights the necessity for improvements in tourism earnings quality.

According to DIRCE (2011), small firms dominate the Spanish service industry. Thomas et al. (2011) argue that small businesses are an important part of the international tourism system yet remain relatively underresearched. Although the determinants of earnings quality have been extensively researched for publicly listed companies, the small business segment is only now gradually beginning to be examined. This article contributes to the discussion on earnings quality with respect to nonlisted companies and identifies specific factors related to the hotel industry that are likely to explain cross-regional differences in earnings quality.

With respect to the determinants of earnings quality, recent empirical research notes that factors such as economic growth, the investor protection, the degree of external economic openness, institutional settings, enforcement systems, and socioeconomic and cultural factors moderate earnings quality (see Burgstahler et al., 2006; Coppens & Peek, 2005; Jara & López Iturriaga, 2011; Leuz et al., 2003; Monterrey & Sánchez-Segura, 2006). Extending this evidence, this article analyses a set of macro-economic factors specific to the tourism industry and relevant for hotels. This question has not been addressed in previous articles.

This study draws on the determinants of earnings benchmarks and identifies regional economic factors that might be associated with earnings. Specifically, the following three drivers that directly affect the corporate performance of hotel firms are examined: the hotel occupancy rates, the number of arrivals in a region, and the number of visitors staying overnight. The analysis is performed at the industry and regional levels to consider the effect of tourism flow and the geographical location of hotel firms.

Our hypothesis assumes that the supply and demand trends of tourist flow in which a hotel is geographically situated influences the hotel’s performance, but there are differences between firms with strong management incentives to alter the earnings figures related to other firms. The evidence suggests a positive relationship between earnings of non–suspect firms and regional tourist flow but a negative association for Suspect firms (firms meeting earnings benchmarks).

Although earnings targets are one of most analyzed topics in the literature, there remain limitations in the current article primarily because many different incentives may influence a manager’s decision to engage in earnings management, such as compensation contracts or tax regulation. The “loss avoidance” strategy appears to be a portion of a more complex story in which firms close to meeting one earnings benchmark may have used their discretion to meet different earnings benchmark goals. Another caveat is that the approach adopted by Burgstahler and Dichev (1997) assumes that the distribution of premanaged earnings should be smooth (discussed in more detail below). Thus, meeting or beating a target is a censored measure of earnings management.

Future studies might focus on an international sample of the service industry in general or on the hotel industry in particular. Because of the importance of economic conditions in an industry such as the service industry, it is noteworthy to know the firm’s strategies to meet earnings targets. The adverse conditions driven by the global financial crisis allow us to test the abilities of the firm, managers, and owners to develop strategies to respond to the external shock and therefore survive.

Additionally, the literature to date identifies two main strategies to manage reported earnings: accrual-based earnings management (that do not affect cash flows) and real actions that affect cash flows. The strategy followed by the service industry and the effect of the global crisis must be answered. Finally, it would be helpful to improve the methodology to detect earnings management.

Footnotes

Appendix

Appendix C

Control of the Deflator (Asset at the Beginning of the Year). Changes in Net Income Distribution (ΔNI_t/A_t − 1)

ΔNI/A_t _{− 1}	Small Decreases				Small Increases				Test Differences
Intervals	N	Mean	Median	SD	N	Mean	Median	SD	Mean	Median
Madrid
Interval 1	125	10,964	2,186	17,286	125	10,094	2,659	15,987	0.680	0.981
Interval 2	59	8,150	2,302	13,892	65	9,645	2,599	16,115	0.583	0.698
Interval 3	58	8,005	1,596	13,456	31	2,896	1,459	4,925	0.012	0.538
Interval 4	39	5,293	1,809	12,271	24	3,177	1,574	6,688	0.441	0.692
Interval 5	24	1,863	1,057	1,738	22	8,062	1,804	12,825	0.035	0.116
Rest	185	5,663	863	16,344	171	4,790	722	15,788	0.449	0.522
Total	490	7,112	1,487	13,723	438	6,690	1,534	13,159	0.634	0.891
Catalonia
Interval 1	670	6,297	2,404	10,485	684	6,813	2,557	10,560	0.368	0.297
Interval 2	351	7,363	2,741	11,826	313	6,341	2,318	10,530	0.243	0.220
Interval 3	253	4,804	2,158	8,109	203	5,646	2,114	9,955	0.331	0.933
Interval 4	171	5,741	2,321	9,904	119	4,542	2,094	7,732	0.269	0.782
Interval 5	126	4,108	1,561	7,232	80	5,327	1,727	10,426	0.323	0.603
Rest	570	3,968	1,049	9,080	484	3,683	829	8,525	0.602	0.066
Total	2141	5,502	1,851	9,966	1883	5,598	1,804	9,905	0.761	0.774
The Balearic Island
Interval 1	381	7,605	3,543	11,925	371	7,144	3,804	10,602	0.576	0.719
Interval 2	267	6,682	3,732	10,125	190	7,301	3,857	11,426	0.542	0.614
Interval 3	166	7,014	3,219	10,568	145	4,953	2,736	7,480	0.046	0.302
Interval 4	105	6,617	3,087	10,420	107	7,453	3,189	12,289	0.594	0.906
Interval 5	91	5,592	2,507	10,220	55	7,372	3,193	13,276	0.365	0.259
Rest	328	4,621	1,558	9,568	258	5,267	2,161	9,922	0.425	0.198
Total	1338	6,401	2,959	10,666	1126	6,499	3,303	10,596	0.821	0.217
Castilla-León
Interval 1	163	2,390	1,338	3,052	141	2,795	1,250	4,000	0.329	0.598
Interval 2	94	2,365	1,191	3,155	70	2,037	1,070	3,871	0.552	0.231
Interval 3	49	2,253	1,310	3,043	52	1,980	1,199	2,066	0.597	0.742
Interval 4	33	2,396	1,238	2,542	34	2,452	1,054	4,491	0.950	0.307
Interval 5	31	2,389	972	4,555	29	2,004	975	2,558	0.691	0.923
Rest	130	1,332	568	1,722	118	1,118	606	1,490	0.300	0.634
Total	500	2,097	1,156	2,899	444	2,056	966	3,287	0.840	0.136
Valencia
Interval 1	213	3,322	1,047	6,441	226	3,211	1,008	6,771	0.860	0.621
Interval 2	116	4,637	1,557	6,895	109	4,403	1,334	8,265	0.818	0.348
Interval 3	92	4,435	1,379	9,312	64	5,200	2,906	7,713	0.589	0.117
Interval 4	49	3,161	1,063	5,155	43	5,483	2,150	8,903	0.124	0.114
Interval 5	57	3,729	1,330	7,454	35	2,293	990	3,276	0.285	0.197
Rest	195	2,181	583	3,908	169	2,020	550	3,718	0.689	0.473
Total	722	3,388	1,093	6,459	646	3,399	1,020	6,628	0.975	0.487
Remainder
Interval 1	470	2,988	1,073	5,562	465	3,488	1,086	7,436	0.244	0.691
Interval 2	234	3,729	1,248	3,729	231	2,890	1,197	2,890	0.164	0.433
Interval 3	169	2,845	1,184	4,988	165	3,178	1,185	5,568	0.565	0.876
Interval 4	106	2,404	1,237	3,729	96	2,285	1,038	4,728	0.843	0.520
Interval 5	89	2,011	880	2,933	74	2,061	877	3,440	0.920	0.984
Rest	404	404	2,224	556	361	1,680	537	3,238	0.090	0.214
Total	1,472	2,779	1,000	5,682	1,392	2,724	969	5,606	0.797	0.565

Notes

Laura Parte-Esteban, PhD (e-mail: lparte@cee.uned.es), is an associate professor in the Department of Accounting and Finance Faculty of Economics and Management, Spanish National University for Distance Education, Madrid, Spain. Pilar Alberca-Oliver, PhD (e-mail: palberca@cee.uned.es) is an associate professor in the Department of Accounting and Finance, Faculty of Economics and Management, Spanish National University for Distance Education, Madrid, Spain.

References

Ali

Chen

T. Y.

Radhakrishnan

. (2007). Corporate disclosures by family firms. Journal of Accounting and Economics, 44, 238-286.

Ball

Shivakumar

. (2005). Earnings quality in UK private firms: Comparative loss recognition timeliness. Journal of Accounting and Economics, 39, 83-128.

Barber

Ghiselli

Deale

. (2006). Assessing the relationship of CEO compensation and company financial performance in the restaurant segment of the industry. Journal of Food Service Business Research, 9(4), 65-82.

Bartov

Givoly

Hayn

. (2002). The rewards to meeting or beating earnings expectations. Journal of Accounting and Economics, 33, 173-204.

Burgstahler

Chuk

. (2012). Correcting disinformation about discontinuity evidence of earnings management (Working paper). Retrieved from http://ssrn.com/abstract=1866008

Burgstahler

Dichev

. (1997). Earnings management to avoid earnings decreases and losses. Journal of Accounting and Economics, 24, 99-126.

Burgstahler

Hail

Leuz

. (2006). The importance of reporting incentives: Earnings management in European private and public firms. Accounting Review, 81, 983-1017.

Chen

M. H

. (2010). The economy, tourism growth and corporate performance in the Taiwanese hotel industry. Tourism Management, 31, 665-675.

Coppens

Peek

. (2005). An analysis of earnings management by European private firms. Journal of International Accounting, Auditing and Taxation, 14(1), 1-17.

10.

Dechow

Schrand

. (2010). Understanding earnings quality: A review of the proxies, their determinants and their consequences. Journal of Accounting and Economics, 50, 344-401.

11.

Dechow

P. M.

Richardson

S. A.

Tuna

I. A

. (2003). Why are earnings kinky? An examination of the earnings management explanation. Review of Accounting Studies, 8, 355-384.

12.

Degeorge

Patel

Zeckhauser

. (1999). Earnings management to exceed thresholds. Journal of Business, 72(1), 1-33.

13.

DIRCE. (2011). Directorio Central de Empresas. Instituto Nacional de Estadística (INE). Retrieved from http://www.ine.es/jaxi/menu.do?type=pcaxis&path=/t37/p201/&file=inebase

14.

Durtschi

Easton

. (2005). Earnings management? The shapes of the frequency distributions of earnings metrics are not evidence ipso facto. Journal of Accounting Research, 43, 557-592.

15.

Durtschi

Easton

. (2009). Earnings management? Erroneous inferences based on earnings frequency distributions. Journal of Accounting Research, 47, 1249-1281.

16.

Financial Accounting Standards Board. (2008). Exposure draft, conceptual framework for financial reporting: The objective of financial reporting and qualitative characteristics and constraints of decision-useful financial reporting information, financial accounting (Reference No. 1570-100). Norwalk, CT: Illinois CPA Society.

17.

Francis

LaFond

Olsson

Schipper

. (2004). Costs of equity and earnings attributes. Accounting Review, 79, 967-1010.

18.

García-Pozo

Campos-Soria

J. A.

Sánchez-Ollero

J. L.

Marchante-Lara

. (2012). The regional wage gap in the Spanish hospitality sector based on a gender perspective. International Journal of Hospitality Management, 31, 266-275.

19.

Graham

Harvey

Rajgopal

. (2005). The economic implications of corporate financial reporting. Journal of Accounting and Economics, 40, 3-73.

20.

Guillet

B. D.

Kucukusta

Xiao

. (2012). An examination of executive compensation in the restaurant industry. International Journal of Hospitality Management, 31, 86-95.

21.

Haktanir

Harris

. (2005). Performance measurement practice in an independent hotel context. International Journal of Contemporary Hospitality Management, 17(1), 39-50.

22.

Healy

Wahlen

J. M

. (1999). A review of the earnings management literature and its implications for standard setting. Accounting Horizons, 13, 365-383.

23.

Holland

. (2004). Earnings management: A methodological review of the distribution of reported earnings approach (Working Paper). Melbourne, Australia: Monash University.

24.

INE. (2011a). Hotel Occupancy Survey. Retrieved from www.ine.es

25.

INE. (2011b). Annual Survey of Services (NSI-ASS, various years). Retrieved from www.ine.es

26.

International Accounting Standards Board. (2008). Exposure draft of an improved conceptual framework for financial reporting. London, England: Committee of European Banking Supervisors.

27.

Jacob

Jorgensen

B. N

. (2007). Earnings management and accounting income aggregation. Journal of Accounting and Economics, 43, 369-390.

28.

Jara

M. L

López Iturriaga

F. J

. (2011). Earnings management and contest to the control: An analysis of European family firms. Journal of Centrum Cathedra, 4(1), 100-120.

29.

Jensen

Meckling

W. H

. (1976). Theory of the firm: Managerial behaviour, agency cost and ownership structure. Journal of Financial Economics, 3, 305-360.

30.

Kerstein

Rai

. (2007). Intra-year shifts in the earnings distribution and their implications for earnings management. Journal of Accounting Economics, 44, 399-419.

31.

Kim

. (2005). A preliminary examination of determinants of CEO cash compensation in the U.S. restaurant industry from an agency theory perspective. Journal of Hospitality & Tourism Research, 29, 341-355.

32.

Kothari

S. P.

Leone

Wasley

. (2005). Performance matched discretionary accruals. Journal of Accounting Economics, 39, 163-197.

33.

Land

Lang

. (2002). Empirical evidence on the evolution of international earnings. Accounting Review, 77(Suppl.), 115-133.

34.

Leuz

Nanda

Wysocki

. (2003). Earnings management an investor protection: An international comparison. Journal of Financial Economics, 69, 505-527.

35.

Monterrey

Sánchez-Segura

. (2006). Las características socioeconómicas como incentivos para la información financiera: evidencia empírica española [The socioeconomic characteristics as incentives for financial reporting: Spanish empirical evidence]. Investigaciones Económicas [Economic Research], XXX(3), 611-634.

36.

Parte

Such

M. J

. (2011). Determinants of earnings benchmarks in the Spanish hotel industry. Tourism Economics, 17(1), 53-72.

37.

Pope

Simonsohn

. (2011). Round numbers as goals: Evidence from baseball, SAT takers, and the lab. Psychological Science, 22(1), 71-79.

38.

Schipper

. (1989). Commentary on earnings management. Accounting Horizons, 3(4), 91-102.

39.

Skinner

Sloan

. (2002). Earnings surprises, growth expectations, and stock returns or don’t let an earnings torpedo sink your portfolio. Review of Accounting Studies, 7, 289-312.

40.

Thomas

Shaw

Page

S. J

. (2011). Understanding small firms in tourism: A perspective on research trends and challenges. Tourism Management, 32, 963-976.

41.

Turner

M. J.

Guilding

. (2010a). Accounting for the furniture, fittings & equipment reserve in hotels. Accounting and Finance, 50, 967-992.

42.

Turner

M. J.

Guilding

. (2010b). Hotel management contracts and deficiencies in owner-operator capital expenditure goal congruency. Journal of Hospitality & Tourism Research, 34, 478-511.

43.

Turner

M. J.

Guilding

. (2011). An investigation of the motivation of hotel owners and operators to engage in earnings management. Qualitative Research in Accounting & Management, 8, 358-381.

44.

Wang

. (2006). Founding family ownership and earnings quality. Journal of Accounting Research, 44, 619-655.