The Nexus of Research and Development Intensity with Earnings Management: Empirical Insights from Jordan

Abstract

Driven by positive accounting, agency, and political and economic theories, this study examines the relationship between research and development (R&D) intensity and earnings management for listed pharmaceutical companies in the Amman Stock Exchange (ASE) between 2008 and 2021. Employing panel regression methods, the results reveal a positive association between R&D investment and earnings manipulation. Specifically, after two or three R&D delays, the association survived. Moreover, firm size negatively affects earnings management, showing that larger firms have less tendencies to conduct earning manipulation. Furthermore, financial leverage and earnings management are strongly connected, showing that firms may utilize earnings management to avoid credit covenants. The findings emphasize distortions in R&D reporting and profit management within Jordan’s financial reporting practices. Enhancing the accuracy of R&D investment disclosures, minimizing profit manipulation, and fostering greater transparency are crucial. Jordan’s regulators should improve capitalization standards, transparency, auditing, and shareholder activism.

1. Introduction

Investing in research and development (R&D) significantly impacts a country’s future prosperity and prominence. It drives economic advancement, technological innovation, national security, and global competitiveness. In today’s rapidly expanding global economy, businesses must prioritize R&D to stay ahead of the competition and enhance market shares through potential technological innovation.

R&D is crucial for national security, technological innovation, economic growth, and international competitiveness, which in turn contribute to a nation’s worldwide position (Dai et al., 2022; Li et al., 2021; Rodrigues et al., 2019; Veselinovic & Veselinovic, 2019). It stimulates innovation, increases profitability, and improves consumer well-being (Freihat & Kanakriyah, 2017; Kruglov & Shaw, 2024). R&D investment unlocks economic success and ensures a global competitive edge.

This study investigates the impact of R&D investment on firms, focusing on whether there is a manipulation of the relationship between R&D and earnings management indicators, as it plays a crucial role in a country’s future economic success and global market standing.

Accounting methodologies concerning R&D and other intangible assets have been thoroughly examined in the extant accounting literature (Zhu et al., 2022). Prior research has indicated that immediate R&D expenditures enhance reliability and diminish earnings management. Other researchers believe that it is more valuable for investors to capitalize on R&D expenditures (Nelson et al., 2003). However, it is equally imperative to establish the relevance of financial reporting. Regarding the deception of financial information, particular earnings management techniques are viewed as lawful, whereas others have been proven to be unlawful. Recent research indicates that conservatism can discourage R&D manipulation, especially in companies that lack strong internal controls and tax enforcement systems (Shen & Ruan, 2022). In the context of financial reporting relevancy, some researchers indicate that R& D will improve the financial statements relevancy in later years (Aboody & Lev, 2000; Dahmash et al., 2009; Healy et al., 2002; Holthausen & Watts, 2001; Lev & Sougiannis, 1996; Wyatt, 2008).

A fascinating correlation exists between R&D spending and earnings management, with managing profitability closely tied to the cost of R&D investment. Since the returns of R&D have become evident, there is a likelihood of firms cutting back on such expenditures. Analysts forecast to predict the outcomes of R&D investment and earnings based on bonuses which determine managerial performance.

Earnings management is defined as the intentionally interfering of financial statements to manipulate financial outcomes. It is conducted to achieve a particular objective or to portray an attractive financial image to investors. This can be accomplished through various methods, including modifying accounting processes, transaction timing, and changing estimates or assumptions.

The cost of R&D expenditures and earnings management has an intriguing relationship. Companies may use earnings management to tame their profits, alleviating criticism from analysts or investors who could interpret earnings volatility as a sign of instability (Baik et al., 2022). In contrast, investing in R&D is generally a long-term strategy that does not yield an immediate return. The outcomes of R&D are visible after two or three years (Lome et al., 2016). As a result, management may feel pressured to cut R&D expenses to boost short-term profitability (Bushee, 1998). In order to increase the firm’s short-term profitability, it is crucial to justify the investments in research activities that may reduce R&D expenditure (Bhojraj et al., 2009). To earn bonuses (Harter & Harikumar, 2004), R&D outcomes determine the assessment of managerial performance.

Previous studies have shown that it is challenging for analysts and investors to evaluate the returns from R&D investments since the terms and returns of R&D spending are unclear (Yang et al., 2022; Guidara & Boujelbene, 2014). R&D investments are therefore considered to be high-risk. Managers in charge of choosing R&D investments will be rewarded for managing earnings that show success or for adopting a big bath plan in the case of failure (Grabinska & Grabinski, 2017).

R&D investments are crucial in defining innovation outcomes and economic advancement, especially in rising economies such as Jordan. Bereskin et al. (2018) emphasize the importance of R&D in promoting the generation of patents and improving the overall effectiveness of innovation. Nevertheless, the utilization of earnings management techniques is a significant obstacle that can unintentionally result in decreased investments in R&D. Company managers often prefer achieving profitability goals over focusing and spending on innovation as confirmed by a study by Marei et al. (2023) in the United Kingdom. Consequently, spending on research, development, and innovation may be considered one of the challenges facing emerging markets such as Jordan. There is a need in Jordan for spending and supporting R&D efforts because this impacts public health outcomes. The ability of research companies to access global distribution channels and supply chains is a key factor in determining the success and introduction of research projects to local and global markets. Investment in research, development, and innovation in pharmaceutical companies is of utmost importance as it leads to the innovation of new drugs or new manufacturing technologies. It also helps combat diseases, reduces production costs, and enhances product quality.

This study is primarily motivated by examining the critical role that R&D plays in fostering innovation and economic advancement, particularly in developing countries like Jordan, where industries often face resource limitations and market-specific challenges. While the pharmaceutical sector in Jordan is highly concentrated, with six companies comprising the entire industry, it contributes approximately 4% to the total GDP in 2023 (1.5 billion JOD). At the same time, the government’s strategy aims to increase its contribution to GDP to 2.4 billion JOD by 2033. Hence, this study fills a significant research gap by uncovering these firms’ tendencies to manage earnings through R&D expenditures.

The contributions of the findings to the literature are both theoretical and practical. Theoretically, the study’s results are believed to fill the research gap by uncovering earnings management strategies through R&D expenditures, which, according to our best knowledge, have not been previously addressed in Jordan. As for the expected practical results, this study will provide information to regulatory bodies, policymakers, such as the Securities Commission and the ASE, and shareholders as well as financial analysts about the importance of transparent accounting practices and policies for R&D. It will also propose appropriate recommendations to limit earnings management through R&D expenditures. Additionally, it is hoped that this study’s results will provide valuable information to lenders, helping them set lending controls for capital projects and integrating them into the investment appraisal process in this field.

2. Literature Review

2.1. Theoretical Background

Capital projects and R&D expenditures are shrouded in ambiguity and asymmetry in the disclosed data that appear in financial reports. As a result, financial analysts and investors alike face difficulty in evaluating the success of research projects and capital investments across different companies. This is due to the uniqueness and variation in research projects and their differences between companies within the same industrial sector or between companies in different sectors in terms of size, cost, nature, context, and activities (Holmstrom, 1989; La Porta et al., 2000; Leuz et al., 2003; Shleifer & Vishny, 1989). Therefore, each R&D project is connected to the nature of a specific company, resulting in asymmetric information and varying types of R&D cost. As a result, executives may gain advantages from knowledge asymmetry to maintain their positions and maximize their wealth.

Changing capitalization accounting rules and choosing the appropriate R&D accounting treatment are other topics that have been discussed (Koch, 1981; Markarian et al., 2008; Nelson et al., 2003; Oswald & Zarowin, 2007; Seybert, 2010; Stadler & Banal-Estanol, 2010; Xiao & Zhou, 2012), which are well represented in the argument between GAAP and IFRS. Accounting for R&D under US GAAPs (SFAS No. 2) requires that all R&D expenses be fully recorded as an expense when incurred. On the other side, International financial reporting standards (IFRSs) necessitate charging research expenditures as expenses when they are incurred and allow for the identification of development costs as intangible assets and capitalization when specific requirements are satisfied.

Positive accounting theory makes an effort to elucidate the theoretical underpinnings of various accounting practices by attempting to comprehend why accountants use various accounting procedures in a variety of contexts and by a variety of businesses. According to Watts and Zimmerman (1990), the positive accounting theory is applied to support accounting procedures and earnings management on the presumption that political and agency costs exist.

In the agency’s opinion, managers and owners have a conflict of interest. In addition, management perks and incentives are tied to bottom-line profit to ensure that they are used in the owners’ best interests. This incentivizes management to work more in order to increase profits. Accordingly, management seeks to exaggerate earnings in the short term to maximize their wealth, contrary to both shareholders’ and lenders’ interests. Moreover, according to economic theory and political processes, diverse accounting practices and regulations result from political demands from various stakeholder groups impacted by yearly financial reports. According to Jensen and Meckling (1976), a manager’s best interests may be served by prioritizing and upholding positive relationships with various stakeholders, including suppliers, employees, and consumers, at the cost of making long-term investments with ambiguous long-term returns.

Consequently, the treatment of R&D expenses remains subjective and is influenced by management’s interest. In this sense, Roychowdhury (2006) views R&D costs as discretionary, implying that executives may strive to lower this spending to increase the declared income, particularly if these expenses do not appear to create profit in the same period.

External auditors are unlikely to request that a CFO provides evidence that the cost of capitalization requirements have been met or have yet to be met in cases where the nature of R&D is complex, accompanied by asymmetric information and significant uncertainty (Smith et al., 2001). The company management can decide whether to capitalize on R&D expenses.

For earnings management, managers use R&D disclosure voluntarily because of its complexity in investments and operations. According to Aboody and Lev (1998), R&D is a productive contribution that is becoming more important but is rarely disclosed.

2.2. Previous Studies About Earnings and RD Activities

Research on the relationship between R&D costs and earnings management in Jordan is needed, as it contrasts with the extensive international focus in advanced countries.

Marei et al. (2023) investigated UK companies and indicated that innovative corporations are more prone to engaging in manipulation, suggesting a greater probability of financial statement manipulation in such enterprises compared to non-creative ones. Markarian et al. (2008) have shown that businesses with high R&D spending are more likely to use earnings management techniques. Seybert (2010) reported that organizations with significant R&D spending tend to employ actual earnings management techniques.

Similarly, Stadler and Banal-Estanol (2010) observed that firms allocating significant resources to R&D are more inclined to adopt actual earnings management strategies to meet or surpass analyst forecasts. Bens et al. (2002) further support the notion that managing profitability through R&D expenditures is possible, highlighting how reducing R&D expenses can act as a protective mechanism against acquisition threats. Additionally, researchers indicate that personal benefits gained by managers, among other earnings management incentives, are important factors in making investment decisions in R&D projects (Baber et al., 1991).

It is worth noting that these previous studies provide a comprehensive and dynamic view of the aspects and ways in which R&D activities and earnings management are related in companies across different countries worldwide. Thus, these studies help in understanding the subject for academic research purposes and practical accounting practices.

Since the tangible results of spending on research, development, and innovation are not visible in the initial years of expenditure and may start to show their effects clearly on business outcomes, product diversity, or innovations after two or three years (Lome et al., 2016), the accounting literature and previous studies largely categorize the reasons for managing earning through R&D into three main themes: Some companies resort to earnings management using R&D expenditures to boost their short-term profitability (Bushee, 1998), while others indicate that some companies use earnings management to justify R&D investments and to meet or exceed financial analysts’ expectations (Bhojraj et al., 2009; Graham et al., 2005). Additionally, it may be used by some management teams to obtain bonuses and compensation linked to achieving profits (Harter & Harikumar, 2004). Furthermore, some management teams use earnings management through R&D when the results of R&D investments are not good and reflect poor management performance.

To demonstrate a company’s manipulation of earnings by choosing accounting practice to capitalize R&D and not expenditures, a recent study conducted in China (Yang et al., 2022) looked at the connection between capitalizing R&D expenses, managing earnings from the viewpoint of key stakeholders for the year 2009–2018. According to the study, the capitalization of R&D expenses and the degree of profit management are positively correlated.

Using the same framework, W. Zhang et al. (2022) concentrated on the problem of earnings management in businesses that gave their employees equity incentives in areas of earnings management which were explored in their research, including the manipulation of R&D intensity and the accounting treatment of such manipulations. China’s Shanghai- and Shenzhen-listed firms from 2014 to 2019 comprise the sample used in their study. Their study’s conclusions show that managers employ earnings management techniques to intentionally lower the benchmarks and exercise prices used in performance evaluations of stock incentive agreements. Managers specifically use expensing to implement accrual earnings management approaches. Additionally, they step up their R&D spending to manipulate actual profitability.

In Turkey, 65 companies that made R&D expenditures and were featured in the Borsa Istanbul stock market index from 2007 to 2018 had their influence on earnings management assessed (Bayraktar & Tütüncü, 2020). The results showed that R&D investment affected earnings management negatively in the short term and favorably in the long run. Management manipulates outcomes after two or three years to convince the public that there is a return from their R&D expense and to defend their investment.

To provide support for the investments made by US corporations in research activities, Chouaibi et al. (2019) conducted a study to assess the correlation between innovation- and real earnings management (REM)-related practices in a sample of 73 publicly traded US businesses from 2000 to 2012. The results of their study indicate that organizations with a creative focus often exhibit a heightened level of involvement in R&D activities. These corporations often resort to the manipulation of earnings.

Several studies pointed out that the management of a corporation may utilize earnings management through R&D to meet the arrangements of the debt covenants, where some banks and creditors stipulate that the company should maintain a minimum level of net income. These conditions push the management to manipulate their annual earnings. In Germany, researchers (Garanina et al., 2016) investigated the motivation underpinned by earning management through the manipulation of R&D. The study was conducted during 2012–2013, and the sample of the study consisted of 121 companies from both Germany and Russia. The study concluded that Russian companies were driven by compliance with loan agreement conditions, while German companies aimed to smooth their profit. Within the same context, researchers (Park & Jung, 2023) also point out that earning management through R&D elevates the rating for credit for firms listed on CRSP during 1992–2019. Accordingly, both studies reached the same conclusion, which stated that maintaining a minimum threshold of profitability through earnings management via R&D is the impetus for some companies to meet loan agreements.

On the other side, R&D can be manipulated by management to beat investor and financial analyst expectations. In Germany, Dinh et al. (2015) investigated the accounting treatment for R&D expenses under the IFRS for 150 firms listed on the stock exchange during the year 1998 and 2012. According to the study’s conclusions, R&D capitalization may be used to increase earnings over projected earnings or increase earnings from the previous year. The study’s findings demonstrated that the choice to capitalize and the amount of capitalization significantly impacted benchmark beating.

Similarly, Dumas (2012) examined the short-term, focused R&D investment made in the French context between 2001 and 2010. The author assumed that managers influence R&D spending to achieve goals such as zero earnings, earnings from the prior quarter, and analyst projections. The results showed that they do actively alter R&D in order to increase profits and, to a lesser amount, increase outcome earnings, despite the managers’ formal lack of a relationship between R&D and analyst forecasts.

In the same perspective, Guidara and Boujelbene (2015) considered whether encouraging R&D reduction after implementing IFRS affected R&D investment. Based on a sample of 800 companies from 2005 to 2014, France’s R&D-intensive industry grew. The results showed that managers routinely cut R&D investment to meet their objectives, which include profitability and positive profit growth.

In order to boost performance and avoid any losses, Tokuga and Tanaka (2011) investigated earnings management and RD cost for Japanese firms from the electronic sector from 1980 to 2006. The study demonstrated that management manipulates R&D spending levels to improve a firm’s immediate performance. In research conducted between 2005 and 2013 on three nations (Italy, Greece, and Spain), Tahinakis (2014) discovered that enterprises often manipulate results by adjusting R&D to avoid losses and declines in short-term profit.

Research indicates that the results of R&D efforts are not observable until a period of two to three years. Furthermore, the aforementioned discussion demonstrates that management is inclined to manipulate earnings through various means, such as concealing any adverse effects on R&D expenditures and attempting to enhance short-term profitability. They may also manipulate earnings to justify investments in research activities to surpass analyst forecasts, meet earning-based bonuses, and mitigate the negative impact of poor performance resulting from R&D investments. The study proposes the following hypothesis based on the argument derived from the literature review:

Hypothesis 1 (RDINT1_L2).

Reporting intensive R&D expenditures relative to total assets lagged by two years (RDINT1_L2) is significantly and positively correlated to earnings management practices for pharmaceutical publicly traded firms listed on the ASE.

Hypothesis 2 (RDINT1_L3).

Reporting intensive R&D spending relative to total assets lagged by three years (RDINT1_L3) and earnings management practices are significantly and positively correlated for pharmaceutical publicly traded firms listed by the ASE.

Hypothesis 3 (RDINT2_L2).

Reporting extensive R&D spending relative to net sales lagged by two years (RDINT2_L2) and earnings management practices are significantly and positively correlated for pharmaceutical publicly traded firms listed on the ASE.

Hypothesis 4 (RDINT2_L3).

Reporting intensive R&D spending relative to net sales lagged by three years (RDINT2_L3) and earnings management practices are significantly and positively correlated for pharmaceutical publicly traded firms listed on the ASE.

3. Research Design and Methodology

3.1. Sample and Data Sources

The primary aim of this study was to examine the potential correlation between earnings management and R&D reporting within the pharmaceutical sector of Jordan, specifically focusing on firms listed on the ASE from 2008 to 2021. This study is supported by data from two sources: primary sources and secondary sources. Secondary sources comprise a thorough examination of prior research and theoretical frameworks that are relevant to the research topic. The information needed to compute company earnings management was acquired from the ASE firm guide. Conversely, the R&D expenditure data were extracted from the annual reports of pharmaceutical companies that are publicly traded on the exchange.

Therefore, to justify the sample and the period of this study, the companies needed to meet particular criteria to be included in the study sample. First, only companies that were listed and actively traded throughout the research period were included. This sets a standard for comparison. Second, mergers were eliminated from the analysis. Mergers change financial performance, making non-merging company comparisons invalid. Third, only organizations with complete variable data were examined. Missing data might affect outcomes. Finally, ASE companies with incomplete financial records were removed to assure data accuracy. These criteria gave this study a representative and credible sample of ASE pharmaceutical businesses for performance examination. The final cohort of five companies comprises the participants of the investigation.

3.2. Accrual Earnings Proxy

In this study, the cross-sectional modified Jones model developed by Dechow et al. (1995) is used as a proxy for the expected earnings management behavior. The total accruals (TACC), which represent the discrepancy between cash flows from operational activities (CFO) and income before exceptional items (NI), are first calculated, as illustrated in Equation (1):

$${TACC}_{it}={NI}_{it}-{CFO}_{it}\dots$$

(1)

Then, we run a regression model. Every year, for each firm, a cross-sectional estimate of the model is made with Equation (2).

$${\displaystyle \frac{{TACC}_{it}}{{TA}_{it-1}}}=a_0+\beta 1{\displaystyle \frac{1}{{TA}_{it-1}}}+\beta 2\left({\displaystyle \frac{\Delta {REV}_{it}-\Delta {REC}_{it}}{{TA}_{it-1}}}\right)+\beta 3{\displaystyle \frac{{PPE}_{it}}{{TA}_{it-1}}}+{\epsilon }_{it}\dots .$$

(2)

The lagged assets are denoted by the variable TA_it−1, which is the book value of all the assets owned by the company at the end of the previous fiscal year (t − 1). ΔREC_it refers to the change in accounts receivable, whereas ΔREV_it indicates the change in sales revenue. PPE_it indicates the property, plant, and equipment that a firm (i) had at the end of year t. TACC is an indicator of total accruals. Estimates of the parameters α, β1, β2, and β3 are obtained through an empirical study. The model’s residual, represented by ε, is computed using the absolute value approach described by (Al-Haddad & Whittington, 2019) and serves as a stand-in for discretionary earnings management.

3.3. The Study’s Primary Regression Model

The intensity of R&D (RDINT) is used in this study as a proxy for R&D investment. RDINT1 and RDINT2 measurements of RDINT are used. In contrast to RDINT2, which is determined by dividing the current year’s R&D spending by net sales lagged by two and three years, RDINT1 is determined by dividing R&D expenditure by total assets lagged by two and three years.

The primary model employs RDINT1 and RDINT2 to test our hypotheses. The model’s equations are as follows:

$${EM}_{it}=\beta 0+\beta 1{RDINT1}_{it-j}+\beta 2{FSize}_{it}+\beta 3{FLev}_{it}\dots$$

(3)

$${EM}_{it}=\beta 0+\beta 1{RDINT2}_{it-j}+\beta 2F{Size}_{it}+\beta 3{FLev}_{it}\dots$$

(4)

R&D intensity is often expressed as a percentage of net sales or as a fraction of a company’s total assets, as stated by (Ding et al., 2007; Dumas, 2012; Grabinska & Grabinski, 2017; Nekhili et al., 2012; Osma, 2008; Osma & Young, 2009; Xu & Yan, 2013; X. Zhang & He, 2013). A higher R&D intensity implies that a firm is devoting a significant proportion of its resources to R&D operations, which can indicate the company’s dedication to innovation and its potential for future development and profitability. Another commonly used measure of R&D intensity is the proportion of R&D expenses relative to a company’s total assets. This metric shows the proportion of R&D costs to the company’s total asset value. The regression model employed in this study has two control variables. First, the firm’s size (FSize) is calculated by calculating the natural logarithm of its total assets. Lee and Choi (2002) discovered that big enterprises are less likely to manage profitability. The ratio of total liabilities to total assets (FLev) is used to monitor financial leverage and identify how management changes profits to meet debt covenants. According to Beatty and Weber (2003), a higher leverage encourages corporations to manipulate results upward.

4. Analysis and Hypothesis Testing

4.1. Descriptive Analysis of Key Variables in Pharmaceutical Companies

This section describes pharmaceutical industry variables, including R&D expenses, earnings management, firm size, and firm leverage. The dataset contains 70 observations from 2008 to 2021 from ASE-listed pharmaceutical businesses. Important results include R&D intensity, earnings management, and company-specific features. Investors, policymakers, and business professionals may learn about the pharmaceutical industry from these findings.

The value mean for accrual earnings management (EM) is 0.081, as shown in

Table 1. Exploring key financial metrics of pharmaceutical companies: A descriptive analysis.

Variable	Obs.	Mean	Std. Dev.	Min	Max	Variance	Skewness	Kurtosis
EM	70	0.081	0.066	0.0040	0.302	0.0043	1.332	4.466
RDINT1_L2	70	0.017	0.019	0.0005	0.155	0.0004	5.647	40.376
RDINT1_L3	70	0.018	0.022	0.0004	0.182	0.0005	5.868	42.927
RDINT2_L2	70	0.040	0.047	0.0013	0.283	0.0022	4.054	20.479
RDINT2_L3	70	0.052	0.083	0.0016	0.490	0.0069	3.967	19.060
FSize	70	17.096	0.962	15.1127	18.501	0.9260	−0.244	2.202
FLev	70	0.519	0.424	0.0930	1.904	0.1800	1.581	5.191

. This indicates that, on average, corporations employ accruals to manage their earnings. A low standard deviation of 0.066 indicates that the dataset’s accrual management earnings level does not vary sufficiently. The proxies RDINT1_L2 and RDINT1_L3 reflect the research intensity, notably the ratio of R&D expenses to total lagging assets. These characteristics illuminate company research resource allocation. The average research intensity is (0.017) to (0.018), showing that organizations dedicate a tiny amount of their assets to research. The dataset’s standard deviations range from (0.019) to (0.022), indicating research intensity variations. The second proxies for R&D intensity, RDINT2_L2 and L3, show research intensity compared to net sales revenue. These factors give a different viewpoint on research funding. The average research intensity is (0.040) to (0.052), suggesting that organizations spend just a small percentage of their sales income on research. The dataset’s standard deviations range from (0.047) to (0.083), indicating a variety in research intensity.

Moreover, the pairwise correlations are reported in

Table 2. Pairwise correlation matrix.

Variables	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
(1) EM	1.000
(2) RDINT1	0.156	1.000
(3) RDINT1_L2	−0.272 *	0.103	1.000
(4) RDINT1_L3	0.217	0.042	0.263 *	1.000
(5) RDINT2	0.284 *	0.911 *	−0.046	−0.092	1.000
(6) RDINT2_L2	−0.112	−0.006	0.691 *	−0.060	0.038	1.000
(7) RDINT2_L3	0.207	−0.060	0.026	0.685 *	−0.097	0.107	1.000
(8) FSize	−0.321 *	−0.134	−0.224	−0.059	−0.210	−0.344 *	−0.187	1.000
(9) FLev	0.141	−0.194	−0.390 *	−0.508 *	0.050	0.188	0.016	0.004	1.000

Notes: the asterisks * indicate significance at 5%.

. Remarkably, earnings management (EM) shows a positive and significant correlation with RDINT2 (0.284), suggesting that research intensity measured relative to net sales revenue may be linked to higher earnings management practices. Conversely, EM has a significant negative correlation with RDINT1_L2 (−0.272) and firm size (−0.321), indicating that larger firms and lagged R&D intensity relative to total assets may reduce earnings management tendencies. There are no indications for multicollinearity between the study variables, except for the correlation between R&D proxies. Hence, these variables shall be included in separate models. Next,

Table 3. Variance inflation factors (VIFs) and tolerances of independent variables.

Variable	VIF	Tolerance (1/VIF)
RDINT1_L2	1.12	0.893
RDINT1_L3	1.12	0.891
RDINT2_L2	1.07	0.937
RDINT2_L3	1.15	0.868
FSize	1.03	0.968
FLev	1.01	0.989

shows an additional test for a multicollinearity assessment between the independent variables.

4.2. Multicollinearity Assessment

Table 3 provides a multicollinearity assessment for independent variables. The regression model’s independent variables show minimal multicollinearity, with VIF values ranging from 1.01 to 1.15. The tolerances are near one, indicating that a significant percentage of volatility occurs independently (Hair et al., 1995). This improves the accuracy of the predictions and provides a more precise evaluation of both independent and dependent variables.

4.3. Stationary Test (Unit Root Test) for Study Variables

In a time series analysis, data can be categorized as stationary or non-stationary. Stationary series exhibit fluctuations around a constant mean over time, without any inherent trend of increase or decrease. Additionally, the relationship between any two points in the series depends solely on the time lag between them, not the specific time at which the connection is measured. Non-stationary series, on the other hand, display a continuously changing mean, which is either increasing or decreasing. Their variance also fluctuates over time. To determine the stationarity of the variables being studied, particularly in the panel data with a time dimension, the Levin–Lin–Chu (LLC) test is employed. This test helps assess the presence of a unit root, a non-stationary component that can distort statistical analysis. If these variables exhibit a unit root, it is necessary to take their differences to render them stationary. This is because many time series may be non-stationary but yield high values of (R2, F, t), which can lead to wrong interpretations and misleading outcomes. Hence, it is necessary to conduct a unit root test in order to assess the stationarity of the time series.

The LLC test determines the presence of a unit root, indicating the non-stationarity of the time series if the derived test value has a significance level larger than 0.05.

The data shown in

Table 4. The results of the panel unit root test for all study variables.

Variable	Statistics at Level	p-Value	Result
EM	−4.411	0.000	Stationary
RDINT1_L2	−1.899	0.029	Stationary
RDINT1_L3	−1.675	0.047	Stationary
RDINT2_L2	−2.287	0.011	Stationary
RDINT2_L3	−2.625	0.004	Stationary
FLev	−3.457	0.000	Stationary
FSize	−3.093	0.001	Stationary

indicate that the research variables remain constant throughout time, as the p-values do not surpass the significance level of 0.05. Therefore, the study utilized stationary time series data because none of the p-values for the variables were above the 5% threshold. Therefore, the hypothesis that there is a unit root is disproven, indicating that the time series is stationary.

4.4. Results of Hypothesis Testing

The hypothesis is assessed using the cross-sectional time series general least squares regression model. The outcomes of the panel regression analyses for the variables RDINT1_L2, RDINT1_L3, RDINT2_L2, and RDINT2_L3, respectively, are shown in

Table 5. The outcomes of a panel regression for the RDINT1_L2 variable.

Em	Coef.	St. Err.	t-Value	p-Value	Sig
RDINT1_L2	0.898	0.362	2.48	0.013	**
FLev	0.131	0.062	2.10	0.036	**
FSize	−0.016	0.008	−2.03	0.043	**
Cons	0.345	0.133	2.59	0.010	***
Wald chi2(3)	15.467
Prob > chi2	0.0015

Notes: The asterisks **, and *** indicate significance at 5%, and 10%, respectively.

,

Table 6. The outcomes of a panel regression for the RDINT1_L3 variable.

Em	Coef.	St. Err.	t-Value	p-Value	Sig
RDINT1_L3	0.732	0.317	2.310	0.021	**
FlEV	0.132	0.063	2.100	0.036	**
FSize	−0.016	0.008	−2.020	0.043	**
Cons	0.347	0.134	2.580	0.010	***
Wald chi2(3)	14.53
Prob > chi2	0.0023

Notes: The asterisks **, and *** indicate significance at 5%, and 10%, respectively.

,

Table 7. The results of a panel regression for the variable RDINT2_L2.

Em	Coef.	St. Err.	t-Value	p-Value	Sig
RDINT2_L2	0.244	0.132	1.850	0.065	*
FLev	0.113	0.064	1.770	0.077	*
FSize	−0.017	0.008	−2.120	0.034	**
Cons	0.364	0.136	2.680	0.007	***
Wald chi2(3)	14.53
Prob > chi2	0.0023

Notes: The asterisks *, **, and *** indicate significance at 1%, 5%, and 10%, respectively.

and

Table 8. The outcomes of a panel regression for the RDINT2_L3 variable.

Em	Coef.	St. Err.	t-Value	p-Value	Sig
RDINT2_L3	0.211	0.107	1.970	0.049	**
FLev	0.110	0.064	1.730	0.084	*
FSize	−0.016	0.008	−1.960	0.050	*
Cons	0.342	0.136	2.510	0.012	**
Wald chi2(3)	12.90
Prob > chi2	0.0049

Notes: The asterisks *, and ** indicate significance at 1%, and 5%, respectively.

.

Table 5 shows that RDINT1_L2 demonstrates a positive correlation with earnings management (coefficient: 0.898, p < 0.05). Likewise, a positive correlation of 0.131 (p < 0.05) between FLev and enhanced earnings management establishes a connection between high financial leverage and earnings management. On the other hand, the negative coefficient of −0.016 (p < 0.05) for FSize indicates that larger organizations might engage in less effective earnings management. The model is statistically significant (Wald chi-square: 15.467, p-value: 0.0015). This study supports that there is a significant positive correlation between earnings management and the intensive reporting of R&D expenditures relative to total assets prolonged by two years (RDINT1_L2) in pharmaceutical firms listed on the ASE.

Table 6 shows that the RDINT1_L3’s coefficient is 0.732 (p < 0.05), indicating a statistically significant positive correlation between RDINT1_L3 and earnings management. FLev demonstrates a positive correlation of 0.132 (p 0.05), indicating that improved earnings management relates to larger financial leverage. Given that FSize has a negative coefficient of −0.016 (p < 0.05), enterprises with higher total assets tend to have less effective earnings management. The Wald chi-square test statistic is 14.53 (p < 0.05). These results support Hypothesis two and confirm a favorable correlation between reporting R&D expenditures relative to total assets lagged by three years (RDINT1_L3) and earnings management practice.

Table 7 points out that the coefficient for RDINT2_L2 is 0.244 (p 0.1), showing a positive correlation between RDINT2_L2 and earnings management. The LEVG exhibits a positive coefficient of 0.113 (p 0.1), suggesting a positive potential association between financial leverage and EM. FSize has a negative coefficient of −0.017 (p < 0.05), indicating that lower levels of earnings management are connected with larger enterprises when assessed against total assets. The Wald chi-square test statistic (14.53) (p < 0.05) indicates that the overall model is statistically significant. These results support Hypothesis three, confirming that reporting R&D spending relative to net sales lagging by two years and earnings management practices are positively correlated.

Table 8 indicates that the RDINT2_L3’s coefficient is 0.211 (p < 0.05) showing a statistically significant positive relationship between RDINT2_L3 and earnings management practice.

FLEV has a positive coefficient of 0.110 (p 0.1), which suggests a potential positive link between financial leverage and earnings management. Companies with more significant total assets often have less effective earnings management, given that FSize has a negative coefficient of −0.016 (p 0.1). The Wald chi-square test result (12.90) (p < 0.05) indicates that the overall model is statistically significant. These results support the notion that reporting R&D expenditures relative to net sales is delayed by three years, and earnings management strategies are positively correlated.

5. Discussion

5.1. Examining the Findings

Pharmaceutical businesses from Jordan listed on the ASE with R&D activities and earnings management are significantly and favorably correlated, both as a share of net sales and as a share of total assets, according to

Table 9. Summary of the study results.

Time Lag	RDINT1	FLev	FSize	RDINT2	FLev	FSize
2	(+) **	(+) **	(−) **	(+) *	(+) *	(−) **
3	(+) **	(+) **	(−) **	(+) **	(+) *	(−) *

Notes: The asterisks *, and ** indicate significance at 1%, and 5%, respectively.

. This connection is still valid when R&D efforts are two or three periods behind. Earnings management is measured using accrual earnings.

These results align with previous studies, including Bushee (1998), Bens et al. (2002), Harter and Harikumar (2004), Markarian et al. (2008), Bhojraj et al. (2009), Seybert (2010), Stadler and Banal-Estanol (2010), Tokuga and Tanaka (2011), Guidara and Boujelbene (2015), Garanina et al. (2016), and Grabinska and Grabinski (2017) for a three-period lag and for two periods as a percentage of total assets. Conversely, other researchers have highlighted negative associations. For instance, Bayraktar and Tütüncü (2020) observed that R&D expenditures in Turkish firms reduced earnings management in the short term, suggesting that R&D spending may enhance transparency and discourage manipulation. Additionally, studies like those by Bushee (1998) indicate that institutional investors often discourage myopic R&D behavior, thereby limiting earnings manipulation. These mixed findings underscore the complexity of the R&D–earnings management nexus, influenced by factors such as firm size, financial leverage, and managerial incentives. Moreover, Grabinska and Grabinski (2017) identified this behavior for two lag periods as a percentage of sales. Bayraktar and Tütüncü (2020) argue that R&D is a key factor in determining how earnings are managed after one and two time delays as a proportion of assets. Their conclusions diverge, regarding R&D as a proportion of revenue, where they have yet to discover a statistically significant association.

This study also demonstrates that the size of the firm negatively impacts earnings management. This shows that when a firm grows, corporate governance improves, leading to a decrease in need for earnings management. Albrecth and Richardson (1990), Scott (1991), and Lee and Choi (2002) all agreed with this conclusion.

Table 9 illustrates the extremely advantageous association between financial leverage and earnings management. This shows that leverage impacts earnings management since businesses may use earnings management to comply with debt covenants. These findings are in agreement with Duke and Hunt (1990), Bartov (1993), Beatty and Weber (2003), Park and Jung (2023), and other pertinent investigations.

5.2. Explanation of the Study Outcomes

Due to the discretionary nature of R&D costs and the difficulty for investors to conclude a firm’s R&D performance in comparison to other companies, this study concludes that there is a positive relationship between R&D activities and earnings management in the Jordanian context that leads to the conclusion that executives may manage earnings through R&D expenditures.

Theoretically, this study’s outcomes support and agree with both the positive accounting theory and agency theory. Both theories explain the constraints and limitations as well as the pressure either from stakeholders or political pressure that could motivate management regarding accounting choices when there are different choices available to be utilized. In the context of Jordan, as well as any other worldwide country, the choice of financial reporting policies and methods and the process of selecting accounting alternatives in the treatment of R&D expenses by Jordanian pharmaceutical companies are elaborated by the positive accounting theory and agency theory. Accordingly, in this context, both theories explain and contribute to the identification of the extent to which the choice of accounting alternatives in practice is influenced by the interests of management, which must be continuously incentivized to achieve the best interests of the economic entity.

This study’s findings further support the claim that management interests still have a role in how R&D costs are treated because it is still subjective. Management has the flexibility to influence claimed profitability because of its discretionary choice to capitalize on R&D spending and the lack of external auditor monitoring. Additionally, and in line with the positive accounting theory and agency costs, the overlapping interests of owners and management and the connection between management benefits as well as bottom-line profit in the context of Jordan as an emerging market motivate managers to work harder to earn higher salaries. Companies may also control their earnings to avoid violating debt covenants and safeguard their interests.

Finally, in the Jordanian context due to the weakness of observing insider trading and due to information asymmetry and insider benefit, this study’s results suggest that managers may exploit voluntary R&D disclosures for earnings management purposes, taking advantage of the intricate nature of R&D spending.

6. Research Limitations and Policy Recommendations

This study provides valuable insights into the relationship between R&D intensity and earnings management in the Jordanian pharmaceutical sector. The primary limitation of this study lies in its limited scope and small sample size. While it includes all pharmaceutical companies operating in Jordan, the small number of firms presents a challenge for the generalizability of the findings. Nevertheless, this study provides key insights for regions with similar contexts, especially in developing countries.

The findings of this research provide far-reaching implications. According to a survey conducted by the European Bank for Reconstruction and Development, Jordanian companies that invest in R&D activities, such as those operating in the pharmaceutical sector, face many obstacles. Among the most significant are investor protection, corporate governance, and external auditing (EBRD, 2017). This necessitates that regulatory and supervisory bodies in Jordan take the necessary measures that would enhance transparency, accountability, and effective governance if implemented. For instance, in terms of auditing, Jordanian regulatory and supervisory bodies, such as the Jordan Securities Commission and the ASE, must improve and enhance the audits of the accounting treatments for R&D expenses. Therefore, periodic reviews must be conducted to ensure companies’ adherence and compliance with international standards, such as the clearly defined capitalization and transparency standards for R&D expenses.

Secondly, shareholders must be integrated into governance and oversight mechanisms. Active and effective shareholder participation and attendance at general meetings of listed companies serve as tools for monitoring a management’s commitment to transparency in conducting business activities. Annual general meetings can also be used to resolve issues related to R&D accounting. Thirdly, independent external auditors have the responsibility to verify the company’s financial statements and reports fairly and honestly because they represent the company’s financial position and performance in all material aspects. There must be an objective evaluation of the accounting policies and procedures for R&D expenses, as well as alternatives available to capitalize these expenses. Financial analysts must also conduct feasibility studies for R&D projects. These studies include reviewing the resources allocated for R&D, the availability of disclosure and transparency for capital project expenditures, the economic feasibility of these investments, and the disclosure of earnings management activities as well as any manipulation of earnings or the smoothing of income.

Finally, interactions between financial institutions, lending institutions, and companies investing in R&D must be strengthened. Investment banks can provide advice and guidance on directing R&D expenditures, which will enhance the efficiency of spending through investment standards and controls and provide insight into the risks of capital projects and R&D from an independent, professional perspective.

7. Conclusions

This study examines the relationship between R&D intensity and earnings management in Jordanian pharmaceutical companies listed on the ASE, showing how discretionary R&D spending is used to influence financial reporting. The findings show that R&D intensity is positively associated with earnings management (the relationship is stronger when R&D investments are lagged by two or three years), emphasizing the importance of regulatory oversight, robust governance, and independent auditing to control earnings manipulation and promote financial reporting transparency. The findings align with the positive accounting and agency theories, emphasizing the influence of managerial discretion and stakeholder pressures in shaping financial reporting practices. By addressing these challenges, Jordan’s pharmaceutical sector can enhance its contribution to the economy while maintaining investor trust.

Future research should expand the scope of the study geographically and temporally to explore this nexus in broader contexts, contributing to the global understanding of R&D accounting practices and financial management. Moreover, researchers should explore the role of regulatory frameworks in influencing earnings management practices across emerging markets. Future studies should utilize up-to-date datasets and investigate the specific role of corporate governance mechanisms in mitigating earning manipulation in R&D intensive sectors. In return, this will reflect the recent economic and financial situations of the pharmaceutical companies and generate a comprehensive analysis of the variables that affect earnings management.