Corporate Social Responsibility, Advertising Intensity, and Performance: The Importance of National Philanthropic Environments

Abstract

This study investigates the relationships among Corporate Social Responsibility (CSR), advertising intensity, and performance in different national philanthropic environments. National philanthropic environment is defined as the propensity of a nation to be philanthropic. Employing institutional theory, we hypothesize that (1) advertising intensity mediates the CSR/performance relationship, (2) national philanthropic environment moderates the relationship between CSR and advertising intensity, and (3) national philanthropic environments moderate the relationship between advertising intensity and performance. Using a sample of 262 firms from 10 countries, the results support the moderated mediation model. We demonstrate that the success of CSR programs depends on their effective integration with advertising intensity. However, this dynamic is relevant only in countries with higher levels of national philanthropic environment. Thus, strategic managers should increase advertising intensity as CSR increases to ensure a positive effect on performance, but only in countries with high national philanthropic environments.

Keywords:

• CSR
• advertising intensity
• national philanthropic environment
• institutional theory

Introduction

Corporate social responsibility (CSR) has gained significant importance in marketing and advertising research in recent decades and continues to be of importance in the current literature (e.g., Curras-Perez et al., 2023; Snipes et al., 1999; Wang et al., 2022). CSR is defined as a firm action that goes beyond economic and legal interests to promote social good (McWilliams & Siegel, 2001). Numerous studies have shown that CSR positively influences key strategic outcomes such as consumer trust (Pérez et al., 2020), consumer attitudes (Vera-Martínez et al., 2022), brand equity development (Huang, 2023), customer behavioral loyalty (Liu et al., 2020), market share (Rahman et al., 2017), and ultimately financial performance (Kiessling et al., 2016). However, the relationship between CSR and performance is complex and contingent upon other strategic constructs (Kim et al., 2018; Sun et al., 2019).

One such construct is advertising intensity, which refers to the proportion of advertising expenditure relative to a company’s overall resource base (Mirzaei et al., 2016; Rahman et al., 2017). Understanding when to invest in advertising intensity has been found to be of strategic importance to firm performance (Huang & Liu, 2022), however, the nature of the relationship between CSR, advertising intensity, and performance remains unclear in marketing research (Hayes & Duff, 2022). This is alarming, considering that in recent years, global consumers have not only become more concerned with the social responsibility of companies, but many consumers specifically report not believing firm CSR efforts until firms communicate such efforts to them, which highlights the importance of understanding the dynamic between CSR and advertising (Cone Communications, 2017; Hayes & Duff, 2022). Further, the rise in stakeholder dialogue regarding corporate hypocrisy or the discrepancy between CSR talk and action, as well as skepticism toward CSR, emphasizes the need for proper CSR communications and advertising in current times (Andersen & Høvring, 2020; Pérez et al., 2019). Understanding this shortcoming, Taylor (2018) makes an urgent call for researchers to conduct more studies on the nature and specificities of the relationship between CSR and advertising, and how this relationship influences performance. However, Hayes and Duff (2022) recently indicated that this shortcoming persisted. Therefore, this study seeks to fill this gap by investigating CSR, advertising intensity, and performance.

Prior studies on this topic have yielded mixed results. Some studies suggest that advertising intensity positively moderates the influence of CSR on performance (e.g., Rahman et al., 2017), while others find a negative moderating effect (e.g., Hu et al., 2018), direct relationship (e.g., Zhang et al., 2010), or no significant link (e.g., Kang et al., 2016). This ambiguity further highlights the need for continued investigation to understand how the CSR/advertising intensity relationship influences performance. Therefore, our first research question is as follows: How does the relationship between CSR and advertising intensity influence performance?

To understand the CSR/advertising intensity link fully, it is essential to consider the contextual nature of these relationships. CSR is contingent on institutional environments (Randrianasolo, 2018), and international marketing research emphasizes the importance of studying CSR in different countries (Eteokleous et al., 2016; Randrianasolo & Semenov, 2022; Semenov & Randrianasolo, 2022a). Thus, we investigate the influence of CSR/advertising intensity on performance in different national philanthropic environments (NPE), which reflect a nation’s propensity for philanthropy. NPE is a national-level construct that is crucial for understanding CSR dynamics within different institutional environments (Randrianasolo & Semenov, 2022; Semenov & Randrianasolo, 2022a). Therefore, our second research question is: How does the relationship between CSR and advertising intensity influence performance within different national-level philanthropic environments?

This study contributes novel insights into the understanding of CSR, advertising intensity, and firm performance. First, it advances knowledge by uncovering the dynamics of the CSR/advertising intensity relationship and its impact on firm performance, addressing the ambiguity of existing research. Second, while the importance of advertising for CSR success has been acknowledged, market contexts that require stronger synergies between CSR and advertising intensity to influence performance have been overlooked. Our study addresses this gap by investigating this relationship in different NPE contexts. Building on Semenov and Randrianasolo (2022a) finding of NPE's moderating influence on the CSR/advertising link, we extend the research by examining its implications for performance, providing strategic insights into where and how CSR/advertising intensity can influence performance. In other words, where Semenov and Randrianasolo (2022a) uncover the moderating role that NPE plays in the relationship between CSR and advertising intensity, we extend this knowledge by examining the mediating role of advertising intensity in the CSR/performance link, as well as the moderating role NPE plays in this model. Finally, although there is an abundance of studies exploring the link between social and financial performance in domestic contexts, international marketing research lacks such studies (Eteokleous et al., 2016). Our study fills this gap by examining the model of CSR, advertising intensity, and performance in various national contexts. To achieve the objectives of this study, the following sections review the relevant literature, develop hypotheses based on theory, empirically test the hypotheses, and discuss the implications of the findings.

Background and Hypotheses

In this section, we use institutional theory to formulate hypotheses regarding the influence of CSR and advertising intensity on performance. Additionally, we propose hypotheses to explore how these dynamics may vary across different national philanthropic environments.

CSR, Advertising Intensity, and Performance

The existing literature presents conflicting findings concerning the relationships among CSR, advertising intensity, and performance. While some studies indicate positive direct relationships between CSR/performance and advertising intensity/performance (Liu et al., 2018; Maury, 2022; Servaes & Tamayo, 2013), others suggest a direct negative relationship (Rashid et al., 2020). Certain researchers have proposed a positive moderating relationship, emphasizing that the interaction between CSR and advertising intensity enhances performance (Assaf et al., 2017; Bashir, 2022; Rahman et al., 2017). However, other studies find a negative moderating interaction between these variables in terms of their influence on performance (Fisman et al., 2007; Hu et al., 2018). To address this ambiguity, we employ institutional theory to propose a mediation model incorporating CSR, advertising intensity, and performance.

Institutional theory posits that institutional environments shape and constrain firm behavior (Campbell, 2007; Randrianasolo, 2021). From this perspective, firms exhibit behaviors aligned with the values of their institutional environments to gain social acceptance and organizational legitimacy (Randrianasolo, 2018). Firms strive to obtain legitimacy from their respective environments to achieve strategic benefits such as improved access to distribution channels, positive consumer attitudes, and supplier relationships (DiMaggio & Powell, 1983; Randrianasolo, 2018). Lack of legitimacy in the institutional environment can have detrimental effects on firm performance (Park et al., 2012). One strategy employed by firms to acquire legitimacy is CSR (Campbell et al., 2012), particularly in institutional environments where there is pressure to be socially responsible (Randrianasolo & Semenov, 2022).

Given that CSR is a strategy that seeks legitimacy, firms that choose this approach are likely to communicate their CSR activities to stakeholders within their institutional environments to attain legitimacy and, ultimately, enhance performance. This perspective is supported by a body of research suggesting that advertising plays a crucial role in CSR success (Mögele & Tropp, 2010). Studies have revealed a positive association between corporate philanthropic giving and advertising intensity (Zhang et al., 2010). Firms investing in CSR tend to have higher levels of advertising intensity to effectively communicate their CSR initiatives to stakeholders and enhance their corporate image (Pomering & Johnson, 2009). Therefore, “advertising plays an important role in raising awareness among individuals interested in purchasing products with CSR attributes” (McWilliams & Siegel, 2001, p.120).

From an institutional perspective, it can be stated that for constituents within an institutional environment to grant social acceptance, or legitimacy, to a firm, the constituents need to be aware of the firm’s attributes that are crucial for such social acceptance. If a firm seeks legitimacy from constituents in the environment by employing CSR, the firm should raise awareness of such CSR activities to legitimacy-granting constituents. Thus, socially responsible firms tend to invest more in advertising (Fernández-Kranz & Santaló, 2010). This perspective is supported by prior research which has found that CSR communications, perceptions, and performance influence consumer stakeholders’ brand preferences, loyalty, and perceptions (Liu et al., 2014; Liu et al., 2014; Woo & Jin, 2016)

Supporting this stance, Maury (2022) found that firms that participate in CSR are likely to adopt stronger prospector strategies, which are reflected in firms with higher levels of overall advertising intensity. This finding indicates that firms seeking legitimacy in their respective environments through CSR are likely to invest more in overall advertising. It is important to note that this position does not state that CSR influences CSR advertising; rather, it states that CSR influences the overall advertising intensity of a firm. This is in line with prior research which posits that “advertising enhances a firm’s information environment, thereby increasing the firm’s (potential) customers’ awareness about the firm and likely prompting them to become further informed about the firm, its products, and practices, including its corporate social performance” (Servaes & Tamayo, 2013, p. 1047). We adopt this perspective and propose that advertising intensity mediates the relationship between CSR and performance.

To hypothesize the mediation of advertising intensity between CSR and performance, it is essential not only to provide theoretical support for the CSR/advertising relationship, as discussed above, but also to justify why this link influences performance. As a legitimacy-seeking strategy, CSR needs to be advertised to constituents in the institutional environment, which boosts legitimacy and ultimately performance. The theoretical position here is that (1) CSR seeks legitimacy from legitimacy-granting constituents; (2) firms that employ CSR boost their overall advertising intensity to enhance their information environments for legitimacy-granting constituents; (3) firms that increase advertising intensity (and, in turn, information environments) increase constituent awareness about their attributes, including their CSR; and (4) such firms gain legitimacy and yield superior performance because gaining legitimacy boosts performance. Rooted in institutional theory and evidence from previous research (e.g., Maury, 2022; Servaes & Tamayo, 2013), we formalize our first hypothesis:

Hypothesis 1:

Advertising intensity mediates the relationship between CSR and firm performance.

CSR, Advertising Intensity, and NPE

Organizational social performance not only depends on internal resources and strategies but also on external environmental factors such as market orientations and institutional environmental forces (Lăzăroiu et al., 2020; Randrianasolo, 2018). Given that CSR is a strategy that firms employ to adhere to institutional pressures of social responsibility (Randrianasolo & Semenov, 2022), we propose that firms are more likely to align their CSR efforts with advertising intensity in institutional environments with a higher pressure to be socially responsible.

Recent research has proposed that NPE indicates such an environment (Randrianasolo & Semenov, 2022; Semenov & Randrianasolo, 2022a). NPE is defined as the propensity of a nation’s people and organizations to voluntarily contribute to the social good through donations of money, time, resources, or other valuable entities (Randrianasolo & Semenov, 2022). It reflects the normative institutional pressures that exist within an institutional environment for firms to be philanthropic. Marketing research proposes that countries with higher NPE levels have higher pressure for firms to employ CSR (Randrianasolo & Semenov, 2022). Further, NPE determines the complementary vs. substitute nature of advertising intensity and CSR (Semenov & Randrianasolo, 2022a). We adopt this perspective, rooted in institutional theory, to hypothesize that NPE moderates the relationship between CSR and advertising intensity. The theoretical position here is that in countries with higher NPE, firms experience higher pressure to be socially responsible, and therefore are more likely to boost their advertising intensity to increase their information environments and prompt their constituents to become further informed about the firm, its products, and practices, including its CSR. The second hypothesis if formalized:

Hypothesis 2:

NPE positively moderates the relationship between CSR and advertising intensity.

Advertising Intensity, Performance, and NPE

In their study on the relationship between advertising intensity and performance, Semenov and Randrianasolo (2022b) propose that the relationship between advertising intensity and performance is not conclusively positive and includes contingencies. This position is reflected in the literature on this relationship, where some studies find a positive link (Liu et al., 2018), others find a negative link (Meyer & Ujah, 2017), and others find no significant relationship (Long et al., 2020). Adopting Semenov and Randrianasolo (2022b) proposition, we consider this relationship to be contextual and have contingencies. In this study, we hypothesized that NPE provides one contextual contingency that determines the influence of advertising intensity on performance. This proposition is made for two reasons:

First, research proposes that the normative institutional environment of countries with higher levels of NPE is composed of organizations that have a stronger propensity to voluntarily be philanthropic, indicating that firms in such environments are likely to have higher levels of CSR (Randrianasolo & Semenov, 2022; Semenov & Randrianasolo, 2022a). Thus, with heightened CSR levels in such environments, it may be more difficult for firms to employ CSR as a differentiation strategy to gain competitive advantages. One method of differentiation that firms may utilize to build advantages is advertising (Zhang et al., 2010).

In countries with high NPE, CSR is expected and demanded by stakeholders (Randrianasolo & Semenov, 2022; Semenov & Randrianasolo, 2022a); therefore, it is a prerequisite for performance. However, since this is an expectation due to institutional pressures in such environments, firms might not be able to achieve competitive advantages by increasing their CSR. Therefore, CSR is necessary in such environments but will likely result in competitive parity. In such environments, firms may rely on advertising as a differentiation factor to develop competitive advantages and ultimately enhance performance. The theoretical position here, therefore, is that when CSR pressure is high (high NPE), more firms are likely to participate in CSR and therefore diminish the differentiation utility that CSR may bring firms in their quest to leverage competitive advantages. Since advertising intensity can also serve as a differentiation method (Hsu, 2012; Samu et al., 1999; Zhang et al., 2010), we posit that the influence of advertising intensity on performance is stronger in countries with high NPE.

Second, since there is stronger normative pressure to be socially responsible in higher rather than lower NPE countries (Randrianasolo & Semenov, 2022), firms in higher NPE countries that increase advertising intensity to complement CSR yield better legitimacy and in turn performance than those from lower NPE countries. Specifically, in countries with higher NPE, where firms are likely to increase advertising intensity along with CSR, as argued in H2, those with higher advertising intensity are proposed here to be more likely to yield superior performance. Thus, building off H1 and H2, we hypothesize that NPE moderates the relationship between advertising intensity and performance and formalize the third hypothesis:

Hypothesis 3:

NPE positively moderates the relationship between advertising intensity and firm performance.

Moderated Mediation Model

Thus far, we have hypothesized that (1) advertising intensity mediates the relationship between CSR and performance, (2) NPE positively moderates the relationship between CSR and advertising intensity, and (3) NPE moderates the relationship between advertising intensity and performance in the context of the focal variables in this study. Holistically, these hypotheses culminate into a moderated mediation model. Specifically, the premise of the proposed model, grounded in institutional theory, posits that the alignment between CSR and advertising intensity as well as its effect on performance is dependent on the external normative institutional pressure represented by NPE. Firms that employ CSR are more likely to align this strategy with advertising intensity in countries with higher NPE and as such yield superior performance in these countries. The final hypothesis is, therefore, the moderated mediation model, as formalized below. Figure 1 displays the conceptual model.

Figure 1. Conceptual model.

Hypothesis 4:

There is a moderated mediation, where NPE moderates the relationship between CSR and advertising intensity as well as the relationship between advertising intensity and firm performance.

Methodology

Sample and Data Collection

We collected data for the years 2016-2018 from multiple sources. CSR ratings from the CSRHub database were used to measure firm-level CSR. CSRHub offers ratings on 8,419 companies worldwide developed from four categories: governance, environment, community, and employees. The overall CSR ratings represent aggregate scores of these four categories and vary between 0 and 100, where 100 is the highest.

We used the Global Philanthropy Environment Index to measure country-level NPE. This index measures five key factors of philanthropic environments: ease of operating philanthropic organizations, tax incentives, cross-border flows, political environment, and sociocultural environment. This Index reflects countries’ philanthropic enabling environment on a scale of 1.0 to 5.0 (IUPUI, 2019).

Firm-level advertising expenditures, net income, total sales, total assets, total equity, return on assets (ROA), return on sales (ROS), and return on equity (ROE) were collected from the Bloomberg database.

Year of incorporation, number of employees, firm age, R&D expenditures, firm size, financial slack, leverage, year of the initial public offering (IPO), industry classification (Global Industry Classification Standard (GICS), and industry profitability were collected from the Bloomberg database to be included as control variables.

Our search of the database for information on the companies with CSR ratings resulted in 1,883 firms from 22 countries. We eliminated firms with no information on advertising and firms from countries without the Global Philanthropy Environment Index. The final sample consisted of 262 companies from 10 countries. No significant differences in company size were found (p<.05) between firms included and excluded from the final sample.

Cultural dimensions’ data was collected from Hofstede Insights (2022). The data on government regulatory conditions of the countries was collected from the Economic Freedom of the World report (Gwartney et al., 2019). Gross domestic product (GDP) per capita was collected from The World Bank (2019). Descriptive statistics, means, standard deviations, and correlations of the variables in the study, and industry/country variables are shown in Tables 1 and 2.

Table 1. Descriptive statistics, means, Standard deviations, and correlations.

Variables	Mean	s.d.	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
1. ROA	5.67	6.91
2. ROS	6.69	9.53	.69**
3. ROE	11.48	8.88	.73**	.57**
4. CSR	53.58	8.23	−0.05	−0.01	.01
5. Advertising intensity	.11	.11	.08	−0.00	.18**	.09
6. NPE	4.07	.80	.19**	−0.03	.20**	−0.02	.23**
7. Firm Size (LN_Employees)	9.50	1.42	−0.15*	−0.01	−0.00	.27**	−0.05	−0.22**
8. Firm Age (LN_Year)	3.36	.83	.15*	.05	.10	−0.10	−0.07	.23**	−0.06
9. R&D Intensity	4.65	5.51	.11	.10	.15*	.03	.32**	.15*	−0.11	−0.13*
10. Financial slack	24.69	71.67	−0.29**	−0.23**	−0.11	.11	−0.13*	−0.18**	.08	−0.08	−0.16**
11. Leverage	1.74	1.01	.30**	.17**	.07	−0.18**	−0.02	.27**	−0.42**	.10	.16*	−0.49**
12. IPO Age (Ln)	2.62	.80	.10	.09	.11	−0.05	−0.13*	.02	.04	.19**	−0.00	−0.04	.08
13. Industry (GICS Dummy)	21.97	13.53	.06	.14*	.03	−0.08	.22*	.09	−0.16*	.02	.33**	−0.08	.15*	.04
14. Industry Profitability (Industry ROA)	.05	.03	−0.05	.10	−0.09	−0.04	−0.17**	−0.73**	.19**	−0.19**	−0.10	−0.04	−0.00	−0.08	−0.14*
15. Individualism	51.39	21.33	.24**	.00	.26**	.06	.35**	.89**	−0.18**	.18**	.19**	−0.14*	.16*	−0.02	.02	−0.71**
16. Masculinity	57.15	27.60	−0.17**	−0.04	−0.18**	−0.21**	−0.27**	−0.15*	.13*	−0.09	−0.05	−0.13*	.23**	.12	.08	.34**	−0.45**
17. Uncertainty Avoidance	55.44	25.23	−0.02	−0.09	−0.05	−0.21**	−0.09	.62**	−0.08	.16**	.02	−0.17**	.31**	.11	.16**	−0.41**	.29**	.52**
18. Long Term Orientation	74.59	16.76	−0.23**	−0.05	−0.27**	−0.17**	−0.30**	−0.41**	.18**	−0.22**	−0.08	.00	.05	.12*	.08	.41**	−0.65**	.81**	.23**
19. Regulation	7.63	.78	.22**	.01	.23**	−0.12	.16*	.91**	−0.24**	.32**	.13*	−0.24**	.36**	.05	.10	−0.56**	.79**	.01	.58**	−0.32**
20. GDP Per Capita	40014.23	21263.96	.22**	.03	.26**	.04	.27**	.85**	−0.23**	.26**	.18**	−0.16**	.25**	.03	.08	−0.58**	.86**	−0.27**	.28**	−0.52**	.87**

*p<.05, **p < 0.01.

Table 2. Descriptive statistics of firms by industry and country.

Industry		Number of Firms
1	Communication Services	10
2	Consumer Discretionary	45
3	Consumer Staples	28
4	Energy	5
5	Financials	1
6	Health Care	27
7	Industrials	73
8	Information Technology	30
9	Materials	35
10	Real Estate	1
11	Utilities	7
Country		Individualism	Masculinity	Uncertainty Avoidance	Long Term Orientation	Regulation	GDP Per Capita (in USD)	NPE	Number of Firms
1	Austria	55	79	70	60	7.67	51486.6	4.41	6
2	China	20	66	30	87	6.42	9905.34	2.75	69
3	Denmark	74	16	23	35	8.5	61591.9	4.42	10
4	Finland	63	26	59	38	8	49988.9	4.80	10
5	France	71	43	86	63	7.58	41592.8	4.67	25
6	Germany	67	66	65	83	8.12	47973.6	4.73	45
7	Japan	46	95	92	88	8.36	39808.2	4.37	41
8	South Korea	18	39	85	100	7.47	33436.9	3.80	4
9	Sweden	71	5	29	53	7.81	54589.1	4.30	34
10	Switzerland	68	70	58	74	8.57	86388.4	4.75	18
	Total								262

Measures

There might be a lag in the effects of the variables in the study on performance, and the variation in the variables across the years exists. Research also finds that lagged models might have issues of simultaneous bias and auto-correlation problems (Ali Shah & Akbar, 2008). Thus, we calculated a numeric average of each firm- and industry-level variable based on the data for the years 2016-2018 as opposed to a one-year lagged model.

Dependent variable

In line with the current literature, performance was measured with ROA, ROS, and ROE (Semenov & Randrianasolo, 2022b). We included all three measures of performance to demonstrate the robustness of our results. We calculated a 3-year average for ROA, ROS, and ROE to address issues with a one-year lagged model. This method is also consistent with current research on the advertising/performance relationship that demonstrates that advertising has “carryover effects” on performance beyond the year in which advertising expenditures are incurred (Eng & Keh, 2007).

Independent variables

CSR was measured with CSRHub ratings (Randrianasolo & Semenov, 2022), and NPE was measured with the Global Philanthropy Environment Index. Advertising intensity was measured as advertising divided by total sales.

Control variables

We included firm age (natural logarithms of years), size (natural logarithms of employees), R&D intensity, leverage, financial slack, company listing, and industry dummy (GICS) as controls because they were previously linked to performance (Bıçakcıoğlu-Peynirci & Morgan, 2022; Saboo et al., 2017; Semenov & Randrianasolo, 2022b). We measured leverage as a debt-to-equity ratio, financial slack as the current ratio, and company listing as the natural log of the year of IPO.

Country-level variables can also influence firm performance. Thus, we measure country regulations as the regulation rating from the Economic Freedom of the World: 2019 Annual Report and level of economic development as GDP per capita from the World Bank: World Development Indicators.

The relationship between CSR and advertising intensity can be influenced by industry (industry, and industry profitability), firm (firm size and age), and country variables (culture, regulations, and GDP per capita). We measure industry profitability as the industry average ROA and culture as four of Hofstede’s cultural dimensions (individualism, masculinity, uncertainty avoidance, and long-term orientation). We did not include power distance and indulgence because they are theoretically unrelated to the CSR/advertising intensity relationship.

Results

Direct and Mediation Effects

We examined the direct effect of CSR on performance and an indirect effect through advertising using Hayes’ PROCESS Model 4 (Hayes, 2018) to test H1 (e.g., Trivedi et al., 2021), where CSR was the independent variable (X), performance was the dependent variable (Y), and advertising intensity was the mediator (M). The results (Table 3) show that the direct effect of CSR on advertising intensity (95% CI[−0.00 to 0.00], t = 0.45), and the direct effects of CSR (H1:95% CI[−0.10 to 0.11]/[−0.15 to 0.15]/[−0.15 to 0.12]) and advertising intensity (95% CI[−6.32 to 9.72]/[−17.23 to 5.65]/[−1.42 to 19.70]) on performance (ROA/ROS/ROE) are not significant. The indirect effect of CSR→Advertising Intensity→Performance is not significant for all three measures of performance, specifically, ROA (95% CI[−0.01 to 0.02]), ROS (95% CI[−0.02 to 0.03]), and ROE (95% CI[−0.01 to 0.03]). Thus, H1 is rejected suggesting that advertising intensity does not mediate the relationship between CSR and performance.

Table 3. Results from PROCESS analyses in SPSS (model 4), n = 262 for ROA/ROS/ROE as outcomes.

					CI (95%)
PROCESS MODEL 4	Coefficient	Standard Error	t-value	p-value	Lower	Upper	Hypotheses
Direct and Total Effects
CSR→Advertising Intensity	.00/.00/.00	.00/.00/.00	.45/.45/.45	.65/.65/.65	−0.00/−0.00/−0.00	.00/.00/.00
Advertising Intensity→Performance	1.69/−5.78/9.13	4.07/5.80/5.36	.41/−0.99/1.70	.68/.32/.08	−6.32/−17.23/−1.42	9.72/5.65/19.70
CSR→Performance	.00/−0.00/−0.01	.05/.07/.07	.07/−0.01/−0.18	.94/.98/.85	−0.10/−0.15/−0.15	.11/.15/.12
Indirect Effects
CSR→Advertising Intensity→Performance	.00/−0.00/.00	.01/.01/.01			−0.01/−0.02/−0.01	.02/.03/.03	H1: Not Supported
Control Variables
Industry (GICS Dummy)	−0.00/.08/−0.03	.03/.04/.04	−0.22/1.89/−0.70	.82/.06/.48	−0.07/−0.00/−0.11	.06/.18/.05
Firm Size (LN_Employees)	−0.18/.35/.35	.33/.47/.43	−0.53/.74/.80	.60/.45/.42	−0.83/−0.58/−0.51	.48/1.28/1.21
Firm Age (LN_Year)	.71/.48/.47	.52/.75/.69	1.36/.65/.69	.17/.52/.49	−0.32/−0.99/−0.88	1.74/1.96/1.84
Leverage	1.13/1.1/−0.05	.52/.75/.69	2.17/1.35/−0.08	.03/.17/.93	.10/−0.46/−1.42	2.16/2.50/1.31
Financial slack	−0.02/−0.03/−0.01	.01/.01/.01	−2.66/−2.71/−0.67	.01/.01/.49	−0.03/−0.04/−0.02	−0.00/−0.01/.01
IPO Age (Ln)	.55/.70/1.19	.51/.73/.68	1.07/.95/1.74	.28/.34/.08	−0.46/−0.75/−0.15	1.56/2.16/2.53
R&D Intensity	.06/.05/.15	.08/.11/.10	.75/.49/1.41	.45/.62/.15	−0.10/−0.17/−0.06	.22/.29/.37
Regulation	−0.63/−3.21/.19	1.16/1.67/1.54	−0.54/−1.92/.12	.58/.06/.90	−2.92/−6.52/−2.86	1.66/.09/3.24
GDP Per Capita	.00/.00/.00	.00/.00/.00	1.34/1.52/1.43	.18/.12/.15	.00/.00/.00	.00/.00/.00

NPE as a Moderator

We used hierarchical linear modeling (HLM) to test H2 and H3 because the outcome variable is at the firm level and predictor variables are at both the firm and country levels of analyses (Qiu & Homer, 2018). HLM is an appropriate technique here for several reasons. While ordinary least squares (OLS) utilize a single-level approach, HLM adopts a two-level approach. OLS regression techniques assume the independence, normal distribution, and constant variance of the random errors. However, this assumption is likely violated when utilizing nested data (Bryk & Raudenbush, 1992). In the OLS approach, all the regression parameters are fixed and level-2 variance components are not separable from the individual level residuals. However, in HLM, the level-1 parameters are allowed to vary across groups and the variance and covariance of the level-2 residuals are also estimated (Bryk & Raudenbush, 1992). In addition to measuring fixed effects, HLM estimates the random effects of the intercepts and slopes in a model. Specifically, firm (level-1) variables (performance, CSR, and advertising intensity) were nested within the country (level-2) variable (NPE). However, before testing the hypotheses, we needed to demonstrate the appropriateness of HLM by demonstrating that variance between firm advertising intensity as well as variance between firm performance exist.

HLM Results for a Cross-Level Interaction between CSR and NPE on Advertising Intensity

An unconstrained (null) model with no predictors was estimated and showed the level-2 residual variance of the intercept (${\tau }_{00}$) of .00 (χ²(9)=67.85, p<.000) and an ICC1 of .21, suggesting that 21% of the variance in advertising intensity is at a country level (level-2) and 79% is at a firm level (level-1). The significance of chi-square indicates that there is variance in advertising intensity by the country grouping, thus justifying the use of HLM. The following steps were taken to test the relationships: first, all firm-level variables were introduced; next, country-level variables were added; and finally, a cross-level interaction between CSR and NPE was introduced. Based on prior research (e.g., Preacher et al., 2006), firm-level variables (CSR, firm age, and firm size) and industry profitability, were group-mean-centered. The industry dummy variable was not centered. The variables at the country level (NPE, individualism, masculinity, uncertainty avoidance, long-term orientation, regulation, and GDP per capita) were grand-mean-centered. The model below was used to test the relationships.

Level 1 Model (Firm-Level): $$T\mathit{\text{cos}}\hspace{0.17em}{\theta }_{\text{BP}}=T_n,\hspace{0.17em}T\mathit{\text{sin}}\hspace{0.17em}{\theta }_{\text{BP}}=T_t$$

Level 2 Model (Country Level): $${\beta }_{0j}={\gamma }_{00}+{\gamma }_{01}\left(\mathit{\text{National Philanthropic Environments}}\right)+{{\gamma }_{02}\left(\mathit{\text{Individualism}}\right)+{\gamma }_{03}\left(\mathit{\text{Masculinity}}\right)+{\gamma }_{04}\left(\mathit{\text{Uncertainty Avoidance}}\right)+{\gamma }_{05}\left(\mathit{\text{Long Term Orientation}}\right)+{\gamma }_{06}\left(\mathit{\text{Regulation}}\right)+{\gamma }_{07}\left(\mathit{\text{GDP Per Capita}}\right)+u}_{0j}$$ $${\beta }_{1j}={\gamma }_{10}+{\gamma }_{11}\left(\mathit{\text{National Philanthropic Environments}}\right)+u_{1j}$$ $${\beta }_{2j}={\gamma }_{20}+u_{2j}$$ $${\beta }_{3j}={\gamma }_{30}+u_{3j}$$ $${\beta }_{4j}={\gamma }_{40}+u_{4j}$$ $${\beta }_{5j}={\gamma }_{50}+u_{5j}$$

Table 4 displays the results of HLM analyses. Firm and country main effects are presented in the second and third sections of the table. At the firm level, only industry significantly and positively affects advertising intensity (γ = 0.00, p<.01). None of the country-level variables is significantly related to advertising intensity.

Table 4. HLM results for advertising intensity as an outcome variable.

Variables	Advertising Intensity^a
	Coefficient	S.E.	t	d.f.	R^2b	Hypotheses
Null Model
Intercept	0.11***	0.01	6.47	9
Level 1 (Firm-Level)					0.12
Intercept	0.06*	0.01	4.39	2
Industry (GICS Dummy)	0.00**	0.00	4.25	9
Industry Profitability (Industry ROA)	0.25	0.44	0.56	9
Firm Size (LN_Employees)	0.00	0.00	0.23	9
Firm Age (LN_Year)	0.00	0.01	0.07	9
Corporate Social Responsibility	0.00	0.00	0.39	8
Level 2 (Country-Level)					0.29
National Philanthropic Environment	0.05	0.06	0.78	2
Individualism	0.00	0.00	1.86	2
Masculinity	0.00	0.00	0.76	2
Uncertainty Avoidance	−0.00	0.00	−1.93	2
Long Term Orientation	−0.00	0.00	−0.16	2
Regulation	−0.02	0.03	−0.73	2
GDP Per Capita	−0.00	0.00	−0.88	2
Level 1 × Level 2 (Cross-Level Interaction)
National Philanthropic Environment × Corporate Social Responsibility	0.01*	0.00	2.42	8		H2: Supported

Note: Restricted Maximum Likelihood Estimation.

^aFirm-Level n = 262, Country-Level n = 10.

^bIndicates the proportion of variance extracted at each level; i.e., level-1 within-country variance, level-2 between-country variance, and cross-level interaction.

$R_{\mathit{\text{total}}}^2=R_{\mathit{\text{within}}-\mathit{\text{group}}}^2\times (1-\mathit{\text{ICC}}1)+R_{\mathit{\text{between}}-\mathit{\text{groups}}}^2\times \mathit{\text{ICC}}1$.

$R_{\mathit{\text{Advertising total}}}^2$=0.04.

*p<.05; **p<.01; ***p<.001.

To assess the effects of the cross-level interaction between CSR and national philanthropic environments, a slope-as-outcome model was estimated. The cross-level interaction was significant (γ = 0.01, p < 0.05). Figure 2 illustrates the pattern of the significant interaction. The values for the end points of CSR for Figure 2 were selected at one standard deviation above and below the mean (Aiken & West, 1991). The results demonstrate that in countries with greater (lower) NPE, there is a positive (negative) relationship between CSR and advertising intensity; therefore, supporting H2.

Figure 2. Moderating effects of NPE in the CSR/advertising intensity relationship.

HLM Results for a Cross-Level Interaction between Advertising Intensity and NPE on Performance

An unconstrained (null) model with no predictors was estimated for ROA/ROS/ROE and showed the level-2 residual variance of the intercept (${\tau }_{00}$) of 6.08 (χ²(9)=30.99, p<.000)/6.69 (χ²(9)=31.89, p<.000)/12.31 (χ²(9)=47.38, p<.000) and an ICC1 of .09/.09/21¹, suggesting that 9%/9%/21% of the variance in performance is at a country level (level-2) and 91%/91%/79% is at a firm level (level-1). The significance of chi-square indicates that there is variance in performance by the country grouping, thus justifying the use of HLM. The following steps were taken to test the relationships: first, all firm-level variables were introduced; next, the country-level variable was added; and finally, a cross-level interaction between advertising intensity and NPE was introduced. Firm-level variables (advertising intensity, firm age, firm size, leverage, financial slack, IPO Age, and R&D Intensity) were group-mean-centered. The industry variable was not centered. The country-level variables (regulation, GDP per capita, and NPE) were grand-mean-centered. The model below was used to test the relationships.

Level 1 Model (Firm-Level): $${\mathit{\text{Performance}}}_{(\mathit{\text{ROA}}/\mathit{\text{ROS}}/\mathit{\text{ROE}})\mathit{\text{Firm}}}={\beta }_{0j}+{\beta }_{1j}\left(\mathit{\text{Advertising Intensity}}\right)+{\beta }_{2j}\left(\mathit{\text{Industry Code}}\right)+{\beta }_{3j}\left(\mathit{\text{Firm Size}}\right)+{\beta }_{4j}\left(\mathit{\text{Firm Age}}\right){+{\beta }_{5j}\left(\mathit{\text{Leverage}}\right)+{\beta }_{6j}\left(\mathit{\text{Financial slack}}\right)+{\beta }_{7j}\left(\mathit{\text{IPO Age}}\right)+{\beta }_{8j}(R\&D\mathrm{ }\mathit{\text{Intensity}})+r}_{\mathit{\text{ij}}}$$

Level 2 Model (Country Level): $${\beta }_{0j}={\gamma }_{00}+{\gamma }_{01}\left(\mathit{\text{National Philanthropic Environments}}\right)+{\gamma }_{02}\left(\mathit{\text{Regulation}}\right)+{\gamma }_{03}\left(\mathit{\text{GDP Per Capita}}\right)+u_{0j}$$ $${\beta }_{1j}={\gamma }_{10}+{\gamma }_{11}\left(\mathit{\text{National Philanthropic Environments}}\right)+u_{1j}$$ $${\beta }_2={\gamma }_{20}+u_{2j}$$ $${\beta }_{3j}={\gamma }_{30}+u_{3j}$$ $${\beta }_{4j}={\gamma }_{40}+u_{4j}$$ $${\beta }_{5j}={\gamma }_{40}+u_{4j}$$ $${\beta }_{6j}={\gamma }_{40}+u_{4j}$$ $${\beta }_{7j}={\gamma }_{40}+u_{4j}$$ $${\beta }_{8j}={\gamma }_{40}+u_{4j}$$

Table 5 displays the results. The firm and country’s main effects are presented in the second and third sections of the table. At the firm level, only leverage is significantly related to ROA (γ = 4.68, p < 0.05) and ROS (γ = 3.66, p < 0.05). At the country level, only regulation was significantly related to ROA (γ = 3.42, p < 0.05) performance.

Table 5. HLM results for performance as an outcome variable.

Variables	Performance (ROA/ROS/ROE)^a
	Coefficient	S.E.	t	d.f.	R^2b	Hypotheses
Null Model
Intercept	6.08***/6.69***/12.31***	0.83/0.59/1.52	7.31/11.34/8.08	9
Level 1 (Firm-Level)					0.43/0.41/0.40
Intercept	5.16***/4.89**/10.72***	0.80/1.23/1.43	6.38/3.95/7.49	6
Industry (GICS Dummy)	0.05/0.09/0.09	0.04/0.05/0.09	1.29/1.68/0.98	9
Firm Size (LN_Employees)	1.99/1.43/3.62	1.46/0.85/2.76	1.35/1.67/1.31	9
Firm Age (LN_Year)	0.22/0.13/−0.14	0.72/1.16/0.95	0.31/0.11/−0.14	9
Leverage	4.68*/3.66*/5.20	1.52/1.50/3.74	3.06/2.43/1.39	9
Financial slack	−0.00/−0.02/0.01	0.01/0.02/0.02	−0.09/−1.10/0.64	9
IPO Age (Ln)	−0.05/−0.17/0.17	0.78/0.93/0.95	−0.07/−0.18/0.18	9
R&D Intensity	0.21/0.13/0.49	0.23/0.18/0.40	0.94/0.74/1.24	9
Advertising Intensity	−1.73/5.42/−7.40	8.50/10.28/19.42	−0.20/0.52/−0.38	8
Level 2 (Country-Level)					0.12/0.33/0.16
Regulation	3.42*/2.36/2.95	1.34/1.78/2.08	2.55/1.32/1.41	6
GDP Per Capita	0.00/0.00/0.00	0.00/0.00/0.00	0.03/1.78/0.51	6
National Philanthropic Environment	−1.05/−4.43*/−2.01	1.28/1.65/1.81	−0.82/−2.67/−1.11	6
Level 1 × Level 2 (Cross-Level Interaction)
National Philanthropic Environment × Advertising Intensity	8.54*/21.65*/13.39*	3.61/6.74/5.20	2.36/3.21/2.57	8		H3: Supported

Note: Restricted Maximum Likelihood Estimation.

^aFirm-Level n = 262, Country-Level n = 10.

^bIndicates the proportion of variance extracted at each level; i.e., level-1 within-country variance, level-2 between-country variance, and cross-level interaction.

$R_{\mathit{\text{total}}}^2=R_{\mathit{\text{within}}-\mathit{\text{group}}}^2\times (1-\mathit{\text{ICC}}1)+R_{\mathit{\text{between}}-\mathit{\text{groups}}}^2\times \mathit{\text{ICC}}1$.

$R_{\mathit{\text{Performance total}}}^2$=0.39/0.41/0.28.

**p<.01; ***p<.001.

To assess the effects of the cross-level interactions between advertising intensity and NPE, a slope-as-outcome model was estimated. The cross-level interaction was significant for ROA (γ = 8.54, p < 0.05), ROS (γ = 21.65, p < 0.05), and ROE (γ = 13.39, p < 0.05) supporting H3. Figure 3 illustrates the patterns of the significant interactions.

Figure 3. Moderating effects of NPE in the advertising intensity/performance relationship.

Moderated Mediation

We examined a mediated moderation (Figure 1) using Hayes’ PROCESS Model 58 (Hayes, 2018) to test H4, where CSR was the independent variable (X), performance (ROA/ROS/ROE) was the dependent variable (Y), advertising intensity was the mediator (M), and NPE was the moderator (W) for the CSR→advertising intensity as well as advertising intensity→performance relationships. The results (Table 6) show that CSR alone does not influence advertising intensity (95% CI[0.00 to 0.00]). Neither CSR (95% CI[-0.12 to 0.09]/[-0.18 to 0.11]) nor advertising intensity (95% CI[-4.08 to 12.17]/[-10.91 to 11.51]) influence performance (ROA/ROS). While CSR (95% CI[-0.17 to 0.11]) does not influence performance (ROE), advertising intensity (95% CI[2.57 to 23.86]) positively and significantly influences performance (ROE).

Table 6. Results from PROCESS analyses in SPSS (model 58), n = 262.

					CI (95%)
PROCESS MODEL 58	Coefficient	Standard Error	t-value	p-value	Lower	Upper	Hypotheses
Outcome Variable: Advertising Intensity
CSR	.00	.00	.14	.89	.00	.00
NPE	.04	.02	2.00	.05	.00	.08
CSR x NPE	.01	.00	1.96	.05	.00	.01	H4: Supported
Control Variables
Industry (GICS Dummy)	.00	.00	3.01	.00	.00	.00
Firm Size (LN_Employees)	.00	.01	−0.38	.71	−0.01	.01
Firm Age (LN_Year)	−0.01	.01	−0.75	.45	−0.02	.01
Leverage	−0.02	.01	−2.08	.04	−0.03	.00
Financial slack	.00	.00	−2.42	.02	.00	.00
IPO Age (Ln)	−0.01	.01	−1.80	.07	−0.03	.00
R&D Intensity	.00	.00	3.03	.00	.00	.01
Regulation	−0.07	.03	−2.80	.01	−0.12	−0.02
GDP Per Capita	.00	.00	3.20	.00	.00	.00
Model fit summary
F(sig.)	7.41(.00)
R^2	.26
ΔR^2(moderation included)	.01
ΔF(moderation included)	3.87(.05)
f^2	.01
					CI (95%)
PROCESS MODEL 58	Coefficient	Standard Error	t-value	p-value	Lower	Upper	Hypotheses
Outcome Variable: Performance (ROA)
CSR	−0.01	.05	−0.27	.79	−0.12	.09
Advertising Intensity	4.05	4.12	.98	.33	−4.08	12.17
NPE	−0.04	1.36	−0.29	.77	−3.08	2.28
Advertising Intensity x NPE	12.54	4.19	2.99	.00	4.28	20.79	H4: Supported
Control Variables
Industry (GICS Dummy)	−0.01	.03	−0.36	.72	−0.07	.05
Firm Size (LN_Employees)	−0.21	.33	−0.64	.53	−0.85	.44
Firm Age (LN_Year)	.81	.53	1.53	.13	−0.23	1.84
Leverage	1.07	.52	2.05	.04	.04	2.10
Financial slack	−0.02	.01	−2.90	.00	−0.03	−0.01
IPO Age (Ln)	.43	.51	.85	.40	−0.58	1.45
R&D Intensity	.06	.08	.76	.45	−0.10	.22
Regulation	.20	1.64	.12	.90	−3.03	3.44
GDP Per Capita	.00	.00	.93	.36	.00	.00
Model fit summary
F(sig.)	4.29(.00)
R^2	.18
ΔR^2(moderation included)	.03
ΔF(moderation included)	8.95(.00)
f^2	.04
					CI (95%)
PROCESS MODEL 58	Coefficient	Standard Error	t-value	p-value	Lower	Upper	Hypotheses
Outcome Variable: Performance (ROS)
CSR	−0.04	.07	−0.48	.63	−0.18	.11
Advertising Intensity	.30	5.69	.05	.96	−10.91	11.51
NPE	−0.2.55	1.88	−1.36	.18	−6.25	1.15
Advertising Intensity x NPE	28.81	5.78	4.98	.00	17.42	40.20	H4: Supported
Control Variables
Industry (GICS Dummy)	.08	.04	1.84	.07	−0.01	.17
Firm Size (LN_Employees)	.25	.45	.55	.58	−0.64	1.14
Firm Age (LN_Year)	.58	. 73	.80	.42	−0.85	2.01
Leverage	.75	.72	1.05	.30	−0.67	2.18
Financial slack	−0.03	.01	−3.33	.00	−0.05	−0.01
IPO Age (Ln)	.38	.71	.53	.60	−1.02	1.77
R&D Intensity	.08	.11	.69	.49	−0.14	.30
Regulation	−0.07	2.26	−0.03	.98	−4.53	4.39
GDP Per Capita	.00	.00	1.29	.20	.00	.00
Model fit summary
F(sig.)	4.21(.00)
R^2	.18
ΔR^2(moderation included)	.08
ΔF(moderation included)	24.82(.00)
f^2	.10
Outcome Variable: Performance (ROE)
CSR	−03	.07	−0.38	.71	−0.17	.11
Advertising Intensity	13.21	5.40	2.45	.02	2.57	23.86
NPE	−2.77	1.78	−1.55	.12	−6.28	.75
Advertising Intensity x NPE	16.78	5.49	3.06	.00	5.97	27.60	H4: Supported
Control Variables
Industry (GICS Dummy)	−0.03	.04	−0.83	.41	−0.12	.05
Firm Size (LN_Employees)	.27	.43	.64	.52	−0.57	1.12
Firm Age (LN_Year)	.46	.69	.67	.50	−0.89	1.82
Leverage	−0.25	.69	−0.37	.71	−1.60	1.10
Financial slack	−0.01	.01	−0.96	.34	−0.03	.01
IPO Age (Ln)	.99	.67	1.48	.14	−0.33	2.32
R&D Intensity	.17	.11	1.55	.12	−0.04	.38
Regulation	3.11	2.15	1.45	.15	−1.13	7.34
GDP Per Capita	.00	.00	1.39	.17	.00	.00
Model fit summary
F(sig.)	3.35(.00)
R^2	.15
ΔR^2(moderation included)	.03
ΔF(moderation included)	9.35(.00)
f^2	.04

NPE moderates the relationship between CSR and adverting intensity (95% CI[0.00 to 0.01]) and adverting intensity and performance (ROA/ROS/ROE) (95% CI[4.28 to 20.79]/[17.42 to 40.20]/[5.97 to 27.60]). As shown in Figures 2 and 3, in countries with greater NPE, CSR is positively related to advertising intensity, and advertising intensity is positively related to performance. However, in countries with lower NPE, CSR is negatively related to advertising intensity, and advertising intensity is negatively related to performance. Moreover, the relationships between CSR, advertising intensity, and performance are significant only when NPE is included as a moderator. These results support H4.

Discussion

The results provide interesting findings. Notably, H1 was not supported, indicating that advertising intensity does not mediate the relationship between CSR and performance. However, NPE was found to moderate the relationships between CSR/advertising intensity (H2) and advertising intensity/performance (H3). Further, the moderated mediation model (H4) was supported. The novelty in the results of these findings is twofold. First, the findings support the notion that market contexts matter in the question of where to employ CSR. Particularly, much of the prior work on CSR in international marketing contexts proposes that CSR is a legitimacy-enhancing firm strategy but fails to consider that this may not be generalizable in all markets (e.g., Campbell et al., 2012; Randrianasolo, 2018). Our findings indicate that CSR indirectly influences performance in higher, rather than lower, NPE countries, and thus may be employed as a more effective legitimacy-seeking strategy in such countries. This finding answers the “where” question. Second, the findings indicate that for CSR to be effective in high NPE countries, it should be complemented with advertising intensity. Thus, simply employing CSR to gain legitimacy is not enough as firms should increase advertising intensity to legitimacy-granting constituents in the institutional environment to reap the fruits of legitimacy. This finding answers the “how” question. The theoretical contributions and managerial implications of these findings are further discussed below.

Theoretical Contributions

We contribute to the marketing and management research literature by offering new insights into the relationship between CSR, advertising intensity, and performance. Our findings deviate from prior research that has demonstrated the positive effects of CSR and advertising intensity on performance (Maury, 2022), as we find no significant mediation of advertising intensity in the relationship between CSR and performance. We posit that the non-significance of this relationship is due to our sample being collected from different countries, indicating that this mediation is not consistent in every institutional environment. In other words, within some countries, advertising intensity may mediate the link between CSR and performance, and it may not in others. The mediation model (without the moderation of NPE) therefore, is not sufficient to explain the influence of CSR and advertising intensity on performance.

From this finding, we encourage future research to refrain from generalizing about CSR, advertising intensity, and performance with empirical results from single-country studies. This finding also supports the notion that CSR is inherently a socially based strategy, and its effects are dependent on environmental factors, such as NPE, as further discussed below.

Our second contribution is that CSR may only be an effective legitimacy-enhancing strategy in higher NPE countries. This finding challenges the notion that CSR is universally applicable for gaining legitimacy, which is a key premise of institutional theory-based CSR research (Campbell et al., 2012). Our study provides support for the contextual nature of CSR and advances international marketing research by suggesting that CSR may indeed confer legitimacy in foreign markets and reduce the liability of foreignness (Campbell et al., 2012), but its effectiveness may be limited to higher NPE countries.

Third, the finding that CSR should be aligned with advertising intensity to enhance performance in high NPE countries holds significant implications for marketing and advertising research strategy literature. This implies that not only should CSR efforts be complemented by advertising intensity to boost performance, but this effect may only be applicable in high NPE countries. Our analyses reveal that while the mediation of advertising intensity in the CSR/performance relationship is not significant, the moderation of NPE on the links between CSR and advertising intensity, as well as advertising intensity and performance, is significant, thus supporting the moderated mediation model.

Finally, our findings extend the findings of Semenov and Randrianasolo (2022a), which provide evidence that NPE moderates the relationship between CSR and advertising intensity. We extend this knowledge by providing support for a model that not only proposes a relationship between CSR and advertising intensity, moderated by NPE, but also that advertising intensity mediates the relationship between CSR and performance, and that this mediation is moderated by NPE. This model enhances our understanding of the interplay between CSR, advertising intensity, performance, and NPE, and provides valuable managerial implications, as discussed below.

Managerial Implications

The findings of this study, particularly the supported moderated mediation model, offer strategic implications for marketing managers. Firstly, managers operating in high NPE countries may effectively employ CSR strategies and complement them with advertising to gain legitimacy and positively influence performance. This strategic framework is particularly relevant for international business and marketing managers concerned with reducing liabilities of foreignness and newness, as legitimacy-enhancing strategies have been found to alleviate both (Campbell et al., 2012; Singh et al., 1986). Overcoming the liabilities of foreignees and newness is imperative for firms to survive and thrive within their respective institutional environments (Campbell et al., 2012; Singh et al., 1986), and prior research indicates that CSR advertising is one method to gain legitimacy to overcome both liabilities (Du & Vieira, 2012; Randrianasolo, 2018). The results of this current research indicate that CSR advertising indeed may be an effective strategy to gain legitimacy, overcome liabilities of foreignness/newness, and boost performance, but only in high NPE countries.

Conversely, managers in low NPE countries may find it ineffective to rely solely on CSR to gain legitimacy, as our results suggest. Therefore, it may be more beneficial for such managers to explore alternative legitimacy-seeking strategies, such as establishing political ties (Bai et al., 2019), employing local workforces (Forstenlechner & Mellahi, 2011), or implementing enviropreneurial marketing strategies (Randrianasolo, 2021). However, further research is necessary in this domain, and we encourage scholars to investigate suitable legitimacy-seeking strategies in low NPE countries. Specifically, firms must gain legitimacy in their respective institutional environments to survive (Campbell et al., 2012), thus we encourage future research to investigate non-CSR legitimacy-gaining strategies for low NPE countries.

Additionally, beyond implications for international business and marketing managers, our findings hold relevance for policymakers within their respective nations. Policymakers and governments may find it advantageous to enact legislation that promotes philanthropy within the institutional environment of their markets, thereby fostering higher levels of NPE. By enhancing NPE, policymakers can incentivize companies to invest in CSR efforts, as such initiatives are valued and expected by stakeholders. This, in turn, may lead to improved company performance, thereby benefiting the government’s revenue. This point here is that by advancing policies and regulations that promote philanthropy and enhance NPE, governments may be able to influence CSR, and ultimately firm performance within their markets, thus yielding positive gains in economic development.

Limitations and Further Research

While our study rigorously examines secondary data from various sources, it is subject to limitations that present opportunities for future research. Recent studies have highlighted the significant role of industry in CSR research (Randrianasolo & Semenov, 2022). Although we controlled for industry effects, we did not explicitly hypothesize the influence of industry on the relationships examined. Future research could expand on our work by exploring these relationships within specific industry contexts.

Furthermore, our current study focuses on firms operating within their institutional environments. Future studies may extend our work by investigating the dynamics of CSR/advertising/performance in the context of internationalizing firms, with particular attention to how home and host country institutional environments (NPE) influence these dynamics.

Although our sample includes firms from ten different countries, it predominantly represents developed economies, with only a few emerging economies included. We encourage future studies to investigate our model using a sample that is more representative of developing countries.

Lastly, while we captured advertising intensity, firm performance, and NPE in the country of operations (host country), we did not account for advertising spending by each firm in the home or other host countries. Future research may explore whether advertising intensity in the home and other host countries influences firm performance in a given country of operations.

Conclusion

The relationships among CSR, advertising intensity, and performance have received extensive attention in business research. However, the findings from these investigations have produced inconsistent and inconclusive outcomes, often due to the reliance on single-country samples. To address this limitation, our study takes a different approach by recognizing the need to consider market-specific dynamics. Drawing on institutional theory, we propose and empirically validate a moderated mediation model that encompasses CSR, advertising intensity, NPE, and performance. The findings offer valuable strategic insights for marketing managers, especially those engaged in international marketing, by highlighting that the impact of CSR on performance is contingent upon both high NPE countries and the presence of complementary advertising intensity. This paper contributes to the existing body of knowledge by advancing our understanding of when and how CSR should be employed as a strategic tool to enhance performance and what role advertising intensity plays in this dynamic.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 ICC > 0, even as small as .10 (Kahn, 2011), suggests that there might be a firm level variable that explains heterogeneity of advertising inensity across the firms.