HAPPINESS AND WEALTH: ATISFACTION, WELL-BEING, AND ROA RELATIONSHIP

We investigated happiness at work reflected in the indexes presented by the 100 best companies to work for in the United States, as well as its impact on Return on Assets (ROA). Indicators used by Glassdoor were collected, as well as financial data from the Eikon. We performed a confirmatory factor analysis (CFA) to obtain the Happiness at Work General Index. The results were applied to regression to understand the relationship between the index and the ROA. Happiness at Work General Index and ROA are positively related, providing evidence that happiness at work positively impacts the operational performance of companies.


Introduction
Work and happiness are two aspects of nature that have long been objects of observation by man. During ancient history, happiness was superficially approached in the field of philosophy, being considered a good motivating human action. It was only around the 1960s that researches and indexes on happiness and social indicators emerged that led scholars, mainly in the field of psychology, to study and systematize in-depth the elements and factors that lead people to be happy (Diener, 1984).
Our study was based on the Huang, Li, Meschke, and Guthrie (2015) analyses, in which the authors propose to answer if there is an association between job satisfaction and companies' operational performance. Originally, the authors proposed to assess the role of corporate culture in family businesses and its implications for company value. To measure job satisfaction, employees' perceptions about organizational culture are used, as assessed by the Glassdoor website. The reason for adopting this metric is explained by the authors pointing out that corporate culture is an unobservable phenomenon in the field of Econometrics and that previous researches such as Edmans (2011Edmans ( , 2012 and Guiso et al. (2015) used data from the Great Place to Work Institute to assess The 100 Best Companies to Work For. Thus, aiming to innovate in studies in this field, the use of Glassdoor's assessments presents itself as a new set of data to examine job satisfaction. The dataset contains indicators such as overall satisfaction, career development, compensation and benefits, work/life balance, senior management, and CEO approval.
Once in psychometrics, the assessment of variables such as satisfaction, well-being, and many others of organizational behavior are measured through instruments that assess the individual perception of employees, at the micro-level of analysis. In turn, operational performance measures are evaluated at the macro-level, which would reduce the effectiveness of relational analyzes between these variables, given that they are not located at the same level of analysis. However, in organizational studies, there is a tendency to understand the aggregation of lower-level variables as a reliable representation of an aggregated higher-level variable, called composition. Given that Glassdoor's indicators represent the combination of employees' assessments, and that a fundamental requirement for aggregating perceptions at the organizational level is the agreement of perceptions among the participants, then it makes sense to consider them, in this case, to compare with the operational performance, as both deal with the organizational level (Klein & Kozlowski, 2000). Therefore, the objective of this paper is to investigate the happiness at work variable reflected in the wellbeing, and job satisfaction indexes presented by the 100 best companies to work for in the USA, as well as its impact on operational performance.
When studying organizations from an organizational behavior perspective, there is a lack of concrete answers about the role of happiness at work and its effects on organizations (Sender & Fleck, 2017). From conceptual aspects and components of the construct to its practical applications in the daily routine of organizations, the theme still presents several research gaps and is of great interest for the evolution of scientific knowledge in organizational studies. Cropanzano and Wright (2001) signaled that research dealing with happiness at work used four different approaches, based on organizational behavior constructs, namely: a) job satisfaction, b) positive and negative affects, c) absence of exhaustion emotional and, d) psychological well-being. It is noteworthy that psychological well-being in the Panaccio and Vandenberghe (2009) view is defined as the presence of positive affect, the absence of negative affect, and the presence of satisfaction, which is very similar to the understanding of happiness that has been applied in organizational studies. According to Loon, Otaye-Ebede, and Stewart (2019), there is theoretical and empirical evidence that demonstrates that a good way to promote successful organizational performance is to promote psychological well-being.
The Glassdoor indexes as a metric to assess employee happiness, expressing whether this is a good company to work for -overall satisfaction, career development, compensation and benefits, work/life balance, senior management, and CEO approval -comprise related factors to the concepts of well-being at work and job satisfaction. Thus, presenting the concept of happiness by joining the aspects of these two concepts is of great academic relevance, as it corroborates the approach to happiness that is widely used in the literature, such as the studies by Fisher (2010), and Fogaça and Coelho Junior (2016), among others. Furthermore, happiness is observed empirically in companies from different sectors, to make the analyzes applicable to different realities.
Following the research agenda such as Sender andFleck (2017), Nangov, Sasmoko andIndrianti (2018), and Nangoy, Mursitama, Setiadi, and Pradipto (2020), this paper understands the need to understand happiness as a strategic force that affects organizational performance. The managerial implications mainly concern support for decision-making in human resources management, cost savings, better organizational climate, business policies, and the company's strategic actions. According to Nangoy et al. (2020Nangoy et al. ( , p. 1041, the increase in work performance mediated by well-being can even impact the social sphere by "playing a role at the microeconomic level, even at the level on how to produce self-sustaining spirals between human productivity and human well-being".

Theoretical Review
Human resources management points to the relationship between performance and happiness as one of its most relevant issues, which although studied for some time, is still a promising and challenging issue (Guerci, Gilardi & Hauff, 2019;Hauff, Guerci & Gilardi, 2020;Warr & Nielsen, 2018). It is common to come across, both in the literature and in the business context, the assertion that satisfied and happy employees will be more productive than others (Taris & Schaufeli, 2015). Other studies also indicate that satisfied employees who work in customer service tend to retain customers, as they tend to be more attentive, helpful, happy, and kind, delighting the customer (Scussel & Fogaça, 2017;Woisetschläger, Hanning & Backhaus, 2016). Fogaça and Coelho Júnior (2016), when investigating the hypothesis "happy, productive worker", observed that well-being helps to achieve better individual job performance and that when employees feel satisfied with the activity performed, the work flows, there is a greater willingness and, consequently, better performance. The authors also found that job satisfaction also plays a role in explaining job performance, supporting the idea that well-being at work and job satisfaction can be interpreted as happiness at work. This relationship was also identified by Nangoy et al. (2020), who highlighted the implementation of objective actions, organizational engineering, and considerable resources in promoting well-being have effects on the organization as a whole and are long-term.

Well-being at work
Following the studies of positive psychology, Siqueira and Padovam (2008) observe that well-being at work is psychological well-being linked to positive affective bonds with work and the organization in which the individual is inserted. Therefore, Paz (2004) presents the concept based on the more cognitive bias of the subject. According to the author, well-being at work involves "the satisfaction of needs and the fulfillment of desires of individuals in the performance of their organizational role" (Paz, 2004, p. 138), that is, a process that considers the cognitive and the affective aspect of well-being and encompasses situational aspects inherent to well-being (Paschoal & Tamayo, 2008).
Tatiane Paschoal and Álvaro Tamayo (2008) systematized the consolidated literature to build and validate an objective measure of well-being at work. According to the authors, and following the line of researchers such as Ryan and Deci (2001), the construct is formed by an affective dimension (emotions and moods at work), and a cognitive dimension (perception of expressiveness and personal fulfillment at work). These dimensions, together, encompass both hedonic and eudaimonic elements of well-being, defining it as "the prevalence of positive emotions at work and the individual's perception that, in their work, they express and develop their potentials/ skills and advances in reaching their life goals" (Paschoal & Tamayo, 2008, p. 6). Martins, Siqueira, and Souza (2015) realized that organizational support actions tend to generate higher levels of well-being at work, as there is a perception that the company cares about the employee. This is also perceived by other institutional attitudes, such as fair rewards for their responsibility, experience, efforts, and quality of work (Loila, Alves & Siqueira, 2017). Consequently, the opposite also occurs. Negative actions, such as negligent attitudes towards harassment, leading to a decrease in employee satisfaction, commitment, and even health (Jacoby & Monteiro, 2016). Vasconcellos and Neiva (2016) conclude that the improvement of HR policies and practices can increase well-being at work and expectations for the organizational future, in addition to retaining talent, maintaining the quality of work, and stimulating career expectations. In this context, the studies understand that programs for the continuous promotion of new interesting learning for the employee, encouragement and health care, and, mainly, greater social inclusion in the work environment lead to higher levels of well-being at work (Lopes, Avila & Beuron, 2018). In addition, even procedural situations in the area can lead to greater or lesser levels of satisfaction, such as a change of position or job position, for example. Attitudes of acceptance and satisfaction with changes are significantly higher when employees perceive that the change was planned and prepared by the organization, while the opposite generates insecurity and distrust in employees (Franco, Neiva, Nery & Demo, 2016).
Carneiro, Leite, Bastos, and Bittencourt (2020) emphasize that some organizational factors contribute both to well-being at work and for organizations to thrive in productivity in the market, such as: relationships with management, colleagues, and customers; the valorization of work; career opportunities; autonomy; organizational support; the fair wage; and identification with the organization.
The bibliographic review by Guimarães and Neto (2021) obtains similar results and highlights that the opposite conditions generate negative results, that is, remuneration incompatible with the position held and the job market; unhealthy organizational environments and conditions of physical and mental risks; and strenuous working hours directly reduce well-being during work. According to the authors, investments in actions and programs for quality of life at work are perceived as fundamental management strategies for maintaining well-being. On the other hand, leadership styles, management models, and working conditions tend to have a positive or negative impact, depending on the performance formats.
By observing antecedents and moderators of well-being, Farsen, Boehs, Ribeiro, Biavalti e Silva (2018) indicates that the promotion of quality of life in the work environment assumes an antecedent role in well-being at work. Hirschle and Gondim (2020) observe that the absence of support, resources and good relationships, and other aspects related to work have a more negative impact on indicators of wellbeing at work than their presence contributes positively. According to the authors, the perception of organizational support and resources at work, especially autonomy and social support from peers and managers, act as moderators, reducing negative impacts on well-being at work.

Job Satisfaction
During the first decades of the 20th century, the studies carried out approximated job satisfaction with the motivational process. Satisfaction and motivation at work were treated as similar theories, with satisfaction being one of the components of motivation that leads employees to behave in a way that increases organizational performance and productivity (Siqueira, 2008).
In this period, around the 1950s, Herzberg's theory of two factors emerged, which argues that the factors that lead to satisfaction are different and independent of factors that generate dissatisfaction at work. In this way, he argues that dissatisfaction is equivalent to no satisfaction, just as satisfaction corresponds to the non-occurrence of dissatisfaction. In other words, job satisfaction is not the opposite of dissatisfaction. According to the author, the absence of hygienic factors, linked to the work context, tends to generate dissatisfaction among employees. However, the presence of these elements is not satisfying. It arises with the insertion of motivational or intrinsic factors, generally linked to the content of the work, and that satisfy psychological needs, such as recognition, responsibility, achievement, progress, and the work itself (Pilatti, 2012).
Years later, at the beginning of the 21st century, researches aimed at defining job satisfaction as affectivity at work gained strength, an affective bond between the individual and his/her work that is the result of the components of the organizational environment. Regardless of the concept adopted as job satisfaction, there is a consensus that this aspect interferes with the health (or illness), the work environment, and the personal life of employees (Marqueze & Moreno, 2005).
Even after 80 years, the concept of satisfaction continues to encompass affectivity and remains a concept integrated by many dimensions (Siqueira, 2008). In a study on the different approaches to the concept, according to Siqueira (2008), job satisfaction corresponds to the degree of contentment with colleagues, compensation, managers, work nature, and promotions. Thus, the concept represents how much the individual experiences pleasurable experiences in the context of organizations.
Among the aspects related to the work environment, some studies sought to reaffirm the relationship of high levels of satisfaction with a pleasant relationship between individuals (Santos, Callefi, Shiozaki & Luz, 2020), as well as the relationship between leader and employee for good planning and execution of the work (Moretto & Silveira, 2021). In addition, the factors of involvement with work activities, alignment of expectations regarding work and reality, financial reward, stability, perception of organizational justice, and career momentum were also reiterated in the surveys as means to obtain high levels of job satisfaction (Beuren, Ribeiro & Silva, 2019). Assunção and Pimenta (2020) highlighted the need for managers to pay attention to environmental and psychosocial work conditions, lacking social support. Another aspect highlighted by Vieira and Bastos (2019) is the alignment between career anchors (needs, reasons, and competencies of professionals) and the reality of the company to obtain satisfaction.
Finally, the latest studies have demonstrated the consequences of job satisfaction within organizations. Moretto and Silveira (2021) concluded that the higher the level of employee satisfaction, the greater the chances of their growth in the company, the greater the professional's dedication, there is more demand for professionalization and, consequently, better performance.
This reality is also presented by Sillero and Zabalegui (2019) in a study with nurses in the perioperative ward. The authors demonstrated the importance of job satisfaction for professional commitment and participation in the unit's activities. With high levels of satisfaction, nurses tend to work with greater commitment and attention.

Method
To achieve the main objective of this study, data were collected from the Glassdoor website, from the list of the 100 best companies to work for, from the assessment carried out in 2020 for companies operating in the North American market. Glassdoor is a website with an international reach where users voluntarily and spontaneously review companies they have previously worked for. The questionnaires, answered anonymously, involve themes such as culture, salary range, quality of life, and other organizational aspects. The main report provided by the institution considers six indicators based on workers' assessments, namely: I) Culture and values; II) diversity and inclusion; III) work/life balance; IV) senior management; V) compensation and benefits; VI) career development.
For this study, and in line with the literature presented so far, indicators related to organizational culture and diversity and inclusion were not considered for the analysis of the proposed model, since these themes are not usually addressed in the literature regarding the concepts of well-being or job satisfaction and are treated as distinct variables. Items III, IV, V, VI are aspects frequently used both in measures of job satisfaction and well-being at work.
For example, in the Siqueira (2008) Satisfaction Scale, one of the most applied to measure satisfaction in the Brazilian scenario, of the five dimensions evaluated, three are satisfaction with the manager, satisfaction with compensation, and satisfaction with promotions, which are theoretically aligned with the indicators (IV) senior management, (V) compensation and benefits and (VI) career development, respectively. In turn, in the Well-Being at Work Scale originally proposed by Paschoal and Tamayo (2008), the three factors identified are positive affects, negative affects, and achievement, and they have theoretical alignment with such factors, being (III) work/life balance related to positive affections and (VI) career opportunities related to achievement.
Given the choice of the Glassdoor ranking, the sectors covered were diversified, that is, there was no control over their selection. However, it was possible to observe that the sample included organizations dedicated to the provision of health services and hospitals, and computer hardware and software.
To obtain the operational data of the companies listed in the ranking, the Eikon database was used, an open platform where it is possible to evaluate news, information, data, critical analysis, and a wide variety of different indicators from more than 40 thousand institutions around in the world. As one of the world's largest providers of financial markets data and infrastructure, using the database is a way to obtain reliable and standardized financial data, favorable for processing and comparison.
At Eikon, gross data for total assets, net income, total income, total liabilities, short-term liabilities, longterm liabilities, equity, monthly share price, and the monthly trading volume for the years 2018, 2019, and 2020 were collected. These were considered because they allow the future construction of a variety of relevant operating indicators, such as Return on Equity, Return on Assets, Return on Investment, among others. Since companies' data advertising and availability policies differ by region of operation and availability of each company, it was possible to collect only data from 41 companies in the proposed initial sample.
For the evaluation of the model in which Happiness at Work is assumed as a latent variable, the confirmatory factor analysis (CFA) was performed, considering that one of its advantages is the property to assess the construct validity of a proposed measurement theory (Hair et al., 2009). To evaluate the latent variable model, and following the indications of Hair et al. (2009), the CFI (Comparative Fit Index) was observed as an incremental index and the RMSEA (Root-Mean-Square Error of Approximation), and the SRMR (Standardized Root Mean Squared Residual), which considers the difference between the observed normalized and the predictable correlations, as absolute indexes.
Thus, it is necessary to pay attention to the necessary criteria for a structural model to be considered satisfactory: the RMSEA must present values lower than 0.05 or even 0.10; the CFI indices must be equal to or greater than 0.90, and; the SRMR is an absolute measure that must have a value less than 0.1 (Hair et al., 2009). Therefore, in line to verify the relationship between the variables, the Ordinary Least Squares (OLS) regression with heteroskedasticity correction via the White's matrix was used. Considering that one of the assumptions for regression analysis is homoscedasticity, that is, the data is less dispersed (or more concentrated) around the regression line, the use of heteroskedasticity correction via White's matrix is justified since we can obtain the heteroskedasticity correction of the standard errors of the estimators (Gujarati & Porter, 2011;Wooldridge, 2011).

Results and Discussion
In this chapter, the results obtained from factor analysis and Ordinary Least Squares (OLS) regression will be presented, as well as the discussion between these data and the literature.

Happiness reflected by 100 companies, from Glassdoor
At this stage, we tried to identify, through confirmatory factor analysis, whether the latent variable Happiness at Work reflects the indexes proposed by Glassdoor. Thus, the factor analysis sought to verify the existence of underlying relationships between the evaluative requirements of each of the indicators used by Glassdoor for the formation of the latent variable Happiness at Work. The option for this method resides in the note made by Hair et al. (2009) in which construct validity is demonstrated by the degree to which a set of measured items reflects the latent theoretical construct that those items should measure.
Among the variables presented by the platform, work/life balance (WL), senior management (SM), compensation and benefits (CB), career development (CD) were chosen, as explained in the method, as they reflect the concept of happiness at work, in line with Cropanzano and Wright's (2001) proposal. In this idea, happiness can be interpreted as job satisfaction, positive and negative affects, and psychological well-being. Thus, the proposed relationship has four variables and the statistical procedure used to test this relationship was multiple linear regressions.
As a result of the factor analysis, the model's fit indices were considered satisfactory for the Happiness at Work General Index: CFI=0.931, RMSEA=0.308, SRMR=0.078. However, the RMSEA presented a result above the satisfactory value, which can be explained by the sensitivity of the index to the sample size, tending to reject true models due to a small sample, such as the one in this study (Hair et al., 2009). The latent variables of the model were then observed, as shown in table 1: With these data, the next step was to standardize the weights, distributing the results found to find the proportion of the percentage importance of each variable against the general happiness index obtained. In this process, the estimates found in the latent variables were multiplied by the respective original index (WLI, CDI, SMI, or CBI). Then, each result was divided by the sum of all the estimates of latent variables, obtaining the single factor, the Happiness at Work General Index of the 100 best companies to work for from Glassdoor, comprising the theoretical dimensions of well-being and job satisfaction. Thus, the final values found for the index were: The unique factor found in this study brings important reflections on the understanding of Happiness at Work. Compensation and benefits were the indicator with the lowest participation in construct. We can turn to Herzberg's theory of motivation to understand this result. According to the author, factors that lead to satisfaction are different and independent from factors that generate dissatisfaction at work, that is, job satisfaction is not opposite to dissatisfaction. For Herzberg, compensation and benefits are part of the dimension of hygienic factors, elements related to the company and that cause dissatisfaction when absent, terrible, or precarious. However, when present they do not necessarily cause employee satisfaction. While motivation generates a desire, satisfaction fulfills a need. In the case of compensation and benefits, for example, it is possible to understand that the salary can satisfy the individual's needs, but it is not enough to motivate him (Pilatti, 2012).
About senior management appearing with the greatest contribution to the construct, the study by Fogaça and Coelho Júnior (2016) had already indicated that satisfaction with the head was the dimension of job satisfaction that best explained the individual job performance. In turn, it ends up going against Herzberg's theory, in which the author points out that the relationship with the boss is a hygienic factor, that is, when absent it generates dissatisfaction, but when present it is not responsible for generating a positive effect.
It is interesting to observe that the influence of managers on the employee's happiness has been pointed out since the first studies on the subject and is constantly highlighted until the most recent studies. In Locke's 1976 Discrepancy Theory, the relationship with bosses was already highlighted as one of the main agents influencing the employee's assessment of his work or work experiences and how much this facilitates and allows him to achieve his goals and values (Marqueze & Moreno, 2005). Moretto and Silveira (2021) disclosed the results of their studies and concluded that the relationship between leader and employee is important for good planning and execution of the work, as competent leaders capable of managing, developing, and adequately remunerating their employees lead to a continuous process of virtuous growth of the organization and the employee.
Regarding the other two indicators, career development and work/life balance, it can be inferred that their contribution to the model reinforces the relationship that well-being at work has with happiness. Ryan and Deci (2001) state that well-being is composed of the affective and cognitive dimensions. In the latter, well-being at work is demonstrated through the perception of expressiveness and personal fulfillment at work. Paz (2004) presents the gratification pole, composed, among other aspects, by the expectation of growth in the company, for the development of personal and professional skills. Warr and Nielsen (2018) return to the eudaimonic aspect of the theme and emphasize that well-being at work is related to self-fulfillment and expression through the exploration of one's potential and the development of personal attributes. It is also noteworthy that both the concept of well-being at work proposed by Paschoal and Tamayo (2008), and that of job satisfaction by Siqueira (2008), both used as the basis for this study, highlight this influence. Paschoal and Tamayo (2008) state that well-being at work involves the individual's perception that, in their work, they express and develop their potentials and abilities. Siqueira (2008), on the other hand, highlights not only the promotions but also how they are carried out.
Therefore, in line with what has been signaled by the literature, the Happiness at Work General Index corroborates the proposal that some authors defend, that well-being at work and job satisfaction can be used as variables observed in the understanding of the happiness construct (Cropanzano & Wright, 2001;Panaccio & Vandenberghe, 2009) since there is still no consensus on how to operationalize it. Thus, the factor analysis confirmed the study's proposal that the four indicators used to integrate the concept of happiness.
Interestingly, the choice of well-being at work and job satisfaction variables were preliminarily anchored in the proposal of scale factors used in the measurement of these variables (Paschoal & Tamayo, 2008;Siqueira, 2008). Considering that positive affects, negative affects, and achievement represent well-being at work, having theoretical alignment with work/life balance and career opportunities, it can be conjectured that well-being at work corresponds to half of the happiness construct. The same can be said about satisfaction, which is aligned with the senior management, compensation and benefits and career development indicators.
Thus, the Happiness at Work General Index can find theoretical support in the concept of Panaccio and Vandenberghe (2009), who understand happiness as the presence of positive affect, the absence of negative affect, and the presence of satisfaction, following the trend in organizational studies, as well as advances in understanding the construct about the study by Huang et al. (2015), which inspired this study and which was limited to considering Glassdoor indicators as factors of job satisfaction.
Another point worth mentioning after confirming the factorial structure of the index is the advance in using a measure analyzed at the organizational level. As explained at the beginning, Glassdoor's indicators represent the combination of employees' assessments that can be interpreted as aggregating perceptions at the organizational level, which enables a linear analysis situated at the same (organizational) level between happiness and operational performance (see Klein & Kozlowski, 2000). Therefore, the next stage of analysis is precisely to verify whether this promising and challenging issue, which is the relationship between happiness at work and performance (Guerci et al., 2019;Hauff et al., 2020;Warr & Nielsen, 2018) is confirmed with the chosen sample.

The influence of happiness on operational performance
This step sought to test the effects of the latent variable happiness at work, found in the results of the previous objective, on the operational performance of the 100 best companies to work for in the USA, according to Glassdoor. As presented throughout this study, it is expected that the Happiness at Work General Index can positively impact the operational performance of the organizations studied.
After obtaining the Happiness Index obtained in the previous objective, an Ordinary Least Squares (OLS) regression was carried out, which allowed the evaluation of the relationship between the dependent variable of operational performance (y), the Return on Assets (ROA), and the independent variable (x), the Happiness at Work General Index.
Return on Assets (ROA) is a business profitability indicator that reveals the return produced by the total investments made by a company in its assets. According to Assaf Neto (2021, p. 85), the "ROA is one of the most important profitability indicators of a company, being calculated by the ratio between the net income generated by the assets (operating income), as defined, and the total assets, owing these amounts must be expressed in the currency of the same date". The author also emphasizes that it is a measure that reveals how many times sales covered the total assets of a company in a year. Thus, it is given by: ROA = Net Income/Total Assets.
Among the tests carried out and the analysis of other possible indicators, it was observed that the ROA was the one with substantial results and the best investigation. This measure was also used in the Huang et al. (2015) paper, the basis for this study. However, the authors indicated that ROA is affected by accounting choices and does not incorporate intangible assets or leverage. The present study also identified that by inserting other elements that impact operational performance, such as control variables, the model does not have significance and becomes meaningless.
The next step was to correct the model through heteroskedasticity via White's matrix. The final result is presented below: The values demonstrated that the results are statistically significant, since the Pr (>|t|) indicates that the angle of the final linear equation is different from 0, that is, there is an average linear relationship between the ROA (y), and the Happiness at Work General Index obtained through factor analysis (x). The larger the angle, the greater the impact of the independent variable on the dependent variable. Similarly, the smaller the angle, the smaller the impact. Considering that the angle is not zero, the estimator (0.04901) indicates that the relationship between the variables is positive, creating the phenomenon represented by the following graph: The blue line represents the linear mean between the points, where the distance between the points represents the model's margin of error. It is possible to observe that many points are outside the acceptability region, represented by the part shaded in gray, which indicates the confidence interval, that is, the trend of the phenomenon. This error occurred due to the low amount of data obtained. In addition to the initial sample of 100 companies being reduced to 41 due to the availability of financial data, the heteroskedasticity correction via White's matrix reduced the sample to just 33 observations, so as not to reveal a marked and defined behavior.
The final result indicates that happiness at work (Happiness at Work General Index) positively impacts the operational performance of companies (ROA). Although the study does not indicate a strong correlation, it is noteworthy that this appears even with the low amount of data, indicating a relevant trend. This finding corroborates the findings by Huang et al. (2015), who concluded that employee satisfaction is positively associated with subsequent ROA, relating performance and job satisfaction. Although this research does not study the causes for a higher performance related to happiness, it is possible to use the literature to suggest some hypotheses. Taris and Schaufeli (2015) state that satisfied employees are also productive. Fogaça and Coelho Júnior (2016), when investigating this hypothesis, explain that well-being helps to achieve better individual performance and that when workers feel satisfied with the activity performed, the work flows, there is a greater willingness and, consequently, better performance.
The studies mentioned above analyzed performance at the individual level, and not at the organizational level as in the present paper. However, it is important to emphasize that individual performance is an important part of organizational performance, and that is why these studies supported the proposal made here of a positive relationship between happiness at work and performance. It is also noteworthy that by observing the predictive relationship of the Happiness at Work General Index on operational performance, this study meets the research agenda proposed by Sender and Fleck (2017), Nangov et al. (2018), and Nangoy et al. (2020), contributing to the understanding that happiness is one of the factors that play a significant role in organizational performance.
Nevertheless, customer satisfaction is sometimes understood as a way of analyzing organizational performance, Moretto and Silveira (2021) demonstrated that when an employee is satisfied, the chances of growth in the company increase, greater demand for professionalization, and professional dedication. Finally, there is evidence that satisfied workers tend to treat their customers with more attention, joy, and kindness, in addition to staying longer in the organization. As a result, customers tend to come back and, in addition to being welcomed again, find a familiar face that connects them to the brand (Scussel & Fogaça, 2017).
Through the results, the need to understand happiness as a strategic force that affects organizational performance (Nangov et al., 2018;Nangoy et al., 2020;Sender & Fleck; and the need to consider it in decision-making in human resources management, in cost savings, in a better organizational climate, in the company's business policies and strategic actions. Obviously, the results obtained here are more indicative than conclusive, but they are enough to support the various studies that have highlighted the importance of promoting organizational practices aimed at well-being at work and job satisfaction (Lopes et al., 2018;Vasconcellos & Neiva, 2016), since investing in environmental and psychosocial work conditions (Assunção & Pimenta, 2020;Carneiro et al., 2020), increases the possibility of achieving positive performance, whether at an individual or organizational level.
This inference emerges from the Glassdoor indicators used, which deal precisely with the aggregate perception of workers about organizational practices developed to promote happiness at work. Thus, this study corroborates the findings of Nangoy et al. (2020), who highlighted the implementation of considerable resources in promoting well-being has effects on the organization as a whole and is longterm.
A properly rich company is not only one that delivers excellent financial results. It is also a company with a rich environment, a rich climate, and employees who are full of happiness.

Conclusion
The main objective of this study was to investigate the variable of happiness at work reflected in the wellbeing and satisfaction indexes presented by the 100 best companies to work for in the USA, as well as its impact on operational performance. Initially, we sought to identify, through confirmatory factor analysis, whether the latent variable Happiness at Work reflects the indexes proposed by Glassdoor, and then test the predictive effects of the latent variable happiness at work on the operational performance of the 100 best companies to work for. It can be said that the proposed objective was fulfilled, since it was proven that the Happiness at Work General Index, the only factor in the factor analysis, and the operational performance, via ROA, are positively related, signaling that happiness at work has a positive impact in the operational performance of companies.
As theoretical contributions, we can highlight the proposition of the Happiness at Work General Index, considering that its construction was anchored in the definitions of two variables, job satisfaction and well-being at work, traditionally operationalized as happiness, in addition to being an index built from collective perceptions of the construct, enabling its relationship with another variable located at the organizational level. Still, in the theoretical field, this study expands the discussion of the relationship between happiness and performance, confirming this predictive relationship based on data obtained from databases known in the business environment.
As a practical contribution, this study reinforces the orientation for organizations to invest in people, since operational indicators are affected by these actions. Broadly speaking, this study signaled that happy companies can be richer. Worrying about promoting happiness can generate returns for companies. The manager can act on this issue by creating means and programs to promote quality of life; enable leaders to develop a good job with their followers and mediate this relationship; provide fair and interesting career advancement opportunities; in addition to satisfactorily remunerating employees with added salary and benefits.
As a limitation, it is highlighted that the sample was reduced to 41 observations due to the availability of data by companies and advertising in the Eikon database. Of these, only 33 were considered after heteroskedasticity correction via White's matrix. However, among all the indicators tested to support the analysis, only ROA showed satisfactory results. For future researchers, it is recommended to replicate the study with attention to the limitations presented above. It is suggested to use other bases of financial indicators, or also to expand the range of analyzed companies. In addition to covering a greater volume of data, including companies other than the 100 best to work for, it is possible to understand the behavior of the phenomenon in a wider range of factors, such as sector, type, and size of companies. In addition, it is recommended that future studies expand the theoretical discussion on the operationalization of happiness at work, applying the factors used here in other samples. Finally, it is hoped that the issues raised here may arouse interest for further research on the subject so that in the future it can contribute in a practical way to the business reality.