Firm value and employee attitudes on workplace quality.
Data Availability: Data used in this study are available from public sources.
Over the past several years, employers have increased their focus on organizational characteristics that enhance employee work experiences and assist employees in balancing their jobs and personal lives (Shellenbarger 1998). For example, companies offer employees amenities such as on-site daycare, concierge services, and flexible work schedules (Levering and Moskowitz 2000). Testimonials from companies such as Sears, MCI, EDS, and Northern Telecom indicate the same trend (see Rucci et al. 1998; Moskowitz 1997). This focus on workplace quality has become significant enough for business press publications such as Fortune and Working Mother to publish annual lists of companies that excel in creating a high quality of work life for their employees.
In order to expend resources to improve the quality of employee work life, companies must believe in the value of such investments. According to the strategic human management resource literature, such beliefs are well founded. Relying on a resource-based theory of the firm (Barney 1991), the literature posits that human resource activities can be a source of competitive advantage that produces improved financial performance (Wright et al. 2001). Evidence from studies such as Huselid (1995) and Fulmer et al. (2003) is generally consistent with this proposition. However, "systematic research on the measurement and valuation of human resources is extremely lean" (Lev 2002), and our understanding of the relation between human resources and firm performance remains limited (Becker and Huselid 2002).
This paper examines the relation between market values and employee attitudes on their company's quality of work life, called "workplace attitudes." We investigate whether companies with higher workplace attitudes have higher market values than companies with lower workplace attitudes. We also investigate whether differing degrees of workplace attitudes among those companies designated as high-quality "workplace attitude" firms are reflected in market values. Finally, we examine whether companies with higher workplace attitudes have greater market returns than companies with lower workplace attitudes.
These investigations are worthwhile for several reasons. First, empirical evidence in this area is lacking (Lev 2002). Second, no study to date examines, in a multivariate setting, the value relevance of workplace attitudes. Third, we introduce into the accounting literature a theory from the strategic human resource management literature that predicts that human resource practices provide a valuable competitive advantage for a firm. Fourth, the investigation identifies another example of unrecognized intangible assets and nonfinancial data that are useful in determining market values.
To test the association between workplace attitudes and firm value, we first compare market values of firms with higher workplace attitudes to market values of firms with lower workplace attitudes. We identify a firm as having higher (lower) workplace attitudes if it appears (does not appear) on Fortune magazine's annual list, "The 100 Best Companies to Work for in America" (hereafter Best Companies List). Regression results indicate that listed firms have higher market values than do nonlisted firms. We then compare market values of firms listed in the top one-third and bottom one-third of the Best Companies List, finding that market values of firms at the top exceed those of firms at the bottom. Finally, we compare market returns of listed and nonlisted firms, finding that listed firms generate greater returns than nonlisted firms in the two-years following inclusion on the Best Companies List. Taken together, these results provide some support that workplace attitude is a source of competitive advantage that leads to a valuable intangible asset.
THEORY AND RESEARCH QUESTION
In their book titled Cracking the Value Code, Boulton et al. (2000) report the results of surveying more than 250 executives from numerous industries about issues concerning value creation in their firms. Approximately 85 percent of respondents indicated that investments in employees and customers are critical to business. Further, executives rated employee retention as the second-most important measure of value creation in the current economy, ahead of items such as technological investment, brand recognition, market share, and new product development. In a similar study of approximately 6,000 managers and executives, McKinsey & Company found that talent, defined as "smart, sophisticated businesspeople who are technologically literate, globally astute, and operationally agile," is considered to be the most important corporate resource for the coming decades (Fulmer et al. 2003).
Strategic Human Resource Management Literature
The strategic human resource management literature formally captures such opinions. Over the past decade, much of this literature embraced the resource-based view of the firm set forth by Barney (1991) and others. According to this view, companies generate sustainable competitive advantages by effectively controlling and manipulating their resources and/or capabilities that are (1) valuable, (2) rare, (3) that cannot be perfectly imitated, and (4) for which no perfect substitute is available (Barney 1991). Human resource activities, including those that improve employee attitudes on workplace quality, meet these four characteristics (Fulmer et al. 2003; Lado and Wilson 1994). As a result, human resource activities can create a competitive advantage by developing a skilled workforce that effectively carries out the company's business strategy (Wright et al. 2001). This competitive advantage leads to improved performance, higher profitability, and greater market values (Becker and Huselid 1998).
Some recent anecdotal studies provide evidence consistent with this logic. For example, Taco Bell found that sales and profits in the 20 percent of stores with the lowest employee turnover were 100% and 55% higher than sales and profits in the 20 percent of stores with the highest employee turnover, respectively (Heskett et al. 1997). Sears calculated that a five-percentage point increase in employee satisfaction translated into a 1 percent increase in customer satisfaction, which led to a one-half percent increase in revenues (Rucci et al. 1998).
Empirical investigations also provide evidence consistent with this logic. For example, Lau and May (1998) find that firms rated as good places to work by Levering and Moskowitz (1993) experience higher sales growth, increased asset growth, and larger returns on assets than firms that were not so rated. Huselid (1995) detects an association between the presence of workplace initiatives to enhance employee morale and increases in productivity, market values, and accounting returns. Delery and Doty (1996) report a positive association between human resource practices and organizational performance within the banking industry.
Research Question Investigated
Despite the growing evidence on the relation between human resource activities and financial performance, knowledge of how these activities result in improved performance is limited (Becker and Huselid 1998). Fulmer et al. (2003) explore this issue by considering employee attributes that may act as a mediator between human resource activities and firm performance. One such attribute is employee attitudes. Fulmer et al. (2003) argue that employee attitudes affect behaviors and, therefore, performance. They argue further that employee attitudes can influence workplace quality through the attraction and retention of good employees. Although the authors cite evidence of a positive association between employee attitudes and employee performance (see Ostroff 1992), they note that a relation between workforce attitudes and organization-level performance has yet to be empirically documented.
Fulmer et al. (2003) attempt to document such a relation by using inclusion on Fortune's initial Best Companies List and inclusion on a similar listing by Levering and Moskowitz (1993) as proxies for firm-level aggregate employee attitudes. They argue that if employee attitudes resulting from high workplace quality produce a sustainable competitive advantage, then listed firms should exhibit better financial performance than do nonlisted firms. Fulmer et al. (2003) find generally better operating performance for listed firms than for a control sample. Further, they find that stock returns of listed companies outperform the returns of a broad market index 70 percent of the time and the returns of a control sample 60 percent of the time. They conclude that financial performance is enhanced when employees perceive that their employers create high workplace quality for them.
In sum, there is some evidence that companies successful in creating positive employee attitudes about their workplace possess a valuable competitive advantage, which leads to improved operating and market performance. However, the evidence has limitations. For example, Huselid (1995) examines workplace initiatives reported by employers, not the successful implementation of such initiatives. Lau and May (1998) and Fulmer et al. (2003) utilize only univariate tests and cannot rule out alternative explanations for their results. Delery and Doty (1996) examine only a single industry and do not control for any factors other than human resource practices.
We continue this line of research on the relation between firm performance and employee attitudes on workplace quality. Like Huselid (1995), Fulmer et al. (2003), and Delery and Doty (1996), we rely on the resource-based view of the firm to predict that firms create competitive advantages with human resource activities. Like Fulmer (2003), we use Fortune's Best Companies List to indicate the existence of positive employee attitudes regarding workplace quality. Accordingly, we predict that listed companies have higher market values than do nonlisted companies.
Research Question: Do firms with higher workplace attitudes receive higher market values than similar firms with lower workplace attitudes?
RESEARCH DESIGN AND EMPIRICAL RESULTS
Measurement of Workplace Attitudes
Testing our prediction that companies with higher workplace attitudes have higher market values than those with lower workplace attitudes requires the ability to observe a company's workplace attitudes. We assume that companies included on Fortune's Best Companies List have (1) high workplace attitudes and (2) higher workplace attitudes than companies not included on the list. The extent to which these assumptions are invalid decreases our ability to detect any true difference in market values resulting from differences in workplace attitudes.
Fortune first published its Best Companies List in January 1998. It has since published a list each January. The list reflects an exhaustive review of company practices described by management and employees of the companies (Levering and Moskowitz 2000). To be considered in these rankings, candidate companies must allow Fortune to administer a survey--the Trust Index developed by the Great Place to Work Institute--to at least 250 random employees. Employee responses to this proprietary survey, which measures trust in management, pride in work/company, and camaraderie, account for approximately two-thirds of a candidate's ranking score. Candidates must also explain their employment philosophies and practices by completing the Great Place to Work Institute's Cultural Audit. Candidates are encouraged to include in their Audits supplementary materials such as company newsletters and videos. Responses to the Audit account for the remaining one-third of a candidate's score.
Our use of the Fortune list is not unique. Lau and May (1998) and Fulmer et al. (2003) utilize similar lists. Fulmer et al. (2003) provide some assurance that the Fortune list is a valid reflection of employee attitudes about workplace quality by comparing survey data from employees of listed firms to survey data from employees of clients of Hewitt Associates. They compare responses on the likelihood that the employee will continue to work for the company in the coming year, concluding that the desire to continue employment is higher for employees of listed firms than employees of nonlisted firms. The authors also estimate that as a group the listed firms are approximately one standard deviation above the population mean for attitudes on workplace quality. Taken together, Fulmer et al. (2003) argue that the evidence provides a compelling case that listed companies have high workplace attitudes.
Sample Size and Selection
To test our prediction that listed companies have higher market values than other companies, we employ a matched sample design that compares listed firms to a matched group of firms that do not appear on Fortune's Best Companies List. Our sample of companies with higher workplace attitudes includes all of the publicly traded firms appearing on at least one of Fortune's lists published from 1998 to 2001, for a total of 186 companies. As Figure I shows, 66 of those firms appear on only one list, 58 on two, 30 on three, and 32 on all four lists. Nine different types of companies appear on the Fortune list. However, most of the listed companies (115, or 62 percent) are publicly traded.
Of the 115 publicly traded firms, we could not obtain financial and stock data from Research Insight for four firms. In addition, we eliminated from the analysis 19 financial institutions and four firms traded as American Depository Receipts, which left a total of 88 usable firms appearing on at least one list. (1) To maximize the sample size, we consider each year a firm appears on the list as an observation, which results in a total of 192 firm-year observations. Table 1 presents a reconciliation between the 88 sample firms and the 192 firm-year observations.
To select a match firm for each sample firm, we identified all firms in the same two-digit SIC code as each listed firm. We then eliminated any firm that appeared on one of Fortune's lists. From that reduced set of nonlisted firms, we selected as a match firm the one with the closest earnings before extraordinary items (EARNINGS) to the listed firm. In cases where the matched firm did not have adequate data to complete the tests, we selected the firm with the next closest EARNINGS. For firms appearing on more than one list, we repeated the matching process separately for each year. We used SIC code as our first matching criterion because the final sample includes firms from a wide variety of industries. We used EARNINGS as our second criterion to control for operating performance and size. We discuss in a later section the sensitivity of our results to the matching procedure and the use of firm-year observations.
Figure 2 reports descriptive statistics for the Fortune and matched control firms. To control for extreme observations when estimating the regression model, we truncated both the Fortune firm and the matched control firm data at 5 percent. Overall, the Fortune firms are larger than the matched firms. The average (median) market value of common equity for the Fortune firms equals $35,273 ($8,472) million, while the market value of common equity for the matched firms is $11,141 ($3,551) million. In addition, the Fortune firms have average assets of approximately $8,012 million compared to the matched firms' average of $6,182 million. Even though we match on EARNINGS, the Fortune firms earn significantly more than the matched firms ($893 million versus $472 million). This overall difference occurs because EARNINGS for a few of the Fortune firms greatly exceed those of the available matches within the same two-digit SIC code. For example, Intel Corporation's earnings are five to seven times greater than its match-firm in the four sample years. Finally, the Fortune firms spend more on R&D than the match-firms.
Tests of Market Values
Our main prediction states that firms with higher workplace attitudes have higher market values than those with lower workplace attitudes. We estimate the following multiple regression model to test our prediction:
(1) MV[E.sub.i,t] = [[alpha].sub.i,t] + [[beta].sub.1][LIST.sub.i,t] + [[beta].sub.2][BVE.sub.i,t] + [[beta].sub.3][EARNINGS.sub.i,t] + [[beta].sub.4][R&D.sub.i,t] + [[epsilon].sub.i,t].
The dependent variable MVE is the market value of equity at the end of the first quarter immediately after publication of the Fortune list. For example, for calendar-year companies appearing on Fortune's list in its January 2001 issue, we measure MVE as the closing price times the number of common shares outstanding on March 31, 2001. We utilize the first quarter stock price because the model includes fiscal year-end financial variables not generally available until well after the close of the fiscal year.
Our primary variable of interest in the model is LIST, which is a dummy variable taking the value of 1 for Fortune firms, and 0 for matched firms. A positive coefficient on LIST supports the hypothesis of higher market values for firms with higher workplace attitudes compared to firms with lower workplace attitudes.
We include earnings (EARNINGS) and book value of equity (BVE) to control for financial variables round to affect market value of equity. We also include research and development expenditures (R&D) to control for unrecognized intangible assets not included on the balance sheet (Hirschey and Weygandt 1985). Each of these three variables is measured at the end of the fiscal year immediately preceding inclusion on Fortune's list. Based on the results of prior literature, we expect the coefficient on each of these control variables to be positively associated with market value of equity.
Table 2, Column A, reports the results of the multiple regression estimation. All reported t-statistics are based on White's (1980) heteroscedasticity-consistent standard errors. The coefficient on LIST is positive and statistically significant (p < .0001), indicating that Fortune firms have higher market values than matched firms. This significant coefficient suggests that the presence of higher workplace attitudes increases the market value of a firm's equity by approximately $8 billion. This result agrees with our prediction that market participants attribute a competitive advantage to those firms that succeed in creating higher workplace attitudes.
The regression coefficients for the control variables agree with prior studies. The coefficients on BVE and EARNINGS are positive and significant. As an indication that stock prices impound another intangible asset, the coefficient on R&D is also positive and marginally significant. The results are substantially identical when we include a size control (assets or sales) in the model (see Table 2, Columns B and C). Finally, the estimated regression produces an adjusted R2 of over 0.80, indicating that the model explains much of the variation in market values.
One concern with the above model is that we force the coefficients on the control variables to be equal across the two samples. As a result, the LIST dummy variable improperly reflects the extent to which Fortune firm control variables affect market values differently than matched firm control variables. To address this, we estimate the regression analysis after including interaction variables for all of the listed firms. The only variable with significantly different coefficients across the samples is R&D. Therefore, we estimate our original regression model again and include an interaction variable between R&D and LIST. As Table 3, Column D reports, the coefficient on LIST and the interaction between R&D and LIST are both positive and significant. The coefficients on BVE and EARNINGS remain positive and significant as well.
Tests of Market Values among Listed Firms
Another concern is that the coefficient on LIST could be due to a bias in Fortune's selection process. Figure 2 indicates that Fortune firms had higher market values and earnings than matched firms in the year prior to inclusion on the list. This generates some concern that Fortune considers prior financial performance in its selection process, thereby biasing the list toward higher performing firms. Fulmer et al. (2003) report a similar concern in their study.
To address this concern and to provide additional evidence on the value relevance of workplace attitudes, we examine the differences in market values among listed firms only. If workplace attitudes provide a source of sustainable competitive advantage, then those firms best able to create higher workplace attitudes should have the greatest competitive advantage and therefore the highest market values. Furthermore, the presence of a selection bias should not affect the interpretation of this comparison test.
One method to examine whether market values reflect a firm's ranking on the Best Companies List is to conduct a rank test. However, a rank test requires an assumption of linearity within rank, or equal difference in scores between adjacent firms. We do not believe that the difference between firms listed as first and second is likely to be the same as the difference between firms listed as, say, tenth and eleventh. Since we cannot assume linearity, we cannot conduct a rank test. Instead, we separate listed firms into three groups--those ranked in the top one-third of the list, the middle one-third, and bottom one-third. We then eliminate the firms in the middle one-third and estimate the following regression with the remaining firms:
(2) MV[E.sub.i,t] = [[alpha].sub.i,t] + [[beta].sub.1][RANK.sub.i,t] + [[beta].sub.2]BV[E.sub.i,t] + [[beta].sub.3][EARNINGS.sub.i,t] + [[beta].sub.4][R&D.sub.i,t] + [[epsilon].sub.i,t].
The variables in the regression model are identical to those in Equation (1) except for RANK, which is substituted for LIST. RANK is a dummy variable taking the value of 1 if the firm was included in the top one-third of the list, and 0 if it was included in the bottom one-third of the list. A positive coefficient on RANK suggests that investors value the degree of workplace attitudes. The results from this regression, reported in Table 3, indicate that firms ranked in the top one-third of Fortune's list have marginally higher market values than firms ranked at the bottom of the list (p-value = 0.0655). In other words, the market values relative levels of workplace attitudes. This result complements the original regression analysis of the Fortune and matched firms and provides additional evidence on the association between market values and workplace attitudes. It also helps to ensure that our results are not solely a function of any potential bias in Fortune's selection process.
Tests of Market Returns
Each of the above tests examines market values at a specific point in time. Taken together, the test results suggest that market values reflect differences in workplace attitudes. To provide more evidence on the value of workplace attitudes, we examine market returns of listed and nonlisted firms. If workplace attitudes lead to a sustainable competitive advantage, we should expect higher returns for listed firms. To test this prediction, we compare two-year market returns for the 1998, 1999, and 2000 Fortune firms and their matches. We do not utilize the 2001 firms as we do not have the necessary returns data for these firms.
Table 4, Panel A reports cumulative two-year forward returns for listed and nonlisted firms. Two-year forward returns equal the change in market values, adjusted for dividends, for two years from the date Fortune listed a firm. For example, for 1998-listed firms, we calculate the market returns from January 1998 to December 1999. Fortune firms realized average returns of 68 percent while matched firms realized average returns of 13 percent. The significant difference between the two suggests that workplace attitudes affect long-term market performance. We also compare the two-year cumulative forward market returns of high- and low-ranked Fortune firms. Firms in the top one-third of Fortune's list realized average returns of 59 percent while firms in the bottom one-third of Fortune's list experienced average returns of 43 percent. Although the difference is not statistically significant, the comparison has very low power due to the small sample size.
Tests of Causality
While greater market returns for listed firms is consistent with the resource-based view of the firm that human resource activities can result in a sustainable competitive advantage, the above comparison does not establish causality. To provide some evidence on whether the greater returns of listed firms allow them to invest in workplace attitudes or whether investments in workplace attitudes result in greater returns, we also compare prior returns for listed and nonlisted firms. Because it is likely that listed firms were in the process of creating high workplace attitudes some time before being listed, we do not calculate returns on the two years immediately prior to appearing on the list. Rather, we calculate and compare returns during the fourth and third years before listing. The results of this comparison, reported in Table 4, Panel B, indicate that Fortune firms realized average two-year cumulative returns of 89 percent while matched firms realized average returns of 59 percent. Despite the large absolute difference, these average returns do not significantly differ from each other at conventional levels. Thus, returns could not be used to distinguish the two groups of firms prior to inclusion on the Fortune list; however, returns after inclusion are significantly different. This result provides support for our hypothesis.
We also compare the differences in monthly returns for Fortune and matched firms for the two-year time periods used above: Fortune monthly return--matched monthly return. The average difference in monthly returns between Fortune and matched firms for the two-year period after inclusion on the list is 1.3 percent. This is significantly different than zero at the 0.01 level and again indicates that the listed firms have higher returns after being listed. In contrast, the average difference in monthly returns between Fortune and matched firms for the two-year period prior to inclusion is 0.6 percent, which is not significantly different than zero at conventional levels. These results provide additional support for the theory that strong workplace attitudes result in better long-term market performance.
In summary, our results suggest that the market values high workplace attitudes. Companies included on Fortune's Best Companies List have higher market values and better returns than matched firms not included on the list. Further, the market values firms ranked in the top one-third of the list more than firms ranked in the bottom one-third of the list.
Additional Sensitivity Tests
To provide assurance that our methodology did not drive our results, we conducted the following sensitivity analyses. We first addressed the sample itself. In our primary test of market values, we considered each year a firm was included on Fortune's Best Companies List as an independent observation. To determine if that treatment affected our results, we estimated the primary regression using data only from the year that Fortune first listed a firm. This reduced our sample to the 88 Fortune firms and their matched firms. The results of the regression estimation using these firms mirror those reported in Table 2. The coefficient on LIST remains significantly positive.
We also addressed the matching process. For our primary regression analysis, we matched first on two-digit SIC codes and then on earnings. However, as reported in Figure 2, nearly all of the variables of interest differed statistically between the two groups. Fulmer et al. (2003) report a similar experience with their matched sample. To address this issue, we reevaluated the results after eliminating the worst matches. In particular, we omitted six listed firms and their matches with earnings differences of greater than $2 billion. Recalculated descriptive statistics revealed no significant differences among sales and total assets; however, listed firms remained significantly larger for all other variables. The regression results excluding these firms mirror those of the original regression. Market values of Fortune firms significantly exceed those of matched firms.
As an additional test, we created a new set of matched firms by choosing the two-digit SIC firm with the closest average return on assets for the prior three years. (2) The results for this group of matched firms approximated those previously discussed. However, we again found significantly different descriptive statistics between the listed firms and the matched firms.
In an effort to remove the size discrepancy, we created a third set of matched firms by choosing the firm with the closest total assets within the same one-digit SIC code as listed firms. Average total assets for these matched firms were not substantially different from sample firms. Regression analysis using this sample of firms confirmed our original conclusions. Listed firms have significantly higher market values than the matched firms (t-value = 4.61, p-value < 0.001). Thus, the size difference of Fortune and matched firms does not appear to be driving our results.
This study investigates the relation between the value of a firm and its employees' attitudes on workplace quality. We assume that firms included on Fortune's annual list, "The 100 Best Companies to Work for in America," have higher workplace attitudes. Regression analysis of firms included on and excluded from Fortune's list suggests that firms with higher workplace attitudes achieve significantly higher market values than firms with lower workplace attitudes. On average, the market value of equity of firms with higher workplace attitudes (the firms included on the Fortune list) is approximately $8 billion higher than the market value of the matched firms. Additional analysis indicates that firms ranked in the top one-third of Fortune's list that presumably have the highest workplace attitudes have higher market values than those ranked lower on the list. Finally, we find that while two-year prior returns do not significantly differ between the high-quality firms and their matches, two-year future returns for high-quality firms exceed the corresponding returns of matched firms.
Our results support the claims of the financial press that employee initiatives benefit the employees, the firms, and investors. They also provide evidence consistent with the prediction in the strategic human resource management literature that human resource activities can be a source of competitive advantage that results in improved financial performance. As such, our results add to a lean area of research (Lev 2002) that is not fully understood (Becker and Huselid 1998). The results also complement testimonials from companies such as Sears and Taco Bell.
The results of our study suggest that workplace attitude is an important intangible asset. Thus, developing reliable measures of workplace attitudes that can be reported in the financial statements or footnotes is important. We do not address in this study the proper method of recognizing and measuring such an asset. Although we find that inclusion on Fortune's list is significantly related to market values, we do not attempt to identify a measure of employee attitude that meets the current definition of an asset. Finns do not recognize many other intangible assets because of the same problem (Lev 2002). Short of recognizing employee attitudes in the financial statements, we believe that a reliable nonfinancial measure that is standardized, objectively verifiable, and disclosed in the footnotes to the financial statements will provide useful information for shareholders to better understand a firm's value creation opportunities (Lev 2002). Perhaps scores from the Trust Index mentioned earlier could serve as this measure. We anticipate that any measure of employee attitude, whether financial or nonfinancial, must periodically be tested for impairment.
An important limitation of this study is the assumption that inclusion on Fortune's list represents workplace attitudes. While Fulmer et al. (2003) provide some validation for this assumption, firms that appear on the list might not actually have high workplace attitudes. Further, it is possible that some firms with high workplace attitudes did not complete the protocol required in Fortune's listing process, perhaps because of the costs associated with this evaluation process. Even if such measurement error and/or self-selection bias is present, we believe that these problems would make it more difficult to find significant results. Nonetheless, our inability to rule out these concerns should be noted as a limitation of the study.
An interesting extension of this study is to investigate the extent to which investors alter their valuations of investments in intangibles such as attitudes about workplace quality when general economic conditions change. This study utilizes data from a time period characterized by a strong economy. Would we find similar results in a weaker economy, such as the recessionary period beginning in late 2001? The resource-based view of the firms and our results suggest that employee initiatives are valuable and should not be restricted because of short-term slumps in the business cycle. Performing this same analysis after an economic downturn and comparing it to the results of this study should provide insights regarding this question.
TABLE 1 Sample Firms Sample Number Percentage Publicly traded firms 115 100 Less: Firms with incomplete data (4) (4) Financial institutions (19) (16) American Depository Receipts (4) (4) Final Sample of Firms 88 76 Firm-Year Observations Firms Observations Firms appearing on one list 33 33 Firms appearing on two lists 23 46 Firms appearing on three lists 15 45 Firms appearing on four lists 17 68 88 192 TABLE 2 Tests of Valuation of Workplace Attitudes MV[E.sub.i,t] = [[alpha].sub.i,t] + [[beta].sub.1] LIS[T.sub.i,t] + [[beta].sub.2] BV [E.sub.i,t] + [[beta].sub.3] EARNING[S.sub.i,t] + [[beta].sub.4]R&[D.sub.i,t] + [[epsilon].sub.i,t] Coefficient (t-statistic) Variable A B C D Intercept -5,207.0 -3,662.76 -3,777.83 -3,799.94 (-4.58) * (-3.13) * (-3.76) * (-3.58) * LIST 7,999.71 7,091.63 6,491.85 6,024.62 (5.20) * (4.49) * (4.19) * (3.87) * BVE 2.22 3.56 3.91 2.07 (2.09) ** (2.42) * (2.29) * (1.91) ** EARNINGS 20.96 21.12 21.00 21.45 (4.59) * (4.84) (4.59) * (4.62) * R&D 10.47 9.04 9.98 3.32 (1.59) (1.25) (1.43) (0.51) LIST*R&D 8.52 (1.87) ** SALES -0.66 (-1.68) ** ASSETS -0.81 (-1.64) ** [R.sup.2] 0.81 0.82 0.82 0.81 n 384 384 384 384 *, ** Indicates t-statistics are significant at the 0.01 and 0.05 levels, respectively, for a one-tailed test. The variables in Table 2 are defined as follows: MVE = market value of common equity at the end of first quarter immediately following inclusion on the list; LIST = dummy variable taking the value of 1 if the company is included on Fortune's list and 0 otherwise; BVE = book value of common stock at the end of the fiscal year-end immediately preceding inclusion on the list; EARNINGS = earnings before extraordinary items for the fiscal year immediately preceding inclusion on the list; R&D = research and development expenditures for the fiscal year immediately preceding inclusion on the list; SALES = sales for the fiscal year immediately preceding inclusion on the list; and ASSETS = total assets at the fiscal year-end immediately preceding inclusion on the list. TABLE 3 Tests of Valuation of Degree of Workplace Attitudes MV[E.sub.i,t] = [[alpha].sub.i,t] + [[beta].sub.1] RAN[K.sub.i,t] + [[beta].sub.2] BV[E.sub.i,t] + [[beta].sub.3] EARNING[S.sub.i,t] + [[beta].sub.4] R&[D.sub.i,t] + [[epsilon].sub.i,t] Coefficient (t-statistic) Intercept -2,344.96 (-0.94) RANK 7,246.33 (1.52) *** BVE 3.49 (2.53) * EARNINGS 20.76 (3.75) * R&D 7.59 (1.17) [R.sup.2] 0.80 n 125 *, *** Indicates t-statistics are significant at the 0.01 and 0.10 levels, respectively, for a one-tailed test. The variables in Table 3 are defined in Table 2 except for the following: RANK = dummy variable taking the value of 1 if the company is ranked in the top one-third of the listed firms, and 0 if it is in the bottom one-third of the listed firms. TABLE 4 Comparison of Market Returns Panel A: Two-Year Cumulative Forward Returns Fortune Control t-statistic n 0.6766 0.1288 2.15 ** 128 High Rank Low Rank t-statistic n 0.5857 0.4291 0.50 51 Panel B: Two-Year Cumulative Prior Returns Fortune Control t-statistic n 0.8937 0.5912 1.15 126 ** Indicates t-statistics is significant at the .05 level. FORWARD RETURNS = a firm's market returns for the year of inclusion and the year after inclusion on Fortune's list; and PRIOR RETURNS = a firm's market returns for the fourth and third years prior to inclusion on Fortune's list.
(1) We eliminated financial institutions because of their strict regulatory environment and their unique reporting requirements. We eliminated firms traded on U.S. exchanges as American Depository Receipts because they are foreign firms that are not required to report GAAP-based financial statements.
(2) We defined return on assets as annual net earnings divided by year-end total assets.
Barney, J. B. 1991. Firm resources and sustained competitive advantage. Journal of Management 17: 99-120.
Becker, B., and M. Huselid. 1998. High performance work systems and firm performance: A synthesis of research and managerial implications. In Research in Personnel and Human Resources, edited by G. F. Ferris, 16: 53-101. Greenwich, CT: JAI Press, Inc.
Boulton, R., B. Libert, and S. Samek. 2000. Cracking the Value Code: How Successful Businesses Are Creating Wealth in the New Economy. New York, NY: HarperCollins Publishers Inc.
Delery, J., and D. Doty. 1996. Modes of theorizing in strategic human resource management: Tests of universalistic, contingency and configurational performance predictions. Academy of Management Journal 39: 802-835.
Fulmer, I., B. Gerhart, and K. Scott. 2003. Are the 100 Best better? An empirical investigation of the relationship between being a "great place to work" and firm performance. Personnel Psychology (forthcoming).
Heskett, J., E. Sasser, and L. Schlesinger. 1997. The Service Profit Chain: How Lending Companies Link Profit and Growth to Loyalty, Satisfaction, and Value. New York, NY: The Free Press.
Hirschey, M., and J. Weygandt. 1985. Amortization policy for advertising and research and development expenditures. Journal of Accounting Research 23 (Spring): 326-335.
Huselid, M. 1995. The impact of human resource management practices on turnover, productivity, and corporate financial performance. Academy of Management Journal 38: 635-672.
Lado, A. A., and M. C. Wilson. 1994. Human resource systems and sustained competitive advantage: A competency-based perspective. Academy of Management Review 19: 699-727.
Lau, R., and B. May. 1998. A win-win paradigm for quality of work life and business performance. Human Resource Development Quarterly 9 (Fall): 211-226.
Lev, B. 2002. Intangibles: Management, Measurement, and Reporting. Washington, D.C.: Brookings Institution Press.
Levering, R., and M. Moskowitz. 1993. The 100 Best Companies to Work For in America. New York, NY: Currency/Doubleday.
--, and --. 2000. The 100 best companies to work for. Fortune (January 10): v141, n1.
Moskowitz, M. 1997. The 100 best companies for working mothers. Working Mother (October): 18-96.
Ostroff, C. 1992. The relationship between satisfaction, attitudes, and performance: An organizational level analysis. Journal of Applied Psychology 77: 963-974.
Rucci, A., S. Kirn, and R. Quinn. 1998. The employee-customer-profit chain at Sears. Harvard Business Review 76 (January/February): 83-97.
Shellenbarger, S. 1998. Companies are finding it really pays to be nice to employees. Wall Street Journal (July 22): B1.
White, H. 1980. A heteroscedasticity-consistent covariance matrix estimator and a direct test for heteroscedasticity. Econometrica 48 (May): 817-838.
Wright, P., B. Dunford, and S. A. Snell. 2001. Human resources and the resource based view of the firm. Journal of Management 27: 701-721.
Brian Ballou is an Associate Professor at Miami University, Norman H. Godwin is an Associate Professor at Auburn University, and Rebecca Toppe Shortridge is an Assistant Professor at Ball State University.
We thank Ella Mae Matsumura and two anonymous referees for their contributions to the paper.
Submitted: November 2001
Accepted: August 2003
Corresponding author: Norman H. Godwin
|Printer friendly Cite/link Email Feedback|
|Author:||Ballou, Brian; Godwin, Norman H.; Shortridge, Rebecca Toppe|
|Date:||Dec 1, 2003|
|Previous Article:||CEO and CFO certifications of financial information.|
|Next Article:||The information content of royalty income.|