The Effects of Non-Expensed Employee Stock Bonus on Firm Performance: Evidence from Taiwanese High-Tech Firms

Authors


Abstract

The choice of whether to expense broad-based stock incentives has been a highly controversial debate in both academic research and practice circles. We provide insightful findings to reconcile certain debates regarding the effectiveness of non-expensed, broad-based stock incentives. Using a unique longitudinal dataset from Taiwanese high-tech firms over the 1997–2008 period, our results indicate that non-expensed employee stock bonus incentives exerted positive effects on short-term organizational value added creation. The dilution effects of broad-based stock incentives in Taiwan, however, exerted a negative influence on profitability and eroded share return. The negative effects were even more severe in the following year, and overexploitation of employee stock bonus also damaged the long-term organizational performance of Taiwanese high-tech firms. This negative aspect of non-expensed employee stock incentives resulted in more evidence for changing the regulatory context of broad-based stock incentives in Taiwan.

1. Introduction

For more than two decades, firms have used employee equity participation as a primary incentive mechanism to motivate employees and to enhance firm performance (Blasi and Kruse 2006; Blasi et al. 2008; Kruse 2002). Firms with high capital investments in information technology software and hardware rely intensively on stocks and options as incentive schemes to align employee and employer interests more closely (Ittner et al. 2003; Murphy 2003; Sesil et al. 2002). The stock incentives adopted by these new economy firms are typically broad-based schemes, in which most firm employees are eligible to participate (Poutsma et al. 2003). Given the increasing difficulty of monitoring employees in new economy firms, employee stock incentives are an efficient alternative mechanism to formal monitoring (FitzRoy and Kraft 1995; Milgrom and Roberts 1995). Based on the agency cost argument, employee stock incentives allow these firms to motivate employees (Core and Guay 2001), attract and retain employees (Oyer and Schaefer 2005), enhance organizational productivity (Sesil et al. 2002), and promote entrepreneurial behaviour (Blasi et al. 2003). Thus, broad-based stock incentives are typically expected to increase a firm's operational and financial performance (Hochberg and Lindsey 2010; Sesil et al. 2002).

However, because incentive schemes in Taiwan have received favourable accounting treatment, certain firms have overexploited them and granted excessive broad-based stock incentives. The perception of lower cost may become the primary reason why firms choose such stock incentives, and these ‘cost free’ perceptions may lead to the abuse of such employee stock incentives (Hall and Murphy 2003). Mimetic or normative forces may cause firms to adopt employee stock incentives that mismatch a firm's monitoring context (Brandes et al. 2005). Therefore, possible abuse and the dilution effects of broad-based equity incentives have persisted as shareholder concerns (Walters and Young 2008).

Are employee stock incentives beneficial to firm financial performance, and to what extent will unrecognized expenses of employee stock incentives influence firm profitability and market reaction? These controversial debates call for more substantive empirical analysis to explain and respond to the effects of stock incentives that high-tech industries have created. Although prior studies have supported the view that employee equity incentives improve firm performance (Hochberg and Lindsey 2010; Kruse 2002; Sesil et al. 2002), empirically examining the overall performance effects of employee stock incentives that operate under favourable accounting rules remains worthwhile.

We contribute to the literature by examining the relation between non-expensed employee stock incentives and various measures of organizational performance. We empirically analyse this relationship by using a unique longitudinal dataset from publicly traded, high-tech firms in Taiwan, where the non-expensed employee stock incentives have a relatively loose vesting period regulation. Taiwanese high-tech firms have mostly used employee stock bonuses (ESB), rather than employee stock options, to share wealth with their employees. ESB plans share similar attributes with employee stock options, which are widely applied in new economy firms in the United States. Until 2008, favourable accounting regulations did not require firms to recognize the expense of ESB plans. Distributing ESB to employees has been accomplished using newly issued stocks, which may lead to dilution effects. Taiwanese ESB incentives typically have short vesting period requirements, and therefore employees have high autonomy to trade granted stocks in the stock market. Therefore, findings from prior studies (Hochberg and Lindsey 2010; Kruse 2002; Sesil et al. 2002) that are easily generalized to Taiwanese ESB incentives involved relatively favourable attributes (i.e. non-expensing accounting treatment and dilution possibilities).

In contrast, the Taiwanese ESB scheme offers a unique opportunity to examine the effect of broad-based stock incentives on organizational performance when compared with earlier comparative studies. Taiwanese high-tech firms have widely adopted broad-based stock incentives as a primary compensation mechanism for more than two decades. The growth of Taiwanese high-tech firms has been incredibly strong in the past 15 years (Dodgson 2009). Prior studies have argued that the wide diffusion of employee equity schemes and the compelling incentives they have created have become imperative forces that promote prosperity within Taiwanese high-tech industries (Chang and Yu 2002; Han and Shen 2007). However, the active debate on the effectiveness of Taiwanese ESB incentives existed between management level and investors, particularly foreign investors, in a long time. To explore whether the effects of ESB incentive on profitability and market return are consistent with efficiency considerations, we used a longitudinal dataset with details on ESB incentive schemes, which were disclosed in firms' annual financial reports, to investigate the performance effects of broad-based equity incentives along with the growth of a new economy sector.

We observed that non-expensed ESB incentive schemes raised short-term productivity; however, organizational profits and stock returns faded because of dilution effects, particularly in the long term. The diverse effects on organizational performance, therefore, may be the root cause of the controversial debates on using non-expensed ESB schemes. We also observed that the overexploitation of non-expensed ESB incentives may not only impede profitability and stock returns, but also produce worse multiple facets of long-term performance.

To present our work on this critical issue, this article is outlined as follows. Section 2 offers the institutional background of employee equity incentive schemes in Taiwanese high-tech firms. Section 3 provides a review of theoretical arguments and existing empirical findings. Section 4 contains a description of our data and our estimation framework. Sections 5 and 6 present our empirical results and a comprehensive discussion, and finally Section 7 offers concluding remarks and study limitations.

2. ESB in Taiwanese high-tech firms

Traditionally, Taiwanese companies have shared their profits with employees in the form of cash bonuses. In 1983, United Microelectronics Corp. (UMC), one of Taiwan's leading semiconductor manufacturers, began to share equity with its employees and became the first company to give employees newly issued stock as their share of the company's profits. Employee equity incentive schemes have since become the main vehicle to motivate employees in Taiwanese high-tech firms.

Similar to other new economy firms worldwide, Taiwanese high-tech firms have primarily distributed equity incentives to employees directly. The practice represents a transformed form of profit-sharing that allows employers to reward employees with company equity. Each employee receives the awarded equity based on his or her salary level, seniority, position and individual performance rating. In Taiwan, these plans resemble restricted stock plans in the United States, but have shorter vesting periods, typically between three months to two years. Employees typically have full autonomy to hold or liquidate their awarded stocks after the restricted period. Because the vesting periods are short, Taiwanese employee equity incentives are similar to both cash profit-sharing and equity participation (Sesil et al. 2002), and are referred to as ESB (Han and Shen 2007).

Figure 1 depicts the ESB usage rate in Taiwanese publicly traded high-tech firms during the 1992–2007 period. The proportion of publicly held corporations launching ESBs increased from 25 per cent at the end of 1992 to 80 per cent in 2000, and then levelled out at approximately 50 per cent in 2007. ESB usage peaked in Taiwanese high-tech firms in 2000 and gradually decreased to 50 per cent from 2001 to 2007. Numerous initial public offerings (IPO) of high-tech firms in the 1990s contributed to the substantial increase in the total usage rate in 2000, because most of these firms adopted ESB before the IPO date. ESB schemes are diffused in Taiwanese high-tech firms and have become an institutionalized practice.

Figure 1.

The Usage Rate of Employee Stock and Cash Bonus Plan in Taiwanese High-Tech Firms.

Taiwanese ESB schemes are broad-based stock incentives and a critical element of employment relationships in Taiwanese high-tech industries. According to a Taiwanese company's articles of incorporation, employee profit-sharing may be allocated in either cash or stocks. Under this system, all firm employees are eligible to participate in the ESB plan. To attract talent, Taiwanese high-tech firms typically claim in their recruitment advertising that they offer employees the opportunity to share in the financial success of the company by becoming shareholders. Employees in Taiwanese high-tech industries also consider ESB a vital part of their total compensation. Therefore, the ‘real money’ for employees in Taiwanese high-tech firms is not in wages. As reported in a 1999 salary survey by the Electronic Engineering Times, the annual payment of company stock to employees was the most crucial financial reward employees received because the salaries may be low. Yearly earnings stock bonuses, however, typically rendered the overall package more attractive (Carroll 1999).

Figure 1 also shows the usage rate of cash bonuses among Taiwanese high-tech firms, and illustrates an interesting trend regarding the substitution between stock and cash bonuses. Before 2000, ESB programmes diffused quickly, and the practice of adopting cash bonuses decreased. However, the practice of adopting cash bonuses increased gradually after 2000, and the practice of ESB began to decrease. Therefore, a certain type of substitution may exist between these two types of employee bonus. Approximately 40 per cent of Taiwanese high-tech firms pay both cash and stock bonuses to employees.1

The phenomenon of widely adopted ESB plans was partially because of favourable accounting regulations, which did not require firms to recognize the ESB expense from revenues. Under such regulations, firms did not, in effect, ‘pay’ for the stock bonuses they gave employees because ESB was treated as a dividend paid out of retained earnings. Stocks were distributed to employees from the ESB plans by issuing new stocks, not through stock repurchase, typically generating equity dilution concerns.

Shareholders and investors, particularly foreign investors, felt that these high-tech firms abused the ESB schemes and accused them of distributing excessive numbers of stock bonuses to employees. For example, an Asian Wall Street Journal article in 2002 noted that technology companies in Taiwan offered the most lavish ESB incentives worldwide. UMC, the first high-tech firm to launch employee equity incentives in Taiwan, distributed approximately 86 per cent of total earnings to employees in 2001.2 The issue of additional common shares from ESB also causes stock dilution. The average increase in the number of shares outstanding from ESB exceeded 1 per cent annually for Taiwanese high-tech firms from 1996 to 2003. Therefore, similar to the intense discussions regarding the effects of accounting standards for employee stock options on economic consequences in the United States (Guay et al. 2003), this accounting treatment also generated extensive debate in Taiwan.3 The rate of stock dilution decreased to 0.8 per cent from 2004 to 2007. In 2008, the stock dilution dropped to 0.18 per cent when firms were required to recognize ESB expense.

3. Theoretical arguments

3..1 Effects on Productivity

As one type of collective incentive, employee equity incentives are particularly appropriate under certain conditions, such as ‘new economy’ work settings (Applebaum and Berg 2000; Ben-Ner and Jones 1995), in which employees are likely to be highly skilled, causing a shift in the balance of power between them and the firm. Employees are in an advantageous position to access customer needs and have more private information from which management can benefit (Levine and Tyson 1990). Co-operation among employees is essential for high-tech firms to improve performance. However, traditional monitoring efforts might be costly and less effective than horizontal monitoring (FitzRoy and Kraft 1995). An individual incentive scheme, such as piece-rate, may discourage co-operation and innovation (Lazear 1998). However, giving employees incentives to monitor themselves and others through group incentive plans may be an effective substitute for formal monitoring and traditional incentives (Sesil et al. 2002). Among the various group incentive schemes, direct employee equity incentives are regarded as the optimal mechanism to motivate employees in high-tech firms.

Taiwanese ESBs should, thus, substantially improve organizational productivity because these schemes contain both cash profit-sharing and equity participation attributes, both of which are powerful motivational mechanisms. As discussed in prior literature, several theories explain how these schemes improve employee productivity. First, the most direct positive effects of employee stock incentive plans result from manifesting enterprise success in a higher stock price, thus rewarding employees with higher wealth in the form of stock (Kato and Morishima 2002). Direct employee equity incentives, which allow employees to sell awarded equity after vesting periods, preserve the characteristics of cash profit-sharing. In line with expectancy theory, these incentive schemes create an additional performance-contingent incentive to motivate greater worker effort (Sesil et al. 2002).

Second, direct employee equity incentives exert subtle, psychologically indirect effects on productivity (Jones and Kato 1995). Most direct employee equity incentives are broad-based, in which stocks are granted to most rank-and-file employees. Employee equity incentives help create a co-operative culture in which employees are more likely to share technological information and accept the introduction of a new technology that improves productivity (Sesil et al. 2002). Third, employee equity might boost productivity as a flexible efficiency-wage scheme. Such ‘x-efficiency’ effects may increase organizational attractiveness for job applicants, and consequently enhance organizational efficiency and productivity through hiring high-quality employees (Baddon et al. 1989).

Fourth, previous studies have suggested that formal ownership is associated with employee attitudes and behaviour, such as high job satisfaction, improved organizational performance, lower employee turnover (Long 1978; Wilson and Peel 1991) and strong organizational commitment (Florkowski and Schuster 1992). Festing et al. (1999) also observed that profit-sharing schemes in Europe decreased absenteeism and employee turnover. Therefore, sharing profits and ownership with employees might foster and protect firms' specific investment in human capital, which is critical to enhance organizational productivity (Blair 1999).

Previous scholars have argued that productivity effects associated with group incentive schemes, such as profit-sharing and company bonuses, might decrease because of the free-rider or 1/n problem (Blasi et al. 1996; Hall and Murphy 2003; Sesil et al. 2002); however, the work design and management style of Taiwanese high-tech firms might overcome the 1/n problem deriving from group incentive plans and foster the benefits of stock incentive plans. Team- or group-based organizational design establishes an organizational structure because a high level of granularity in the production chain is a critical strength of Taiwanese high-tech firms (Fuller et al. 2003). Therefore, we propose that the introduction of an employee stock incentive plan would generate positive effects on productivity.

Researchers have indicated that favourable accounting treatments stimulate overexploitation of broad-based stock incentives because firms perceive that these types of stock incentive carry lower costs than other compensation schemes do. The lack of mandatory expensing requirements for employee equity incentives may result in ineffective corporate governance and allow management to use employee equity incentives to extract excessive compensation (Guay et al. 2003). Therefore, from an economic-cost standpoint, overusing broad-based stock incentives may be an inefficient method to convey compensation, and to attract, retain and motivate employees (Hall and Murphy 2003). The ESB vesting period in Taiwan is particularly short. Thus, when the ESB plans pay a higher proportion of employee total compensation, employees might focus excessively on the wealth effects of their stock compensation. Extrinsic motivation may become more salient, thereby diminishing the intrinsic motivation effects of an employee ownership plan (Klein 1987). In other words, employees may be more interested in financial gain than in ownership, and the stock incentives may grant little incentive effects in aligning employee interests with those of shareholders (Oyer and Schaefer 2005; Rosen et al. 2005). Therefore, overexploiting broad-based stock incentives may function as an opposing force to organizational productivity, causing an inverted U-shaped relationship between the intensity of the broad-based stock incentives and organizational productivity.

3..2 Effects on Profitability

In addition to the possible productivity gains, can Taiwanese ESB schemes create more profits for firms? There are two sides to this theoretical argument. One emphasizes that employee equity incentives create productivity gains, reduce exit costs (such as training and replacement expenses), and consequently reduce indirect costs and improve profitability (Jones and Kato 1995; Klein 1987; Sesil et al. 2002). However, certain researchers have argued that employee equity incentives might increase labour costs and offset profit, consequently rendering employee equity incentives cost-ineffective. The increased costs may result from increased wages and intensive administration costs, including increasing skill training, communicating information and measuring team performance (Ben-Ner and Jones 1995). Empirical evidence has also demonstrated uncertain findings on the profitability effects of employee equity incentives. A study by Pugh et al. (2000) examined establishing employee stock ownership plans (ESOPs) and determined that they exert small, positive short-term effects on organizational accounting performance. Borstadt and Zwirlein (1995) examined 85 US publicly traded firms that established ESOPs and observed no profitability improvement after adopting ESOP. Frye (2004) determined that employee stock option plans exerted a positive effect on accounting returns in the early 1990s, but the performance effects became non-significant in their 1999 sample. Thus, firm profitability depends on whether the magnitude of productivity growth can offset the increased cost of employee incentives. These findings are consistent with the emerging conclusion in the literature that, whereas certain independent effects of employee ownership are evident, employee equity primarily works with a supportive corporate culture to impact productivity (Kruse et al. 2010).

Because Taiwanese firms have not been required to recognize the expense of ESB plans, it thus bears no accounting charge. Taiwanese ESB incentives are paid out through issuing new shares, resulting in the raise of total number of shares. Thus, the dilution effect may still exist, which should exert negative effects on profitability if measured as return on equity (ROE). When and if firms overuse broad-based stock incentives and issue excessive numbers of new stocks, the dilution effects may worsen. Thus, it is possible that productivity gains might be neutralized because of the dilution effects of the Taiwanese ESB plans, although firms have not been required to recognize the cost of ESBs. Thus, the negative impact of dilution effects on profitability may worsen when ESBs are overexploited.

3..3 Effects on Market Reaction

Finally, we argue that the effects of Taiwanese ESBs on market reaction should be more positive, although ESB plans result in a wealth transfer from shareholders to employees and cause a dilution effect (Ikäheimo et al. 2004). If the dilution effect outweighs the motivational effect and reduced costs, employee stock incentives will diminish shareholder returns. Therefore, broad-based employee equity incentives might not provide a positive shareholder pay-off (Callaghan et al. 2006). In the case of Taiwanese ESB, shareholders might not see the potential dilution effect of broad-based stock incentives because the costs of employee equity incentives have not been deducted from annual earnings; thus, profits may be overstated. Even if firms publicly disclose the details of employee equity incentives, shareholders or investors may fixate on reported earnings and ignore information regarding costs that financial statements do not explicitly recognize. Investors may consider using non-expensed ESB to relax firm financial constraints, allowing the firm to retain financial resources for other growth opportunities (Babenko et al. 2011; Core and Guay 2001). Therefore, investors could overprice stocks (Guay et al. 2003).

Employee equity incentives could also become a profitability signal for investors. The higher the amount employees obtain from a firm's stock incentive plan, the more the public believes in this firm's ability to create profits, and motivate and retain valuable talent (Conte et al. 1996). Equity incentive plans, thus, create a signalling effect for future growth and profitability. Bhagat et al. (1985) also indicated that the market might respond to employee stock incentive plans positively if it sees strong incentive effects for employees. Their empirical findings indicated that employee stock purchase plans exert a positive effect on shareholder wealth for reasons other than tax reduction. They argued that incentive or signalling effects are present in the case. This effect is more substantial in high-tech firms because investors are convinced that stock incentive pay plans can increase the value of the firm's intangible assets (Sesil et al. 2002). Hence, we argue that ESB in Taiwanese high-tech firms should exert positive effects on market reaction.

4. Methodology

4..1 Data Sources

Our dataset consisted of 844 public-traded high-tech firms in Taiwan, obtained from the Taiwan Economical Journal (TEJ) database, which is similar to the COMPUSTAT dataset. We also combined data from firm annual financial reports, obtained from the Market Observation Post Systems online services. We included both firms with and without ESB incentive schemes in our dataset.

The data consisted of pooled cross-sectional and time-series firm-level data for Taiwanese publicly listed, high-tech firms covering 12 years, from 1997 to 2008, a period of widespread ESB. After deleting observations with missing records, our actual dataset was an unbalanced panel consisting of 844 high-tech firms and 5,216 observations. These sample firms are classified into nine industries, namely integrated circuit (17.05 per cent), photonics (12.34 per cent), computers and peripheral equipment (15.6 per cent), telecommunications (10.14 per cent), electronic components and parts (21.72 per cent), other electrics manufacturers (8.57 per cent), electronics vendors and retailers (3.81 per cent), information technology services (4.96 per cent), and biotechnology (6.05 per cent). The average sale of sample firms was NT$10.6 billion, with an average employee base of 779 over the study period.

4..2 Measurements

(a) Dependent variables

To investigate the diverse effects of ESB on multi-dimensional organizational performance, we used the different performance indices of productivity, profitability and market reaction. We assessed firm productivity by measuring the value added, which attempts to capture ESB effect on productive efficiency. This was measured as gross income, which is sales revenue minus production costs (cost of sales), normalized (divided) by total assets. Value added provides a top-line view of a company's production or (in the case of a merchant) sales-related cost structure. It is a measure of how well (or badly) a company uses its capital, capacity and other resources, and shows its competitive strengths and weaknesses, compared with other companies in the same industry. Therefore, we used value added as a proxy to measure firm productivity in prior employee participation studies (Jones and Kato, 1993, 1995; Kato and Morishima 2002; Kruse 1993).

We measured profitability by ROE, which assesses organizational performance (i.e. net income) relative to the value of stockholder equity (retained earnings plus other equity) in the company. This accounting measure of performance is a valid measure of overall company performance (Jensen and Murphy 1990). Firms were not required to recognize ESB expenses during the period that our data covered, according to Taiwan accounting regulations. Thus, the ROE we used presented profitability before expensing the ESB cost. Market reaction is a forward-looking measure of firm performance, which captures firm-perceived growth opportunities and the expected future stream of earnings. We employed annual stock return as the measure of market reaction.

(b) Independent variables

We used two measures to assess the main independent variable, ESB, of this study. The first measure, ESBD, is a dummy variable that indicates whether the firm distributed stocks to employees as incentive rewards. This followed a similar research approach to Kruse (1993), in which ESB was measured as a dichotomous variable. The second independent variable, ESBR, assessed the intensity of ESB used as employee compensation. ESB intensity was calculated as the ratio of the market value of ESB to the total payroll, which was the sum of cash and stock bonuses and the year's fixed salary expenses. Both ESBD and ESBR were measured by year t−1.

(c) Control variables

The variables K and L represented the capital and labour inputs, respectively. We measured the proxy of capital input as the ratio of fixed assets to total assets, with labour input being the ratio of the number of employees to total assets. Because our study sample features all high-tech companies, we controlled for R&D intensity (R&D) when estimating the performance effect. We measured R&D intensity as the ratio of R&D expenditure to net sales and controlled for firm debt ratio, which was calculated as the ratio of total debt to total assets, denoted as DEBT. Considering that firms may use incentive plans as a means to provide financing (Babenko et al. 2011; Core and Guay 2001), we included firm investment rate, measured as the ratio of long-term investments to total assets (LINVEST), to control for potential differences in performance among firms with different investment rates. Finally, we adopted 11-year dummy variables to capture technological change and other shocks that are common to all firms during the 12-year period. Table 1 presents the definition of each variable used in this study.

Table 1. Variable Definitions
VariableDefinition
Dependent variables 
QtProductivity. Value added in year t
ROEt (%)Profitability. The ratio of net income relative to the value of stockholders' equity in year t
Stock ReturntMarket response. Annual stock return in year t
Independent variables 
ESBDt−1The usage of employee stock bonuses in year t−1, where the use of ESB was measured as a dichotomous variable
ESBRt−1Percentage of market value of employee stock bonus to total payroll expenses in year t−1
ESBSQRt−1The squared value of the variable ESBRt−1
Control variables 
Kt (%)Capital intensity. The ratio of fixed assets to total assets
Lt (%)Labour input. The ratio of the number of employees to total assets in year t
R&Dt (%)R&D intensity. The ratio of R&D expenditure to net sales in year t
DEBTt (%)Leverage intensity. The ratio of total debt to total assets in year t
LINVESTt (%)Investment rate. The ratio of long-term investments to total assets in year t

4..3 Regression Models

Because our dataset was longitudinal, it granted more confidence to examine the link between ESB and organizational outcomes. The other benefit is that multiple observations of the same firm allowed us to control for certain unobserved characteristics of individual firms. Because we could not assume that the observations are independently distributed across time, correction of default ordinary least square standard errors was necessary. We used a fixed-effects model to control the possible heterogeneity after conducting the Hausman test to differentiate between the fixed-effects model and the random-effects model in the panel data. In this case, the fixed-effects model provides consistent estimates and is preferred. When using the fixed-effects model, we assumed that something within the individual observation may affect or bias the predictor or outcome variables, and it was necessary to control for this. The model also removes the effect of time-invariant characteristics from the predictor variables, enabling us to assess the predictors' net effect and to control for differences among firms in managerial abilities and worker quality (Kato et al. 2005). In addition, for robust t values of the estimated coefficients, we report the standard errors that are clustered at the firm level.

For each dependent variable (productivity, profitability and market reaction), we estimated three models. In the first model, we tested the effects of a one-year lag in adopting an ESB scheme (i.e. ESBDt−1) on the dependent variable. In the second model, we tested the effects of a one-year lag of ESB intensity (i.e. ESBRt−1). In the third model, we examined whether overexploitation of ESB exerted an effect on organizational performance. We added the squared value of ESB intensity (ESBRSQRt−1) to the model. All three models included K, L, R&D, DEBT, LINVEST and year dummies as control variables. We also created a two- and three-year lag of the independent variables, that is ESBDt−j and ESBRt−j for j = 2 and 3, to examine the possible longer-lag effects of ESB on organizational performance.

5. Results

5..1 Descriptive Statistics

Table 2 presents a summary of the sample firms' descriptive statistics of ESB incentives and firm demographics in different periods from 1997 to 2008. The sample firms have larger total assets and numbers of employees before 2000 compared with the other periods. Table 2 shows that the total assets and employment size were larger in the late 1990s than in the other three periods. This might be caused by the increasing number of high-tech firms going public after 2000 because the newly IPO firms were typically small in total assets and employment size. We also observed growing net sales and R&D intensity over time, which demonstrates that Taiwanese high-tech industries were flourishing after the late 1990s.

Table 2. Descriptive Statistics for Employee Bonuses and Firm Characteristics in Selected Time Period, 1997–2008
 1997–20002001–20042005–20081997–2008
  1. aAs in NT thousand dollars.
  2. ESB, employee stock bonus; ROE, return on equity.
Employee stock bonus    
ESB adoption0.72340.59720.42170.5111
(0.4478)(0.4906)(0.4939)(0.4999)
ESB valuea703,564177,792178,905226,199
(2,841,723)(870,316)(1,311,046)(1,406,431)
ESB intensity0.29270.18550.13710.1684
(0.2928)(0.2334)(0.2156)(0.2343)
Firm demographics    
Total assetsa15,200,00010,800,00011,500,00011,600,000
(36,500,000)(37,800,000)(47,400,000)(43,300,000)
Employment size1,329755705779
(2,703)(1,991)(1,953)(2,052)
Net salesa10,100,0008,441,13212,000,00010,600,000
(18,700,000)(26,700,000)(59,400,000)(47,400,000)
ROE0.03970.03880.02040.0286
(0.1495)(0.2311)(0.4866)(0.3906)
Stock return0.09510.06550.09300.0835
(0.8513)(0.6258)(0.7718)(0.7316)
R&D intensity0.03900.06640.12110.0943
(0.0430)(0.4246)(2.7949)(2.0989)
Leverage intensity0.35900.37430.34600.3572
(0.1439)(0.1507)(0.1591)(0.1554)
Investment rate0.20720.18190.21570.2030
(0.1505)(0.1482)(0.1848)(0.1704)

Importantly, Table 2 indicates the change of ESB plans from the late 1990s to the late 2000s. We observe a continuous drop in adopting ESB incentives from the late 1990s to the present. The intensity of ESB usage also declined over time, from 29.27 per cent in 1997 to 13.71 per cent in 2008, indicating that Taiwanese publicly held high-tech firms did not rely on stock shares as employee incentives to the degree that they did in the late 1990s. However, ESB remained extremely popular for Taiwanese high-tech firms, with the average usage rate at approximately 51.11 per cent, and the proportion of ESB to total compensation at approximately 16.84 per cent.

5..2 Organizational Performance Regressions

Table 3 reports the fixed-effects regression results of the one-year lag effects of ESB on organizational performance. The performance indices were productivity, profitability and company stock return. For each performance index, we presented three model specifications (outlined in the previous section) to estimate and test the robustness of the results. Models 1, 4 and 7 estimated the performance effects of ESB by using a dummy variable (ESBDt1), which denoted the presence of an ESB plan in the previous year. Models 2, 5 and 8 presented the effects of ESB intensity (ESBRt1) by measuring the proportion of ESB to the total compensation cost in the previous year. Finally, Models 3, 6 and 9 examined the possible curvilinear relationships between ESB intensity and organizational performance indices.

Table 3. Performance Effects of Employee Stock Bonuses: Fixed-Effect Regression Results
 ProductivityProfitabilityMarket reaction
Value addedReturn on equityStock return
Model 1Model 2Model 3Model 4Model 5Model 6Model 7Model 8Model 9
  1. Notes: Numbers in parentheses are values of robust standard error clustered at the firm level. See Table 1 for variable definitions.
  2. +p < 0.10; * p < 0.05; ** p < 0.01; *** p < 0.001.
Kt−0.194***−0.188***−0.188***−0.375***−0.393***−0.394***−1.1132***−1.159***−1.157***
(0.019)(0.019)(0.019)(0.089)(0.090)(0.091)(0.165)(0.168)(0.166)
Lt317.999***338.336***338.329***−1313.850−1374.940−1374.688419.961353.177352.665
(33.821)(33.690)(33.709)(1046.110)(1069.285)(1068.542)(294.966)(302.498)(300.665)
R&Dt−0.00001***−0.00001***−0.00001***−0.00002**−0.00002**−0.00002**8.56e−68.41e−68.51e−6
(1.53e−6)(1.50e−6)(1.50e−6)(7.79e−6)(7.90e−6)(7.94e−6)(9.23e−6)(9.26e−6)(9.18e−6)
DEBTt−0.101***−0.097***−0.097***−0.987**−1.005**−0.994**−0.197+−0.194+−0.215*
(0.013)(0.013)(0.013)(0.326)(0.331)(0.330)(0.107)(0.110)(0.109)
LINVESTt−0.178***−0.178***−0.178***−0.245*−0.241*−0.238*−0.583***−0.593***−0.600***
(0.017)(0.016)(0.016)(0.101)(0.101)(0.101)(0.131)(0.133)(0.131)
ESBDt−10.011***  0.013  −0.176***  
(0.002)  (0.011)  (0.030)  
ESBRt−1 0.054***0.046** −0.1010.187** −0.360***−0.937***
 (0.008)(0.016) (0.064)(0.063) (0.083)(0.169)
ESBSQRt−1  0.012  −0.437***  0.874***
  (0.023)  (0.146)  (0.232)
Year dummiesYesYesYesYesYesYesYesYesYes
Observation5,2165,2165,2165,2165,2165,2165,2165,2165,216
Firm844844844844844844844844844
R20.34110.37820.37800.04970.04490.05040.23040.22640.2263

One of the more robust findings in the literature on employee ownership effects is the positive effect of employee ownership on productivity (Kruse 2002). Our results also indicated that ESB incentives exert significant positive effects on firm productivity. The results presented in Table 3 (Model 1) indicate that firms with ESB plans exhibited higher productivity than did firms without ESB plans (β = 0.011, p < 0.001). The results from Model 2 also demonstrate significant and positive associations between ESB intensity and firm productivity in value added creation (β = 0.054, p < 0.001). These results are consistent with the findings reported by Sesil et al. (2002), indicating that broad-based stock options enhance firms' value added creation activities. They also indicate that ESB, as a broad-based stock incentive similar to other employee equity compensation schemes (i.e. ESOPs, employee stock options), can tie employee motivation with compensation to increase productivity. These results support the hypothesis that broad-based stock incentives positively influence the productivity of Taiwanese high-tech firms. However, we did not observe a non-linear relationship between ESB intensity and organizational productivity. This may indicate that the overexploitation effects on productivity are not substantial.

Beyond statistical effects, the practical effects of ESB are also essential considerations for broad-based stock incentive research. According to Model 1, the estimated coefficient on the ESB dummy is 0.011, which means that introducing ESB in the present year increases productivity (ratio of value added to total assets) the following year by 0.011. For a sample firm with productivity in the sample median, in which the ratio of value added to assets was 0.16, the increase of 0.011 represents 6.9 per cent of productivity growth. The findings from Model 2 also suggest that the increase of 1 per cent ESB to total compensation raises the 0.054 per cent increase in value added.

Findings from Table 3 also indicate the profitability effects of employee stock incentives. In Model 4, the findings indicate that the presence of an ESB plan (ESBD) was unrelated to the firm's ROE (β = 0.013, n.s.). In Model 5, ESB intensity also did not correlate with the ROE measure (β = −0.101, n.s.), indicating that productivity gains did not lead to greater profitability. However, the results indicated an inversed U relationship between ESB intensity and ROE. In Model 6, the effect of ESBRt1 on ROE was significantly positive (β = 0.187, p < 0.01), with ESBSQRt1 having a negative coefficient (β = −0.437, p < 0.001). This might show that the positive productivity effect did not sustain for accounting profitability. As the proportion of ESB to total compensation of a firm increased, the dilution effects might diminish profitability, while ESB incentives were not expensed. Moreover, overexploitation effects existed for profitability. The newly issued stocks from ESB plans considerably diluted the increased possible profitability from productivity gain, although the ESB was not recognized as an operating expense.

Models 7–9 also showed the effects of ESB on market performance measured by share return. The results showed that the coefficients for the ESBDt1 and ESBRt1 variables were both negatively significant (Model 7, β = −0.176, p < 0.001; Model 8, β = −0.360, p < 0.001). Model 9 presents ESB with a U-shaped relationship with stock return, where the ESBRt1 coefficient was negatively significant (β = −0.937, p < 0.001) and the ESBSQRt1 coefficient was positively significant (β = 0.874, p < 0.001). The results contradict our prediction that the decision to pay employees equity compensation might convey optimistic information regarding the firm's growth and profitability to investors, and they might react to the market positively. However, investors may harbour concerns regarding the hidden costs of ESB and the dilution effects of newly issued stocks. The results also suggest that the negative market reaction of ESB decelerates when the ESB intensity levels are high.

Studies have suggested that reverse causality may exist between organizational performance and broad-based equity incentives. We conducted two tests to examine the possibility of endogeneity between organizational performance and ESB schemes. We controlled for past performance (stock return in the previous year) in the models to control for possible effects and determined that the coefficients of employee stock-incentive variables in the estimated models remained in similar directions to the findings of the one-year lag model in Table 3. Following Jones and Kato (1995), we also re-estimated the models by adding a one-year lead of the ESBR variable as an additional right-hand side variable into each model. The estimated coefficients on ESBRt+1 were significantly positive, suggesting possible positive causality effects. While controlling for ESBRt+1, the estimated coefficients on ESBRt−1 remained significantly negative in profitability and stock return regressions. Hence, we should not overemphasize the reverse causality effect.

We also present two- and three-year lag effects of ESBD and ESBR on organizational performance in Table 4. In contrast, the findings indicated that ESB schemes exerted significantly negative effects on organizational performance, including productivity, profitability and share-turn in the following two and three years. The results demonstrate that ESB effects are short term in Taiwanese ESB high-tech firms. To explore the issue of short-term effects, we split the sample based on whether the firms had ESB in year t−2 and examined the one-year lag ESB effects. We observed that when firms did not adopt ESB in year t−2, ESB use in year t−1 did not affect the organizational performance. However, firms using ESB in both years exhibited enhanced productivity compared with firms that terminated ESB in year t−1.4 This might indicate that firms should use ESB for a period of time to obtain performance gains. In the nature of employee ownership, ESB is a long-term incentive scheme. Giving away ESB erratically will not facilitate organizational performance. In summary, firms might take advantage of non-expense ESB incentives to create short-term motivation for employees; however, neglecting the inherent long-term character of ESB and overusing the schemes might damage subsequent performance.

Table 4. Performance Effects of Employee Stock Bonuses: Two-Year and Three-Year Lag Regression Results
 ProductivityProfitabilityMarket reaction
Value addedReturn on equityStock return
Model 1Model 2Model 3Model 4Model 5Model 6Model 7Model 8Model 9
  1. Notes: Numbers in parentheses are values of robust standard error clustered at the firm level. See Table 1 for variable definitions.
  2. +p < 0.10; * p < 0.05; ** p < 0.01; *** p < 0.001.
Two-year lag         
ESBDt−2−0.007**  −0.023+  −0.220***  
(0.003)  (0.012)  (0.029)  
ESBRt−2 −0.014+−0.041** −0.088*−0.109+ −0.706***−1.112***
 (0.008)(0.015) (0.043)(0.058) (0.077)(0.166)
ESBSQRt−2  0.039+  0.031  0.607**
  (0.020)  (0.082)  (0.222)
Year dummiesYesYesYesYesYesYesYesYesYes
Observation4,3724,3724,3724,3724,3724,3724,3724,3724,372
Firm777777777777777777777777777
R20.30610.30090.30110.05270.05290.05270.23840.22690.2285
Three-year lag         
ESBDt−3−0.011***  −0.034  −0.099**  
(0.003)  (0.021)  (0.033)  
ESBRt−3 −0.028***−0.061*** −0.119*−0.160+ −0.384***−0.431*
 (0.008)(0.017) (0.046)(0.084) (0.076)(0.186)
ESBSQRt−3  0.049  0.062  0.071
  (0.025)*  (0.125)  (0.225)
Year dummiesYesYesYesYesYesYesYesYesYes
Observation3,5953,5953,5953,5953,5953,5953,5953,5953,595
Firm722722722722722722722722722
R20.30550.28310.28430.06070.06140.06110.25800.25350.2538

6. Discussion

Using a unique longitudinal dataset, we elucidated the effects of non-expensed, broad-based stock incentives on key organizational performance indices. The findings may reconcile certain debates surrounding the effectiveness of broad-based stock incentives. First, our findings provide evidence to support the efficiency arguments of broad-based stock incentives. Based on a unique longitudinal dataset from Taiwanese high-tech firms, we observed that broad-based stock incentives exerted positive effects on value added creation. This result is consistent with most prior literature in that broad-based stock incentives are crucial compensation schemes to attract, motivate and retain employees (Core and Guay 2001; Sesil et al. 2002). However, we observed that Taiwanese employee stock incentives did not exert deferred effects on productivity. Firms dispensed non-expensed ESB with short vesting periods in lieu of cash compensation to motivate employees; however, this motivation did not last long. We also observed that firms sporadically adopting ESB cannot even enjoy such a short-term productivity effect. Although our results confirmed that Taiwanese high-tech firms considered broad-based stock incentives as crucial compensation schemes to generate greater organizational productivity, these firms neglected the management complexity of ESB incentives. The productivity effect may diminish if firms use them as short-term schemes, and the intermittent distribution of ESB may cause two disadvantages. First, it may convey inconsistent signals to employees and impede the co-operation culture (Bowen and Ostroff 2004). Second, the frequent changes of incentive programmes may increase overall adjustment costs (Hannan and Freeman 1984). Both disadvantages might cause the short-term productivity effects to fade for these firms.

Although we found that non-expensed ESB incentives exerted significant and positive effects on short-term productivity, it remains unclear whether using non-expensed ESB incentives is beneficial to business. The positive effects on ROE or return on assets (ROA) might result only from a smaller amount of wage expense, and thus should mechanically result in higher profit. In contrast to prior studies on broad-based stock incentives, we observed that firms might overuse non-expensed employee stock incentives because of variable accounting treatments, resulting in negative effects. Taiwan's favourable non-expense accounting rules might render the perceived cost of broad-based employee stock incentives much lower than cash rewards, and certain firms might give stock incentives because of financial (cash) constraints. Thus, efficiency might not be the primary reason why firms use ESB incentives (Core and Guay 2001; Hall and Murphy 2003). These firms may grant broad-based stock bonuses that exceed an optimal level. Therefore, incentive schemes can become an ineffective method to convey compensation (Murphy 2003), which may subsequently result in dilution effects and damaged organizational profitability and market performance. Our findings provide evidence for these possibilities. Although it bears accounting charges and incurs no cash outlay, a large amount of newly issued stocks dilutes firm ROE. We reasonably estimate that organizational profitability would worsen if firms were required to recognize cost, causing incentive schemes to increase wage expense at the expense of profits (Ben-Ner and Jones 1995).

For the impacts of ESB on market performance, it is an empirical issue that depends on how shareholders and market participants perceive the net benefits of the incentive schemes. In our study, the stock market did not respond positively to the possible dilution effects. We observed negative effects of broad-based stock incentives on company share return, which means that the market might censor the implicit wage cost of broad-based stock incentives. The findings were contrary to the results of Park and Song (1995), who observed enhanced firm market value after they adopted ESOPs. Because of the short ESB vesting period and heavy trader speculation in the Taiwanese stock market (Barber et al. 2004) and wealth rebalancing motivation (Core et al. 2003), employees typically liquidate their granted stocks as soon as possible, which might render the dilution effects even more substantial in the stock market. Although only profitable firms are allowed to pay bonuses in Taiwan, which may convey a type of intangible information that may be associated with growth prospects (Daniel and Titman 2006), the hidden cost of ESB and the dilution effects might still lead to pessimistic market reactions (Elayan et al. 2005; Espahbodi et al. 2002).

Finally, the negative effects of non-expensed ESB plans were even more severe in the subsequent year, and overexploitation damaged the long-term organizational performance of Taiwanese high-tech firms. However, this notion has typically been neglected because of the apparent short-term productivity effects. Recent research on Taiwanese ESB incentives, such as that of Han and Shen (2007), has elucidated the positive effects of ESB schemes, but they have not discussed the possible negative aspects of the ESB system. The methodology problem might cause Han and Shen to obtain such optimistic results. They did not use fixed-effects technique to controlling for time-invariant unobserved firm characteristics. These characteristics, for example the quality of management and corporate culture, may affect the use of ESB schemes and productivity simultaneously. By controlling for the effects of unobserved firm characteristics through the individual-specific term of the fixed-effects regression, our study provided robust results that could address the potential negative effects, and mean that companies should be alert to potential problems.

7. Conclusion

Our findings show that Taiwanese high-tech firms attempted to use non-expensed ESBs to attract, retain and motivate employees to foster productivity. We observed that non-expensed ESB incentive schemes raised short-term productivity; however, organizational profits and stock returns faded because of dilution effects, particularly in the long term. We also observed that a short-term productivity effect was not promising for firms adopting ESB erratically. We argue that the effects of non-expensed ESB on organizational performance of Taiwanese high-tech firms are not positive in the long term.

The diverse effects on organizational performance might be the root cause of the controversial debates on using non-expensed ESB schemes and might highlight a mixture of divergent motives in an emerging new knowledge economy. Because profitability and operational productivity growth are essential business objectives, productivity growth is critical for firms at an emerging stage (Baum et al. 2001). Firms should ideally trade certain financial returns and transfer wealth to employees to motivate productivity. Over time, the trade-offs among diverse performance indices began to concern shareholders and the public. Prior studies have argued that certain firms frequently overexploit ESB incentives to the detriment of organizational efficiency (Hall and Murphy 2003). Our econometric results are in line with this argument. We observed that the overexploitation of non-expensed ESB incentives might impede profitability and stock returns, and produce worse multiple facets of long-term performance. The decoupling of ESB incentives and efficiency objectives may evoke a response from shareholders and the public, requesting greater transparency in financial reports (Brandes et al. 2005; Walters and Young 2008), which explains why non-expensed ESB incentives gradually lost their legitimacy in Taiwan. By influencing the regulatory context, the debate over controversial accounting for ESBs in Taiwan has recently ended. Beginning in 2009, Taiwanese firms have been required to recognize the cost of ESB as an expense and reflect employee compensation costs fully in their financial statements. This dramatic change in ESB regulations has introduced new challenges to firms in Taiwan. Firms must now establish a new compensation policy that is best suited to the corporation, and that will retain and motivate elite talent and create a triple-win scenario that benefits the enterprise, its shareholders and employees alike. However, considering the inherent lag between performance and compensation policies, investigating how the new regulations affect ESB practices and the compensation policies in Taiwanese high-tech firms must await future data.

Our study has several theoretical and empirical limitations. First, measurement problems must be considered. We are uncertain whether all publicly available information concerning stock profit-sharing plans covers most employees in a firm as opposed to selected employee groups. Second, ESB incentives are typically part of the implementation of an overall high performance work practices (HPWP) system. However, it is impossible to obtain HPWP information of all the firms in our data across time to form a panel dataset. Thus, HPWP might be an omitted variable that results in the overestimated effect of the ESB incentive, and this concern should be addressed.5 Finally, the dataset contains only publicly listed firms in the Taiwan stock exchange. Future studies should exercise caution when seeking to extend the validity of our results to small or privately held companies.

Acknowledgements

For helpful comments on drafts, we thank Joseph Blasi, Douglas Kruse and the participants of the 2010 Beyster Mid-year Fellows Workshop at Rutgers University. We also thank two anonymous reviewers whose critical comments have shaped this article. This study was funded in part by National Science Council of Taiwan (Grant Number 98-2410-H-008-040-).

Notes

  1. 1

    US high-tech firms typically use stock option plans to attract, motivate and retain employees. According to the summary by Kruse et al. (2010), 33 per cent of firms provided broad-based stock options for employees in the 2002 National Organizations Survey. The usage rate was apparently lower than Taiwanese statistics. Another study from Sesil and Lin (2011) indicated that the usage rate of the broad-based stock option was 23 per cent.

  2. 2

    Comparing two corresponding US firms, Qualcomm Inc. and Maxim Integrated, in integrated circuits industries, the cost of stock-based compensation was 26 per cent and 19 per cent, respectively, in 1999 (Hess and Lüders 2000). The ratio of ESB to net income was also approximately 1.5–2 per cent for Taiwanese firms in traditional industries from 2004 to 2007 (Liu and Lee 2009). Thus, UMC was regarded as having a generous ESB policy at that time.

  3. 3

    Local shareholders and foreign investors have become increasingly concerned regarding the expensing of employee equity incentives within Taiwanese technology firms. After a long process of revising laws, a new regulation passed the legislative body in May 2007 requiring firms to expense employee equity incentives at market value from 2008, a change that will make Taiwan's accounting system consistent with international accounting standards.

  4. 4

    The supplement analysis was conducted according to the suggestion of one of our reviewers. The analysis results are available upon request.

  5. 5

    According to the reviewer's suggestion, we attempted to understand the connection between the adoption of the ESB and HRM system in Taiwanese high-tech firms. One of our authors conducted a HPWP survey in Taiwanese listed firms in 2005. The dataset contained 114 high-tech firms, and 60 of them distributed ESB in that year. The ESB usage rate was 52.63 per cent, which was relatively close to the distribution of the sample frame, that is our present dataset. We also conducted analysis of variance tests to examine the differences in the use of work-team design, employee involvement practices, training, performance management and work–family initiatives between firms with and without ESB. We observed that firms with ESB exhibited a slightly higher score in these dimensions; however, the differences were not statistically significant.

Ancillary