Chapter 2: Literature review and hypothesis development
2.2 Organizational slack and Performance
2.2.1 The moderating role of absorbed slack
Recently, some studies identify that different types of slack resources, such as absorbed and unabsorbed and unabsorbed slacks, may have different effects on innovations (e.g., Greve, 2007). Absorbed slack is organizational slack in the form of administrative resources beyond what is necessary for the short-term operation and maintenance of the organization (Singh, 1986; Greve, 2007) and is directly and useful for developing innovations. For example, organizations with more absorbed slack
among other staff that are needed for stimulating innovations (Greve, 2003). Thus, absorbed slack in the evolution of a technology-based firm commonly acts as internal shock absorber to provide the needed resources to make the planned innovation projects progress over time and therefore achieve the benefits of economy of scope across different technology fields (Huang and Chen, 2010). In addition, absorbed slack not only includes large administrations, costly facilities, and high wage levels that are directly for innovation development, but also provides a cushion or buffer to integrate different technological knowledge across disciplinary frontier by resolving latent goal conflicts and reducing resource competition among the innovation projects and accordingly result in a higher level of cross-fertilization of heterogenous knowledge that are useful for developing innovations (Greve, 2003).
Since Greenhalgh (1983) hypothesized that organization slack would have a positive influence on innovation, further empirical studies have found the positive effects of slack on innovation and performance (Singh, 1986; Damanpour, 1987;
Bromiley, 1991; Zajac, Golden and Shortell, 1991; Majumdar and Venkataraman, 1993; Greve, 2003). Chen and Hambrick (1995) has found that organizational slack tends to suppress initiative actions and promotes responsive actions. Furthermore, organizational slack allows firms to respond in more creative ways (Smith et al., 1991). Therefore, based on the important and positive relationship among
technological diversity, absorbed slack, and innovation performance found by Huang and Chen (2010), this study extends Huang and Chen‟s research and hypothesized that absorbed slack positively moderates the relationship between technological diversification and firm performance.
Thus, for the technological diversification, this study propose absorbed slack positively moderates the relationship between technological diversity and innovation performance. The above reasoning leads to the following hypothesis:
Hypothesis 2: Absorbed slack will positively moderate the strength between the
technological diversification and firm performance.
2.2.2 The moderating role of unabsorbed slack
Unabsorbed slack is organizational slack in the form of financial reserves (Singh, 1986), which is more easily redeployed everywhere and allowing for greater managerial discretion (Tang and Peng, 2003). While absorbed slack is directly useful for developing innovations, unabsorbed slack is not directly helpful in the development of innovations. However, unabsorbed slack affects managerial decisions
performance monitoring of uncertain projects (Greve, 2007). While unabsorbed slack can stimulate the management to support new innovative activities (Thompson, 1967;
Nohria and Gulati, 1996; Huang and Chen, 2010), a higher level of unabsorbed slack may cause premature termination of innovation projects due to its strict performance monitoring (Greve, 2007).
In other words, before the management in a technology-based firm has accumulated enough experience to know whether they will eventually improve its performance (Lounamaa and March, 1987), a higher level of unabsorbed slack may cause inefficient problems in an over-diversified technology base (Tang and Peng, 2003; George, 2005; Huang and Chen, 2010). Therefore, this study extends Huang and Chen‟s (2010) research and hypothesize in absorbed slack negatively moderates the relationship between technological diversification and firm performance.
Hypothesis 3: Unabsorbed slack will negatively moderate the strength between
the technological diversification and firm performance
Chpater3: Research method
3.1 Research framework
Our three hypotheses are summarized in Figure 1. In the following section we shall confront this model with the empirical data.
Figure 3-1: The hypothesized model
3.2 Survey procedures and sample
This study is to investigate the relationship between the slack resource of firm, technological diversification, and performance. Thus, this research focus on there has more R&D capabilities and science-based of the industry in Taiwan. The total industry in Taiwan over the years have the most investment amounts is in the electronic industry. Especially in communication equipment industry in Taiwan, the design and production of world famous. In 2009, the communication industry of Taiwan had revenue about 28.5 billion U.S. dollars, ranked the 7th highest in the world. The output value of Taiwan's communications equipment for the 20.4 billion U.S. dollars (Industrial Development Bureau, 2010). Therefore, this study chooses smart phone industry-related firms as of the object, use the TEJ database, selected the study period from 2004 to 2009, total of 7 years. This study selected 55 smart phone industry-related firms from listed firms in Taiwan. According to the classification of Taiwan Stock Exchange Corporation (TSE), including the 5 firms in semiconductor industry, 17 in photovoltaic industry, 3 in other electronics, 3 in communication network industry, 1 in electronic channel industry, 18 in electronics components industry, and 8 in Computer and peripheral equipment manufacturing.
Table 3-1 Taiwanese smart phone industry classification and the number of
sample firms
IT industry classification No. of firms
Semiconductor 5
Photovoltaic 17
Other electronics 3
Communication network 3
Electronic channel 1
Electronics component 18
Computer and peripheral equipment 8
This study collected the patent data of 55 sample firms from the database provided by the U.S. Patent and Trademark Office (USPTO). Data collected period from 2003 to 2008, total of 3,635 patents to calculate technological diversification.
Exclude the missing data, this study totally collect 294 samples. The USPTO oversees the process of granting property rights to inventors for inventions that are „useful‟ and
available continuously across time. A number of previous scholars have used patent data as a proxy for innovation (Argyres and Silverman, 2004; Fleming and Sorenson, 2004; Gittelman and Kogut, 2003; Henderson and Cockburn, 1994; Rosenkopf and Nerkar, 2001). Patent file lists the corporation and business unit that applied for each patent, the technological class to which each patent belongs. In addition, a patent document contains a list of citations to other patents which represent the technological antecedents to the particular innovation. Thus, we use U.S. patent data to compare the performance of patent diversification of Taiwanese smart phone industry.
3.3 Methodology
3.3.1 Hierarchical regression analysis
Hierarchical regression analysis method is take the single effect of important predict variable successively put in regression model. In order to understand the various class of the regression model to total explained variance capacity of dependent variable and individual prediction of variables and its explanation of variability.
This study used hierarchical regression analysis to examine firms from 2003-2008. Try to find the relationship between technological diversification, organization slack, and firm performance. Consider the resources, structure, and strategy of firms may impact on the performance of a deferred nature, so the strategy of the company during the performance of the next variable t, used t +1 period.
This study try to examine Hypothesis 1, 2, and 3 using following regression:
E(8)
E(12)
: Firm size (number of employee) of firms during t period.
: R&D intensity of firms during t period.
: Absorbed slack of firms during t period.
: slack of firms during t period.
: Technological diversification of firms during t period.
: ROA of firms during t+1 period.
: Tobin‟s q of firms during t+1 period.
: Market Value Added of firms during t+1 period.
: Economic Value Added of firms during t+1 period.
3.4 Measure of variable
3.4.1 Organizational slack
Organizational slack has been widely conceptualized into two primary dimensions: absorbed slack and unabsorbed slack (Sharfman etal.,1988; Tan and Peng, 2003). Absorbed slack is due to the hard-to-redeploy nature whereas unabsorbed slack reflects the easy-to-redeploy elements of the organizational resources (Singh 1986;
Tan and Peng, 2003). This study following Tan and Peng (2003) and Huang and Chen (2010), this study use three items: major repair fund, inventory fund, and accounts payables, to measure absorbed slack. The absorbed slack dimension is measured as the sum of the standardized estimations of these three items. Likewise, unabsorbed slack is measured as the sum of the standardized estimations of five items:
depreciation fund, reserve fund, loans, sales expenses, and retained earnings.
3.4.2 Technological diversification
The entropy measure of technological diversification is derived from product diversification (Jacquemin and Berry, 1979; Palepu, 1985). Combining the entropy measure of diversification and technological diversity proposed by Miller (2006) and Chen (2009) followed to calculate the index of technological diversification.
This study collect data from the United States Patent and Trademark Office, and to take patent data of smart phone industry manufacturers in Taiwan from 2003-2008. The classification of data is according to the U.S. Patent Office Patent classification code (US_Class) provided, taking patent classification codes in the first four digits of UPC codes as the class number to calculated value patent diversification.
Each cited patent is identified by U.S. Patent Class (UPC) and related to the distribution. Adding up within 4-digit patent codes gives a citation-weighted count of
„patent equivalents.‟ Made most in number of (core patents) to and for each value of
the difference, if any patent equivalent in the same 4-digit SIC as the core patent is assigned a zero, the counts sharing the same 3-digit SIC code are assigned a one, the same 2-digit code a two, the same 1-digit code a three, and in different 1-digit codes the patent equivalents are assigned a four. To summarize, the index is a measure of the
dispersion of patent applicability across firms, with those patents weighted by adjusted citation counts and depreciated over time. The variable may be higher because the firm has many different classes of patents. The variable as , i is the core patent and j to measure patent dispersed in the enterprise level. If the business is more focused on the development of the patent category, then the patent values will be lower, if the company adopted a strategy to technology diversification, the development of many types of patents, the values will be higher.
For example, a firm have 5 patents in 2008:
Table 3-2: The calculation of technological diversification
Code Difference Sum T.D. index
2211 3 11 2.2
2143 0 2143 0 3706 4 3051 4
The most frequent U.S. patent code is 2143, so used it as core patent of firm.
3.4.3 Return on total assets (ROA)
ROA is to measure the efficiency of the use of asset management companies.
It‟s also called DuPont Analysis. Measure of the study variables using the following formula:
ROA = [Net income + Interest expense (1-25%)] ÷ Average total assets
3.4.4 Tobin’s q
Lang and Stulz (1994) are using Tobin's q as a measure of corporate performance variables. In addition, the evidence also shows that the Tobin's q measure of market value, not only stable but also fit in response to the revenue of R&D activities (Ayadi, Dufrene, and Obi, 1996).
Tobin's q used the company's market value as the numerator, the company's replacement cost of tangible assets are calculated as the denominator of the ratio. The higher the value of intangible assets of the company, Tobin's q values are more higher.
That is, when firms have the stronger monopoly power, value of goodwill, and skilled manager more, the company's Tobin's q values are also larger. However, due to the complexity estimates of the replacement cost of assets must be calculated separately
inventory, land, plant, and other equipment, then must be price inflation. Other factor need to be concerned, like the real rate of depreciation, the value of such capital expenditures and investment into consideration. Thus, this study refer to Chung and Pruitt (1994), developed a relatively simple approximate Tobin's q (approximate q) of the formula, simply through the basic financial and accounting information can be
calculated.
Approximate q =
3.4.5 MVA
Mentioned in Fortune magazine: MVA can show some of the increase in equity, so the MVA is an external measure of financial performance. It‟s also a good measure of overall performance to firms. MVA reflects the company's overall operating performance and make business aggregated its limited resources distribution and adequate treatment management. To create the greatest value of the enterprise, thereby enabling the shareholder wealth maximization, and often applied by scholars at the attention and use. Lehn and Makhija (1996) point out that the traditional performance evaluation assessment indicators have neglected the cost of surplus investment funds, while the MVA but took into account the investment cost of
premium of shareholder value is the measure of the market value of equity less the book value of equity, while the MVA is the market value of equity and debt net of equity and debt book value of the economy. The company's MVA and can be expressed as equity MVA (market value of equity less book value of equity of the economy) and liabilities of the MVA (market value of debt less liabilities of the economic book value of debt).
The market value increases when the company implemented a NPV> 0 the project case, will increase its market value, NPV calculation is the company's future cash amount of the discount period to reduce the sum derived from the original capital invested. MVA calculation, the capital's economic book value is equivalent to the investment company's past and present invested capital combined. NPV method of the future cash flows discounted value of the aggregation, that is the market value of capital of the firm by the MVA method. In other words, the MVA of firm increase, that means the company has invested NPV> 0 of the plan.
MVA＝Market Value of equity－Book Value of equity
3.4.6 EVA
Economic Value Added (EVA) was developed by the New York Stern Stewart & Co. financial consultants for the purpose to firm's financial performance
measures. The measure was derived from the concept of residual income. EVA emphasizes the firm should earn more than the return of the cost of capital to create shareholder value (Stewart and Bennett, 1990; Stern and Joel, 1993). In general, the increase in shareholder value is the value from the enterprise economy value creation, and economic value of the enterprise can through increase the profitability, improved working capital management, or effective project investment to be created. Thus, the EVA is used to measure a certain period in the economic value created is higher than the cost of capital assets that use as a specialized technique (Bennett and Linda, 1995;
Fisher, 1995; Grant, 1996).
EVA has moved from academic jargon evolved into a financial phenomenon. EVA is a performance management as an indicator, can use in various departments, project managers and the performance evaluation of each firm, and it is also has the effect of incentives. As long as the firm adopted to EVA, the various departments in order to achieve better performance that will try to reduce capital costs, improve return on invested capital (ROIC) and weighted average cost of capital (WACC) the difference between make more efficient use of capital, and economic value added increased. Therefore, value-based as the management foundation of the firm, will use EVA as a measure of corporate value creation assessment system to determine the management of funds, long-term financial planning, management
objectives, performance measurement, shareholder communication and incentive pay and other issues. According to scholars of American and British studies that the stock market, EVA is the stock price changes and stock prices and the relationship between the performance measures most closely.
EVA= (Return on Invested Capital – Cost of Capital) × Beginning Total Invested
Capital
= EBIT(1－Tax)－WACC × (Total assets － Current liabilities)
Weighted Average Cost of Capital
＝ ＋ ＋
＝ － ＋ ＋
Cost of equity Capital： This study used capital asset pricing model to approach cost
of equity.
=
＋ β × (
－ )Risk-free interest rate is using rate of Taiwan Treasury bill for one year period.
Risk interest rate is using Taiwan market rate weighted index from TEJ database.
β: With the firms since listing to 2004 years until the rate of return on stocks of the
Taiwan Stock Exchange on the weighted stock index return of the regression coefficients.
3.5 Control variable
3.5.1 Firm size
Firm size has long been found to be an important factor affecting firm survival and performance (Porter, 1980). (Hitt et al., 1997) indicated that firm size is associated with economies of scale and, hence, is expected to have a positive association with firm performance. Follow Hitt et al. (1991), this study using number of employees to measure firm size.
3.5.2 R&D intensity
Total R&D expenditures divided by total sales is the most commonly used measure in the studies of R&D intensity (Hambrick and MacMillan, 1985; Baysinger and Hoskisson, 1989).
Chapter 4: Empirical Results
4.1 Descriptive statistics
Table 4-1: The number of patent of Taiwanese smart phone industry from
2003-2008:
Annual 2003 2004 2005 2006 2007 2008 Number
of patent
255 388 388 640 834 1,130
Table 4-1 presents the number of patents and its trend from the sample firms of this study during the year 2003 to 2008. The trend of annual patent numbers can be understood from table 4-1, the number increased year by year in Taiwan, such as a patents in 2008 had a 35.49% growth in the number.
Table 4-2: Notice the technological diversification index of Taiwanese smart
phone industry in from 2003-2008:
Annual 2003 2004 2005 2006 2007 2008 Patent
diversification index
1.03 1.09 1.16 1.22 1.27 1.49
Table 4-2 and figure 4-2 show the technological diversification index and its trend during the year 2003 to 2008.
Table 4-3: Descriptive statistics:
Minimum Maximum Mean Standard
deviation
49 550,000 18,180.87 53,019.71
0.00 37.24 3.79 4.92
-12.22 25.90 0.06 2.88
-2.31 35.62 -0.13 4.46
0 4 1.15 1.31
-33.31 47.34 7.71 10.72
0.20 1.81 0.71 0.22
-388,921,660 2,827,978 -21,462,402.89 49,073,483.49
-384,515,753.47 288,678,497.12 -1,143,911.62 40,254,256.35
The total sum of sample is 282.
Table 4-3 presents descriptive statistics including means, standard deviations, maximum, and minimum for all measured variables in this study.
4.2 Pearson correlation analysis
Table 4-4 Pearson correlation coefficient analysis:
1 2 3 4 5 6 7 8 9
1
-0.14* 1
3. 0.84** -0.12* 1
4. 0.85** -0.12 0.83** 1
5. 0.01 -0.06 0.14* 0.16** 1
6. 0.00 -0.05 -0.02 0.03 0.00 1
7. -0.08 -0.14* -0.04 -0.11 -0.09 -0.63** 1
8. -0.84** 0.11 -0.81** -0.94** -0.13* -0.11 0.18** 1
9. -0.03 -0.00 -0.14* -0.10 -0.01 0.10 -0.03 0.11 1
* Correlation is significant at the 0.05 level (2-tailed)
** Correlation is significant at the 0.01 level (2-tailed)
N=282
Table 4-4 show all variables in this study using Pearson correlation coefficient matrix. Through the correlation coefficient can understood the correlation and direction with individual variables.
4.3 Measurement model
Table 4-5 Results of regression analysis (Technological diversification on
dependent variable: ROA, Tobin‟s q, MVA, and EVA)
Independent
Table 4-5 displays the results of regression analyses regarding the effects of technological diversification on firm performance that using four type of indicator:
ROA, Tobin‟s q, MVA, and EVA. Each of Performance indicators has four model:
Respectively, Model 1 is the base model that includes two control variables: Firm size and R&D intensity. Models 2 try to capture the direct effect of technological diversification on the dependent variable.
Table 4-5 displays the results of the effects of technological diversification and on four types of performance indicator. For ROA, E(1) indicates that the
combination of control variables does not have significant impact on the dependent
variable (F =0.32, =0.00). It show that the control variable don‟t have directly
effect on ROA. E(2) does not have significant and can explain an additional 0.2% of variance over what the control variables alone explain. The coefficient of
technological diversification is negative and not significant. For Tobin‟s q term, E(5) shows that the combination of control variables has significant impact on the
dependent variable (F =4.31, P<0.05, =0.03). It show that the control variable have directly effect on Tobin‟s q. E(6) and have significant and can explain an
additional 1% (F =3.83, P<0.01) of variance over what the control variables alone explain. The coefficient of technological diversification is negative (0.1, P<0.1) and significant. The finding indicate that technological diversification has an negative relationship with firm's performance, the results in Tobin‟s q terms support Hypothesis
1. For MVA terms, E(9) indicates that the combination of control variables have significant impact on the dependent variable (F =323.01, P<0.01 =0.70). It show
that the control variable have directly effect on MVA. E(10) have significant and can explain an additional 1.5% (F =230.81, P<0.01) of variance over what the control variables alone explain. The coefficient of technological diversification is negative
(-0.12, P<0.01) and significant. The finding indicate that technological diversification has an negative relationship with firm performance, the results was support
Hypothesis 1. For EVA term, E(13) indicates that the combination of control variables does not have significant impact on the dependent variable (F =0.09, =0.001). It show that the control variable don‟t have directly effect on EVA. E(14) does not have significant and can explain equal to what the control variables alone explain. The coefficient of technological diversification is negative and not significant.
Followed Table 4-5, the second part, this study examines the contingent role of organizational slack between technological diversification and firm performance.
Model 4 adds the two dimensions of organizational slack: absorbed slack and unabsorbed slack and their two interaction terms with the technological diversification dimension. All the results was shown from table 4-6 to 4-9.
Table 4-6: Results of regression analysis (Dependent variable: ROA) Independent variable Dependent variable: ROA
In the term with the dependent variable of ROA, table 4-6 displays the results of the effects of technological diversification and organizational slack on ROA.
The E(4) is not significant (F =0.64, =0.02) and explains an additional 1.6 percent of variance over what the control variables alone explain. The coefficient of the
interaction term between technological diversification and absorbed slack in Model 4 is negatively signed and not significant. The absorbed interaction term was inconsistent with the predict positive signed. The coefficient of the interaction term
interaction term between technological diversification and absorbed slack in Model 4 is negatively signed and not significant. The absorbed interaction term was inconsistent with the predict positive signed. The coefficient of the interaction term