THE IMPACT OF INFORMATION TECHNOLOGY ON FINANCIAL PERFORMANCE: THE IMPORTANCE OF STRATEGIC CHOICE

Transcription

1 THE IMPACT OF INFORMATION TECHNOLOGY ON FINANCIAL PERFORMANCE: THE IMPORTANCE OF STRATEGIC CHOICE Namchul Shin School of Computer Science and Information Systems Pace University One Pace Plaza New York, NY Phone: Fax: European Journal of Information Systems, vol.10, no.4, 2001

2 THE IMPACT OF INFORMATION TECHNOLOGY ON FINANCIAL PERFORMANCE: THE IMPORTANCE OF STRATEGIC CHOICE Abstract Information technology (IT) may not automatically improve firm profitability. It is an essential tool, but not sufficient in itself, and should therefore be coupled with organizational factors such as business strategies. A firm can maximize the value from its IT investments by aligning them with business strategies because IT improves scope economies and coordination. This paper examines empirically the contribution of IT to financial performance as measured by net profit, ROA, and ROE by focusing on the alignment of IT with business strategies such as vertical disintegration and diversification. Empirical analysis shows that IT does not directly improve financial performance. In conjunction with vertical disintegration and diversification, however, it does improve financial performance as measured by net profit. Financial performance ratios such as ROA and ROE, however, are not correlated with the alignment (or interaction) factor of IT with vertical disintegration and diversification. The results indicate that increased IT spending improves net profit, but not performance ratios such as ROA and ROE, of firms with decreased vertical integration and higher diversification. KEY WORDS: Information Technology, Financial Performance, Vertical Disintegration, Diversification, Strategic Choice 1

3 1. INTRODUCTION Firms invest in information technology (IT) such as computers and telecommunications technologies in order to improve their economic performance. IT can improve information sharing, decision-making, coordination, product quality, responsiveness, and distribution (Malone et al. 1987, 1989; Gurbaxani and Whang 1991; Brynjolfsson 1993; Brynjolfsson and Hitt 1996). However, it is difficult for companies to capture those benefits as profits. Even though IT may facilitate better coordination and output performance (or productivity), these benefits seldom result in improved financial performance (or profitability). One explanation is that IT investments by firms are paralleled by IT investments by buyers. The potential increases in profits are thus offset by the fact that buyers are able to use IT to reduce costs for searching for low-cost products or services and selecting alternative suppliers. IT investments, in other words, are necessary simply to keep up with market changes, but by themselves they are not enough to get ahead of these market changes. Thus, the lower price that buyers pay for products or services may reduce profitability, even though the reduced input costs may contribute to productivity increases (Hitt and Brynjolfsson 1996). Another explanation is that new IT applications will be copied by competitors in the industry, and thus all performance gains will disappear or be redistributed (Brynjolfsson 1993; Tam 1998). Another explanation is that investment in computers and other types of IT may not automatically improve financial performance; instead, it is one essential tool, but needs to be coupled with organizational factors such as business strategies to be truly effective 2

4 (Henderson and Venkatraman 1993; Brynjolfsson and Hitt 1995, 1998; Rai at al. 1997; Tam 1998). Previous research has found that, on average, IT increases productivity (Loveman 1994; Lichtenberg 1995; Brynjolfsson and Hitt 1996; Dewan and Min 1997; Rai et al. 1997). However, the overall benefits of IT investments vary enormously from firm to firm. Some firms are highly productive with high IT investments, but other firms are less productive with similar investments (Brynjolfsson and Hitt 1998). To explain the difference, Brynjolfsson and Hitt (1995) examined the sources of this variation using the statistical technique of a firm-effects model. They found that about half the benefits from IT investments derive from the unique characteristics of individual firms. Compared to the issue of IT productivity gains, the question of whether IT contributes to financial performance has not yet been clearly answered. Most previous research has shown IT s contribution to be minimal, negative, or mixed (Cron and Sobol 1983; Turner 1985; Bender 1986; Markus and Soh 1993; Rai et al. 1997; Tam 1998). This indicates that IT investments fail to improve business performance for some firms while others show marked improvement. Previous research has thus shown that IT does not automatically improve firm profitability. However, this research has not considered organizational factors such as strategic choices that can moderate the effect of IT on firm profitability. Strategic choices, such as vertical integration and diversification, are closely related to business performance (Bakos and Tracy 1986). A firm can maximize the value derived from its IT investments by aligning them with business strategies because IT improves scope economies and coordination (Henderson and Venkatraman 1993). This paper examines 3

5 empirically the contribution of IT to financial performance as measured by net profit, ROA, and ROE by focusing on the alignment of IT with business strategies such as vertical disintegration and diversification. 2. PREVIOUS LITERATURE AND RESEARCH QUESTIONS 2.1 Previous Empirical Studies on IT and Financial Performance Most previous research on the value of IT to firm profitability has focused on the direct relationship between the two. Cron and Sobol (1983) examined the impact of IT investment on financial performance for medical wholesale suppliers. They found that, on average, the impact of IT was not significant, and that there was either very strong or very weak effects on financial performance for firms with large IT investments. Strong financial performance was also found in larger firms. In his study of mutual savings banks, Turner (1985) found little evidence to suggest that there was a strong relationship between organizational performance and IT expenses or usage. Like Cron and Sobol, Bender (1986) also found a curvilinear relationship between IT investment and firm performance in the insurance industry. Firms with either very low or very high IT expenses performed poorly relative to those with IT expenses in between. They also found that firms with IT expenditures of 15% to 20% of total expenses were the best performers. Markus and Soh (1993) examined the relationship between firm profitability and a range of IT-related variables including IT expenditure, extent of computerization, and proportion of IT services outsourced while controlling for bank size and diversity of banking activities. They found that the larger banks performed worse in realizing returns 4

6 on their IT spending than the smaller banks did. But when they considered IT spending lagged and accumulated over four years, they found that more extensive computerization was associated with greater firm profitability in the larger firms than in the smaller firms. In their recent study, Rai et al. (1997) examined the performance effects of different types of IT investments by using multiple performance measures. They used three different measures of IT investments aggregate IT, client/server systems, and IT infrastructure and three different types of performance measures firm output (sales and value-added), financial (ROA and ROE), and intermediate (labor productivity and administrative productivity) performance. They found that IT investments are positively associated with firm output, but, as in earlier studies, that a clear association is lacking between IT investments and business performance. Only IT capital and client/server expenditure are positively associated with ROA. According to Rai et al, the reason for these results is that financial performance may be significantly affected by variation in the links between IT, business strategy, and competitive context across firms. They argue that incorporation of these contingencies might explain better the relationship between IT investments and financial performance. Tam (1998) examined the impact of IT investments on business performance ratios such as ROA, ROE, and ROS (return on sales) and the stock market return of the firms in four newly industrialized countries. He found that the associations between IT and business performance ratios are mixed and that IT does not correlate with return on stock value in all four countries. Based on these findings, he argues that performance gains can easily disappear because novel applications become routine operations and are copied by 5

7 competitors, and that the link between IT investment and performance ratios could be disrupted by institutional and societal factors, among others. In summary, there has been little evidence of strong improvements in financial performance derived from IT. Most previous research has considered the factor of firm size in examining the performance effects of IT. However, little previous research has considered other factors, such as strategic choice, that are associated with IT investments in the analysis of IT performance gains. 2.2 IT and Strategic Choice IT can be an effective means of both internal (intra-organizational) and external (interorganizational or market) coordination. According to previous research by Brynjolfsson et al. (1994), IT is correlated with a decrease in firm size. They used value-added as a supplementary measure of firm size and showed that reductions in firm size did not result from the substitution effect of IT but from more reliance on outsourcing due to lower market (or external) coordination costs. Firms that vertically integrate various stages of economic activity will create more value-added than firms that do not because vertical integration 1 brings several stages of economic activity under common ownership and management. The findings of Brynjolfsson et al. (1994) imply that IT leads to vertical disintegration by reducing external coordination costs (i.e., market transaction costs) relatively more than internal coordination costs. Several other researchers also argue that 1 Vertical integration is defined as the degree to which multiple stages of economic activities are combined under common ownership and coordinated within a single firm as opposed to a sequence of smaller firms. Vertically integrated firms coordinate most of their economic activities within their firm boundaries. On the contrary, vertically disintegrated firms do most of their economic activities outside their firm boundaries, relying more on market-mediated coordination. 6

8 IT leads to less vertical integration and facilitates value-adding partnerships (Johnston and Lawrence 1988; Malone et al. 1989; Bakos and Brynjolfsson 1993, 1997). Firms pursue vertical integration to avoid high market (external) coordination costs (Williamson 1975). Because IT also reduces these costs, combining IT investments with vertical integration may not be the best strategic choice. Smaller firms building up a value-adding partnership can exploit benefits such as economies of scale, scope and specialization by using IT. However, because IT also reduces internal coordination costs, firms may pursue vertical integration, which facilitates better control. Even though previous researchers provide empirical evidence for the association of IT with vertical disintegration (Brynjolfsson et al. 1994; Dewan et al. 1998), we have little evidence that the association of IT with vertical disintegration improves financial performance, e.g., net profit. Further research is required to determine whether the alignment of IT with vertical disintegration (or market coordination) improves business performance. Firms diversify their operations into multiple business lines for efficient utilization of surplus resources e.g., physical assets, managerial and technical expertise, and market information which are insufficiently utilized in the firms' current operations (Clarke 1985). However, potential economic benefits from sharing business resources are often not realized because of the costs of coordinating resources across multiple business lines or markets (Hill and Hoskisson 1987; Hill 1994; Montgomery 1994). Because IT is an effective means of coordinating business resources across multiple business units, it can improve the economic performance of highly diversified firms. In other words, the economic benefits of diversification can be leveraged by the capabilities of IT, which facilitates better coordination and communication. Recent research by Dewan et al. 7

9 (1998) partially supports this argument. Their findings imply that firms that are more diversified, especially in related lines of business, make greater investments in information technology. According to Hitt and Brynjolfsson (1996), although IT investments produce a significant increase in firm output, they do not improve business profitability. The reason is that IT investments by buyers reduce their costs for searching for low-cost products and services and selecting alternative suppliers. As a result, the lower price that buyers pay for products or services may reduce profitability, offsetting the productivity gains that IT brings to companies by reducing input utilization costs through automation and flexible manufacturing. In response to these market challenges, firms attempt to protect their profitability by relying on business strategies such as diversification and product differentiation (Bakos 1998). In other words, the economic benefits of IT can be leveraged by appropriate business strategies. Based on these considerations, this study argues that business performance can be improved by IT investments in conjunction with diversification. These theoretical considerations lead to the following research questions: 1. Does IT directly improve financial performance? 2. Does IT improve financial performance in conjunction with vertical disintegration? 3. Does IT improve financial performance in conjunction with diversification? To answer these questions, the next section uses an econometric model to perform an empirical analysis of the contribution of IT to financial performance as measured by net 8

10 profit, ROA, and ROE. The analysis includes the alignment (or interaction) of IT with vertical disintegration and diversification. 3. ECONOMETRIC APPROACH This study uses an economy-wide US firm-level data set. The data are divided into six sectors: durable and non-durable manufacturing, transportation and utilities, trade, finance, and other services. Several regressions are run on these data to identify the direction and magnitude of the relationship between IT and net profit, ROA, and ROE while controlling for vertical integration, diversification, the interaction effects of IT with vertical disintegration and diversification, other firm-specific characteristics such as R&D, and industry- and year-specific effects. 3.1 Data Sources and Variable Construction This research draws on two data sources: the data set of IT spending by large U.S. firms included in the Computerworld annual survey from 1988 to 1992 and the Compustat database. The data on IS spending was collected annually in a survey of IS executives from Fortune 500 and other selected firms. The data set includes data on the market value of central processors used by each firm (mainframes, minicomputers, and supercomputers), the total central IS budget, the percentage of the IS budget devoted to labor expenses, the number of PCs and terminals in use, and a variety of other financial and IT-related information. The total central IS budget figure reported in the survey is used as a measure 9

11 of IT spending. This figure includes labor expenses, materials, purchased services and software, and capital spending for the central IS department. Data items such as net profit, ROA, ROE, vertical integration, diversification, and R&D expense are obtained from the Compustat database for the same firms included in the IT data set. Financial performance is measured by net profit, and performance ratios such as ROA, and ROE. As mentioned in section 2.2, vertical integration refers to the degree to which multiple stages of economic activities are combined under common ownership. The traditional measure, the ratio of value-added to total sales, is used to specify the degree of vertical integration (Adelman 1955; Gort 1962). Value-added is derived from subtracting the costs of raw materials from the value of production (the value of the finished goods calculated as the sum of total sales and remaining inventory). This measure of vertical integration is sensitive to changes in the prices of input and output over time. This study uses the traditional measure of vertical integration, the ratio of value-added to total sales, because it is the most commonly used measure in the study of vertical integration. This study also conducts a firm-level analysis including industry and year (5 years) dummies that can reduce the limitations of the measure mentioned above (the effect of changes in input and output prices over time), thereby producing less biased measures of both vertical integration and disintegration. The inverse of vertical integration is used as a measure of vertical disintegration because less vertical integration implies more vertical disintegration. Diversification refers to the extent to which a firm operates in multiple business lines or markets. For the measure of diversification, we employ the sales-weighted entropy 10

12 measure introduced by Jacquemin and Berry (1979). The entropy measure for product diversification can be obtained from the weighted average of the sales shares of the different four-digit SIC code industries, where the weight for each industry is the logarithm of the inverse of its share. The measure refers to the extent of diversification arising from operations in several industries of the different four-digit SIC code industry groups. (See Dewan, Michael, and Min [1998] for details on the construction of the vertical integration and product diversification measures.) The alignment of IT with vertical disintegration and diversification is operationalized by employing the interaction term of IT with vertical integration and the interaction term of IT with diversification respectively. The sample includes 312 observations over five years. The sample statistics are shown in Table 1. Table 1. Sample Statistics: Mean and Standard Deviation (Five Years) Manufacturing Service Full Sample Variables Mean St. Dev. Mean St. Dev. Mean St. Dev. Profit 249 m 992 m 226 m 476 m 235 m 819 m ROA ROE IT Intensity R&D Intensity DIV VI SGA Expenses 1,067 m 1,529 m 1,391 m 2,162 m 1,131 m 1,712 m Sales 6,980 m 13,480 m 6,632 m 8,088 m 6,750 m 11,530 m Number of Obs Note: IT intensity is the ratio of the total central IS budget to selling, general, and administrative (SGA) expenses; R&D intensity is the ratio of R&D expenses to SGA expenses; DIV is diversification; VI is vertical integration; m is millions of dollars. 3.2 Methodologies and the Model 11

13 To analyze the relationship between IT and net profit, ROA, and ROE, an analysis of the combined data set for all five sample years is performed by using an ordinary least squares (OLS) regression and two-stage least squares (TSLS) regression. Since it is difficult to apply time-series models and complete lag structures to the residuals due to the short data period, TSLS regression is employed with the use of one-year lagged variables as instruments. This minimizes potential bias caused by the simultaneity problem (reverse causality). For example, instead of increased IT spending leading to an increase in net profit, increased net profit could be responsible for increases in IT spending. The TSLS regression technique is employed to correct this potential bias. To derive consistent estimates, once-lagged variables of IT, R&D, vertical integration, diversification, and interaction variables are employed as instrumental variables because they cannot be affected by the dependent variable in the following year The Model for Net Profit This model measures the relationship between IT and net profit while controlling for vertical integration, diversification, the interaction effects of IT with vertical disintegration and diversification, and other firm-specific characteristic such as R&D, and industry- and year-specific effects. LnProfit it = β 0 + β 1 ITint it + β 2 VI it + β 3 DIV it + β 4 IT*VDI it + β 5 IT*DIV it + β 6 RDint it + β 7 INDUSTRY it + β 8 YEAR it + ε where LnProfit it ITint it = Natural Logarithm of net profit of the ith firm in year t = IT intensity (IT spending/selling, general, and administrative expenses) 12

14 of the ith firm in year t VI it = Vertical integration of the ith firm in year t DIV it = Diversification of the ith firm in year t IT*VDI it = The interaction term of IT spending and vertical disintegration (the inverse of vertical integration) of the ith firm in year t IT*DIV it = The interaction term of IT spending and diversification of the ith firm in year t RDint it = R&D intensity (R&D expenses/selling, general, and administrative expenses) of the ith firm in year t INDUSTRY it = A dummy for industry YEAR it = A dummy for year ε = An error term with zero mean Since the size of each firm varies significantly in the data, the assumption of constant error variance is violated. This problem of heteroskedasticity is corrected by using intensity measures such as IT spending and R&D divided by selling, general, and administrative expenses. IT spending is used for the interaction variables; diversification is a ratio variable, and the variable for vertical disintegration is derived from vertical integration (also a ratio variable). The natural logarithm of net profit is used as a dependent variable. The ratio of net profit to selling, general, and administrative expenses is not used because of the problem of spurious correlation. 2 In order to control for firmspecific characteristics and industry- and year-specific effects, the R&D intensity variable along with dummy variables for each industry that is categorized by the SIC code and for each year are included. The model is estimated for the full sample and also for the manufacturing and service industry sectors separately in order to determine if the interaction effects of IT differ across sectors. 2 Spurious correlation refers to a situation in which correlation is found to be present between the ratios of variables even though the original variables are uncorrelated or random (Gujarati 1988). To avoid this problem, the ratio of net profit to total sales (return on sales or ROS) is not used. Another reason for not using this ratio is that diversification and vertical integration (or disintegration) are derived from total sales. 13

15 3.2.2 The Model for ROA and ROE This model measures the relationship between IT and performance ratios such as ROA and ROE while controlling for the interaction effects of IT with vertical disintegration and diversification, and industry- and year-specific effects. The model does not include variables such as R&D, vertical integration, and diversification because of the presence of high multicollinearity associated with other variables. The problem of heteroskedasticity is corrected by using the natural logarithms of IT spending and the interaction variables. This approach is widely used in the earlier research on IT and performance ratio (Hitt and Brynjolfsson 1996; Rai et al. 1997; Tam 1998). Performance Ratio it = β 0 + β 1 LnIT it + + β 2 Ln(IT*VDI) it + β 3 Ln(IT*DIV) it + β 4 INDUSTRY it + β 5 YEAR it + ε Performance Ratio it LnIT it Ln(IT*VDI) it Ln(IT*DIV) it INDUSTRY it YEAR it ε = ROA and ROE of the ith firm in year t = Natural Logarithm of IT of the ith firm in year t = Natural Logarithm of the interaction term of IT spending and vertical disintegration (the inverse of vertical integration) of the ith firm in year t = Natural Logarithm of the interaction term of IT spending and diversification of the ith firm in year t = A dummy for industry = A dummy for year = An error term with zero mean The models are designed to test the following hypotheses: Hypothesis H1: There is a positive relationship between financial performance and IT spending. Hypothesis H2: There is a positive relationship between financial performance and the interaction of IT spending with vertical disintegration. 14

16 Hypothesis H3: There is a positive relationship between financial performance and the interaction of IT spending with diversification. Based on theories discussed in section 2, it is expected that the coefficients of both interaction terms (β4 and β5 in the net profit model and β2 and β3 in the performance ratio model) will be positive and significant, and that the coefficient of IT (β1 in both net profit and performance ratio models) will be either positive or negative, but not significant. 4. EMPIRICAL RESULTS AND DISCUSSION The models are estimated with three different specifications: IT only, the interaction terms only, and all the variables. This allows us to isolate the effects of IT and those of the interaction terms. 4.1 The Impact of IT on Net Profit Results for the Full Sample Analysis of the relationship between net profit and IT shows that IT is positively associated with net profit, but not significant (column 2 in Table 2). The null hypothesis of zero effect of IT cannot be rejected for the full sample. This indicates that IT alone does not improve financial performance. Analysis of the relationship between net profit and the interaction terms shows that the interaction term of IT and diversification is strongly associated with an increase in net profit (p <.01) (column 3 in Table 2). The 15

17 estimate is consistent with our hypothesis that there is a positive relationship between financial performance and the interaction of IT with diversification. The null hypothesis of zero effect of the interaction term can be rejected for the full sample at the.01 (twotailed) confidence level. This result indicates that IT improves financial performance measured by net profit for firms with higher diversification, implying that firms can improve their performance by increasing IT spending in conjunction with higher diversification. The interaction term of IT spending and vertical disintegration also has a strong positive relationship with net profit (p <.01) (column 3 in Table 2). The estimate is also consistent with our hypothesis that there is a positive relationship between financial performance and the interaction of IT with vertical disintegration. This result indicates that IT improves financial performance measured by net profit for firms with a lower degree of vertical integration. It also implies that financial performance can be improved when firms increase IT spending associated with vertical disintegration. These results are consistent with previous research that shows increased IT investment to be associated with increased diversification and decreased vertical integration (Brynjolfsson et al. 1994; Dewan et al. 1998). The analysis controlling for other firm-specific variables provides similar results, except for the sign and the significance of the estimate of IT, which are negative and slightly significant (p <.1) (column 4 in Table 2). This further supports the conclusion that IT does not increase net profit by itself. The negative sign of the estimate indicates that IT, on average, has a negative impact on net profit. One possible explanation is that a number of IT projects have failed and thus the success of IT investments are tenuous, on average, when they are not coupled with organizational changes such as vertical 16

18 disintegration and diversification. Another explanation, as we saw earlier, is that IT usage in the marketplace reduces buyers' search costs, thereby intensifying price competition and reducing firm profitability. However, firms can protect their profitability with vertical disintegration and diversification. The explained variance (adjusted R 2 = 42.8%) of the controlled model has not been increased much from the model using the interaction terms only (adjusted R 2 = 39.1%). 3 This indicates that the variance in financial performance measured by net profit is explained more by the interaction of IT with vertical disintegration and diversification. Table 2. OLS Regression Results (Full Sample) IT/SGA IT*VDI IT*DIV R&D/SGA VI DIV Dummy Adjusted R 2 N.013 (.149) % 312 Key: ***(p<.01), **(p<.05), *(p<.1) 1 The values in parentheses are standard errors..359*** (.000).203*** (.000) 39.1% * (.137).393*** (.000).207*** (.000).172*** (.295).131** (.281) (.142) 42.8% Results for the Manufacturing and Service Industries A separate analysis for the manufacturing industry (Table 3) shows that the interaction of IT spending with vertical disintegration has a positive and significant relationship with net profit (p <.01). The interaction of IT and diversification has also a positive and 3 The goodness of fit of the regressions is relatively low. The reason is that there are other variables, mostly macro-level variables such as interest rates, competition, and antitrust regulation, which cannot be controlled by the firm-level analysis. The explained variance of this regression is, however, greater than those of previous studies on IT performance gains, which ranged from 0.01% to 30.0% (Cron and Sobol 1983; Turner 1985; Bender 1986; Markus and Soh 1993; Rai et al. 1997; Tam 1998). 17

19 significant relationship with net profit (p <.01). However, IT does not have a strong positive relationship with increased net profit. The results are consistent with our hypothesis that IT alone does not improve financial performance, but that it does improve financial performance for firms with higher levels of vertical disintegration and diversification. Table 3. OLS Regression Results (Manufacturing) IT/SGA IT*VDI IT*DIV R&D/SGA VI DIV Dummy Adjusted R 2 N (.149) % 282 Key: ***(p<.01), **(p<.05), *(p<.1) 1 The values in parentheses are standard errors..349*** (.000).215*** (.000) 43.1% ** (.138).339*** (.000).234*** (.000).223*** (.302).190*** (.364) (.158) 48.0% 282 A separate analysis of the service industry does not show the same result for the interaction of IT with vertical disintegration and diversification. The small sample number (30 observations) might affect the magnitude of the estimate of the interaction term. One possible explanation is that service firms have relatively higher economies of scale and scope around IT than manufacturing firms; thus, the higher level of IT spending with higher vertical disintegration and diversification does not necessarily lead to higher financial performance in the service industry. This means that aligning IT with vertical disintegration and diversification might be a more important factor for improved financial 18

20 performance (i.e., increased net profit) in the manufacturing industry than in the service industry. Table 4. OLS Regression Results (Service) IT/SGA IT*VDI IT*DIV R&D/SGA VI DIV Dummy Adjusted R 2 N (2.722) % 30 Key: ***(p<.01), **(p<.05), *(p<.1) 1 The values in parentheses are standard errors (.002).131 (.007) 43.1% ** (4.923) (.009).868 (.022) * (39.5).908** (1.501) (2.781) 74.5% TSLS Regression Results The results of the TSLS regression analysis are similar to the results of the OLS regression analysis given above. 4 However, the number of observations in the TSLS analysis is lower than the OLS figure because each observation requires data for both the current period and the previous period. This eliminates observations for all of 1988 and some observations in other years. The standard errors for the coefficient estimates of the independent variables are also substantially larger since instrumental variables are used. The lower number of observations and the associated larger standard errors reduce the significance levels of the coefficient estimates of the variables. The results of the TSLS regression analysis are shown in Table 5. 4 This study does not apply the time-series model because the number of observations is too small for this kind of analysis. Instead, the model is estimated separately for one-year lag and a two-year lag variable. The results of the analyses are similar to the OLS and TSLS regression results. 19

21 IT/SGA IT*VDI IT*DIV R&D/SGA VI DIV Dummy Adjusted R 2 N Table 5. TSLS Regression Results 1 Variable Manufacturing Service Full Sample.053 (2.340) (80.6) (1.835).394*** (.000) (.025).431*** (.000).217** (.000).410 (.038).195** (.000).387 (3.144) (84.6).197 (1.578).307* (.840) (12.6).187* (.546) (.254) (10.6) (.211) 45.6% 96.3% 41.2% Key: ***(p<.01), **(p<.05), *(p<.1) 1 Instrumental variables: once-lagged independent variables (IT/SGA, R&D/SGA, vertical integration, diversification, and interaction variables). 2 The values in parentheses are standard errors. 4.2 The Impact of IT on ROA and ROE Analysis of the relationship between performance ratios such as ROA and ROE and IT variable only shows that IT is not correlated with ROA and ROE (Tables 6 and 7). The analysis also shows that the interaction terms of IT and diversification and vertical disintegration are not correlated with ROA and ROE. Only the interaction term of IT and diversification is significantly associated with an increase in ROA for the full sample (column 7 in Table 6). Because the estimates of the OLS regressions are not significant, TSLS regression is not performed. For the same reason, the model specification including all three variables (IT and the two interaction variables) is not estimated. 20

22 Table 6. OLS Regression Results for ROA Manufacturing Service Full Sample LnIT Ln(IT*VDI) Ln(IT*DIV) Dummy Adjusted R 2 N.001 (.149) (.349) (.235) 26.5% 26.6% 307 Key: ***(p<.01), **(p<.05), *(p<.1) 1 The values in parentheses are standard errors..038 (1.014).070 (.816) (.377) 59.5% 60.0% (.264) -.115* (.293).155** (.165) 7.9% 9.1% 371 Table 7. OLS Regression Results for ROE Manufacturing Service Full Sample LnIT Ln(IT*VDI) Ln(IT*DIV) Dummy Adjusted R 2 N (1.751) (2.598).051 (1.746) 5.3% 5.1% 307 Key: ***(p<.01), **(p<.05), *(p<.1) 1 The values in parentheses are standard errors..232* (2.330).213 (1.913) (.883) 30.9% 29.0% (1.400) (1.567) (.884) 1.5% 1.7% Discussion The results clearly show that IT does not automatically improve financial performance as measured by net profit. The findings imply that IT is necessary for competitive parity, but not sufficient for competitive advantage (i.e., increased profit). The estimates of the interaction of IT and vertical disintegration indicate that increased IT spending improves financial performance of firms with decreased vertical integration. Because IT can be used for efficient market coordination in vertically disintegrated firms, this capability of IT might contribute to increased net profit for these firms. The estimates of the interaction of IT and diversification also show that increased IT spending improves financial 21

23 performance of firms with higher diversification. This implies that firms can increase their net profit with IT investments by aligning them with diversification. The reason may be that IT leverages the economic benefits of diversification by facilitating better coordination of business resources and economies of scope across multiple business units. The results also show that the alignment of IT with both vertical disintegration and diversification are more effective in improving financial performance as measured by net profit in the manufacturing industry than in the service industry. Results from the analysis of ROA and ROE show, however, that even when business strategies such as vertical disintegration and diversification are associated with increased IT spending, they have little effect on improvement in performance ratios such as ROA and ROE. Compared to the presence of a relationship between net profit and the interaction of IT with vertical disintegration and diversification, the insignificant results of the same interaction effects on ROA and ROE are interesting. These results imply that vertical disintegration and diversification associated with IT spending may have little impact on how effectively a firm uses its capital investments and financial leverage. Although there has been a dramatic increase in IT investment in recent years, IT investment is still very small compared to total assets or capital investments (Hitt and Brynjolfsson 1996). As a result, the interaction effects of IT with vertical disintegration and diversification may not be detectable with ROA and ROE measures. Compared to net profit, performance ratios such as ROA and ROE may not be appropriate for estimating the performance effects of IT because of the low ratio of IT to total assets or capital investments. 22

24 5. CONCLUSION This paper performed an empirical analysis of the contribution of IT to financial performance as measured by net profit, ROA, and ROE by focusing on the interaction (or alignment) effects of IT with vertical disintegration and diversification. Data were analyzed for the full sample and for both the manufacturing and service industry sectors for the five years from 1988 to The main contribution of this research is that it provides empirical evidence for the importance of aligning IT with business strategies such as vertical disintegration and diversification. Previous research suggests that the relationship between IT investment and financial performance (or profitability) is not a direct one and is likely to be moderated by business strategy, management practice, and competitive context across firms (Henderson and Venkatraman 1993; Brynjolfsson and Hitt 1995, 1998; Rai et al. 1997; Tam 1998). Rai et al. (1997) argue that incorporation of these factors in the analysis of performance effects of IT may provide a better explanation of the relationship between IT and financial performance. This is one of the first empirical studies that consider business strategies in conjunction with IT investment in the analysis of financial performance. It does this by examining the interaction effects of IT with vertical disintegration and diversification. This study sheds light on how firms can improve financial performance with IT investment, a question that has not been clearly answered by previous research. Another contribution is that, compared to previous studies, this research controlled for firm-specific characteristic such as R&D in addition to the strategic variables such as vertical integration and diversification. 23

25 The results show that IT does not directly improve financial performance, but that IT can improve business performance when used in conjunction with vertical disintegration and diversification. These findings imply that IT alone does not bring success. Although it is an essential component, it is not sufficient in itself and should be coupled with organizational changes. Firms that do not make appropriate organizational changes and develop appropriate business strategies may fail to take full advantage of IT capabilities. The results are also consistent with recent findings showing that increased IT investment is associated with increased diversification and decreased vertical integration. Furthermore, the findings provide partial support for the argument that by reducing buyers' search costs, IT lowers the price paid for products or services and thereby reduces firm profitability. Our findings also support the argument that by improving scope economies and coordination IT can shape appropriate business strategies, and at the same time the economic benefits of IT can be leveraged by such business strategies. 24

26 REFERENCES Adelman, M.A. "Concept and Statistical Measurement of Vertical Integration," in Business Concentration and Price Policy, edited by G.J. Stigler, Princeton: Princeton University Press, , Bakos, Y. "The Emerging Role of Electronic Marketplaces on the Internet," Communications of the ACM, 41, 8, 35-42, August Bakos, Y. and Brynjolfsson, E. "From Vendors to Partners: The Role of Information Technology and Incomplete Contracts in Buyer-Supplier Relationships," Journal of Organizational Computing, 3, 3, , Bakos, Y. and Brynjolfsson, E. "Organizational Partnerships and the Virtual Corporation," in Kemerer C.L. Information Technology and Industrial Competitiveness: How Information Technology Shapes Competition, Kluwer Academic Publishers, 1997 Bakos, Y and Tracy, M.E. "Information Technology and Corporate Strategy: A Research Perspective," MIS Quarterly, , June Bender, D.H. Financial Impact of Information Processing, Journal of Management Information Systems, 3, 2, 22-32, Brynjolfsson, E. The Productivity Paradox of Information Technology, Communications of the ACM, 36, 12, December Brynjolfsson, E. and Hitt, L. "Information Technology as a Factor of Production: The Role of Differences Among Firms," Economics of Innovation and New Technology, 3, 4, , May Brynjolfsson, E. and Hitt, L. Paradox Lost? Firm Level Evidence on the Returns to Information Systems Spending, Management Science, , April Brynjolfsson, E. and Hitt, L. "Beyond the Productivity Paradox," Communications of the ACM, 41, 8, 49-55, August Brynjolfsson, E., Malone, T.W., Gurbaxani, V., and Kambil, A. Does Information Technology Lead to Smaller Firms? Management Science, 40, 12, , December Clarke, R. Industrial Economics, Basil Blackwell Ltd., Oxford, UK, Cron, W.L. and Sobol, M.G. The Relationship Between Computerization and Performance: A Strategy for Maximizing the Economic Benefits of Computerization, Information and Management, 6, ,

27 Dewan, S., S.C. Michael, and C. Min, "Firm Characteristics and Investments in Information Technology: Scale and Scope Effects," Information Systems Research, 9, 3, , September Dewan, S. and Min, C. "Substitution of Information Technology for Other Factors of Production: A Firm-Level Analysis," Management Science, 43, 12, , December Gort, M. Diversification and Integration in American Industry, Princeton: Princeton University Press, Gujarati, D.N. Basic Econometrics, McGraw-Hill, Inc., New York, NY, Gurbaxani, V. and Whang, S. The Impact of Information Systems on Organizations and Markets, Communications of the ACM, 34, 1, 60-73, January Henderson, J.C. and Venkatraman, N. "Strategic Alignment: Leveraging Information technology for Transforming Organizations," IBM Systems Journal, 32, 1, 4-16, Hill, C.W.L. "Diversification and Economic Performance," in Fundamental Issues in Strategy, R.P. Rumelt, D.E. Schendel, and D.J. Teece, Eds. HBS Press, Boston, MA, , Hill, C.W.L. and Hoskisson, R.E. "Strategy and Structure in the Multiproduct Firm," Academic Management Review, 12, 2, , Hitt, L. and Brynjolfsson, E. Productivity, Business Profitability, and Consumer Surplus: Three Different Measures of Information Technology Value, MIS Quarterly, , June Jacquemin, A.P. and Berry, C.P. "Entropy Measure of Diversification and Corporate Growth," Journal of Industrial Economics, 27, 4, , Johnston, R. and Lawrence, P.R. "Beyond Vertical Integration: The Rise of the Value- Adding Partnership," Harvard Business Review, , July-August Lichtenberg, F.R., "The Output Contributions of Computer Equipment and Personnel: A Firm-Level Analysis," Economics of Innovation and New Technology, 3, 4, , May 1995 Loveman, G. W. Assessing the Productivity Impact of Information Technologies, in Allen, T.J. and Scott-Morton, M. Information Technology and the Corporation of the 1990s, Oxford University Press, New York, New York,

28 Malone, T.W., Yates, J. and Benjamin, R.I. Electronic Markets and Electronic Hierarchies, Communications of the ACM, , June Malone, T.W., Yates, J., and Benjamin, R.I. "The Logic of Electronic Markets," Harvard Business Review, , May-June Markus, M.L. and Soh, C. Banking on Information Technology: Converting IT Spending into Firm Performance, in Banker, R.D., Kauffman, R.J., and Mahmood, M.A. Strategic Information Technology Management: Perspectives on Organizational Growth and Competitive Advantage, Idea Group Publishing, Harrisburg, PA, Montgomery, C.A. "Corporate Diversification," Journal of Economic Perspectives, 8, 3, , Rai, A., Patnayakuni, R., and Patnayakuni, N. Technology Investment and Business Performance, Communications of the ACM, 40, 7, 89-97, July Tam, K.Y. The Impact of Information Technology Investments on Firm Performance and Evaluation: Evidence from Newly Industrialized Economics, Information Systems Research, 9, 1, 85-98, March Turner, J. Organizational Performance, Size and the Use of Data Processing Resources, Working Paper, Center for Research in Information Systems, New York University, Williamson, O.E., Markets and Hierarchies: Analysis and Antitrust Implications, New York: Free Press,