Testing productivity paradox: the Slovenian case.ABSTRACT This study examines the relationship of information technology (IT) investments and productivity of Slovenian economy. Based on panel data for all 15 Slovenian industries (SIC: A to O) in period from 1996 to 2000 we tested for presence (or absence) of the productivity paradox The productivity paradox (also known as the Solow computer paradox) is the observation made in Computer Supported Cooperative Work and other business process analysis that, as new information technology is introduced, worker productivity may go down, not up. in the field of IT investments in Slovenia. Specifically, this project tests multiple regressions Multiple regression The estimated relationship between a dependent variable and more than one explanatory variable. with 3 information and telecommunication technology investment measures and investments in research as independent variables (hardware, telecommunications, software and research) and value added Value Added The enhancement a company gives its product or service before offering the product to customers. Notes: This can either increase the products price or value. as measure of productivity (dependent variable). The methodology consists of EDA--stem-and-leaf plots, correlation coefficients Correlation Coefficient A measure that determines the degree to which two variable's movements are associated. The correlation coefficient is calculated as: , OLS OLS Ordinary Least Squares OLS Online Library System OLS Ottawa Linux Symposium OLS Operation Lifeline Sudan OLS Operational Linescan System OLS Online Service OLS Organizational Leadership and Supervision OLS On Line Support OLS Online System and panel regressions (fixed effects model and random effects model In statistics, a random effect(s) model, also called a variance components model is a kind of hierarchical linear model. It assumes that the data describe a hierarchy of different populations whose differences are constrained by the hierarchy. ). The findings show mixed results-significantly positive association between hardware and value added, slight but positive association between investment in telecommunication and value added and non-significant results for impact on software and research investments on value added. However, weight of first two independent variables is well above 75%. Therefore, the productivity paradox is rejected. 1. INTRODUCTION Nobel Prize-winning economist Robert Solow Robert Merton "Bob" Solow (born August 23, 1924) is an American economist particularly known for his work on the theory of economic growth. He was awarded the John Bates Clark Medal (in 1961) and the 1987 Nobel Prize in Economics. (1987) has said that we see computers everywhere except in the productivity statistics. This phenomenon was called 'productivity paradox' and asserts that IT investments do not result in productivity gain (Navarette and Pick, 2002). Attention was first drawn to the productivity paradox by Morgan Stanley's chief economist The Chief Economist is a single position job class having primary responsibility for the development, coordination, and production of economic and financial analysis. It is distinguished from the other economist positions by the broader scope of responsibility encompassing the Steven Roach roach: see cockroach. roach Common European sport fish (Rutilus rutilus) of the carp family (Cyprinidae), found in lakes and slow rivers. A high-backed, yellowish green fish with red eyes and reddish fins, the roach is 6–16 in. who published study "America's Technology Dilemma: A Profile of the Information Economy" in Morgan Stanley's April 22, 1987 economics newsletter series. His conclusion was that the tremendous increase in computerization com·put·er·ize tr.v. com·put·er·ized, com·put·er·iz·ing, com·put·er·iz·es 1. To furnish with a computer or computer system. 2. To enter, process, or store (information) in a computer or system of computers. has had little effect on economic performance, particularly for those sectors of the economy with large numbers of "information workers". First of all let us question ourselves why productivity is so important and do information technologies (in our case hardware, telecommunications and software accompanied with investment in research--knowledge) contribute to productivity growth. Productivity growth determines our living standards living standards npl → nivel msg de vida living standards living npl → niveau m de vie living standards living npl and wealth of the nations. Productivity growth is simple concept. It is the amount of output produced per unit of input. While it is easy to define, it is notoriously difficult to measure, especially in the modern economy (Brynolfsson and Hitt, 1998). Those two authors believe that one of the main reasons for productivity paradox is mismeasurement Mis`meas´ure`ment n. 1. Wrong measurement. . While on the one hand all the cost can be measured and included in statistics, their benefits are difficult to measure due to its intangibility and due to fact that they normally occur after some time lag. Stratopoulos and Dehning (2000) report on another possible reason for productivity paradox. They conducted quasi-experiment comparing successful users of IT and less successful users and found out that how you manage your IT assets is more important than how much you spend on IT. According to according to prep. 1. As stated or indicated by; on the authority of: according to historians. 2. In keeping with: according to instructions. 3. their research, mismanagement mis·man·age tr.v. mis·man·aged, mis·man·ag·ing, mis·man·ag·es To manage badly or carelessly. mis·man  age·ment n. can be another reason for productivity
paradox. Otherwise, this relation between investment in IT and
productivity has been the center of numerous research projects,
editorials, articles and books. This paper has the objective to analyze
the effect of IT investments on productivity in Slovenia (analyzing data
for industries) for period from 1996 till 2000.Our article is consisted out of three main parts. In first we introduce 4 different streams of research regarding productivity paradox in the field of IT investments and present their main findings. Beside that we form a basis for variables implemented in the model. In second part we proceed with presentation of methodological frame of research by presenting research model, hypotheses to be tested and sources of data used for empirical investigation, which forms the third part of our paper. We start by exploring individual variables using Exploratory Data Analysis Exploratory Data Analysis - (EDA) [J.W.Tukey, "Exploratory Data Analysis", 1977, Addisson Wesley]. (Stem-and-Leaf plots), correlation coefficients and descriptive statistics descriptive statistics see statistics. . In next step we conduct confirmatory analysis--multiple regression based on panel data (Ordinary Least Squares, Fixed Effects Model and Random Effects Model accompanied with Hausman test The Hausman test is a test in econometrics named after Jerry Hausman. The test evaluates the significance of an estimators versus an alternative estimator. If the linear model ). We conclude with main realizations regarding our research question. 2. THEORY Previous research in the field of productivity paradox testing can be systemized into four broader categories: firm-based, industry-based, countrywide coun·try·wide adv. & adj. Throughout a whole country; nationwide: launched a fundraising campaign countrywide; a countrywide search. Adj. 1. and international research. Results supporting or refusing productivity paradox vary strongly. 2.1. International level of research Research at international level can be mainly contributed to Kraemer and Dedrick (1996, 2001) and Dewan de·wan n. Any of various government officials in India, especially a regional prime minister. [Hindi d and Kraemer (1998, 2000). All of their work in that field rejects productivity paradox or at least does not support it. Kraemer and Dedrick (1996) found out (based on sample of 12 Asian-Pacific states in period from 1984-1990) that IT investment is positively associated to GDP GDP (guanosine diphosphate): see guanine. and productivity growth. However, in 2001 they realized (studying 43 states) that level of IT investment (as percentage of GDP) is not statistically significant correlated to productivity growth. IT investment growth on the other hand has positive impact on productivity growth. Dewan and Kraemer (1998, 2000) examined 36 states and concluded that IT capital is positively associated to productivity of workforce in developed countries, while in developing countries it has no significant impact. Plice (2001) examined 6 industrial sectors in 38 developed countries and realized that IT capital demonstrated 5-8 times higher return on investment (ROI (Return On Investment) The monetary benefits derived from having spent money on developing or revising a system. In the IT world, there are more ways to compute ROI than Carter has liver pills (and for those of you who never heard of that expression, it means a lot). ) as non-IT capital. 2.2. Countrywide level of research At the countrywide level of research Roach (1987) was the pioneer and 'founder' of productivity paradox, He attempted to explain why the measured productivity growth rate in the U.S. economy has slowed substantially since 1973. Roach observed that the amount of computing computing - computer power per white-collar worker white-collar worker n → oficinista m/f white-collar worker n → employé(e) de bureau white-collar worker white n in the service industry was growing dramatically over 1970s and 1980s, yet the measured productivity of this part of US economy was flat. As already mentioned in introduction, his conclusion was that enormous increase in computerization had little impact on economic performance, particularly for those sectors of the economy with large numbers of information workers. Strassman (1997a, 1997b, 1999) and Schrage (1997) show disappointing impact of IT investment on the U.S. economy. In contrary, Dewan and Kraemer (1998) report other developed countries (e.g. Germany) to have positive returns on IT investments. Dos Santos Santos (sän`t  s), city (1996 pop. 412,288), São Paulo state, SE Brazil, on the island of São Vicente in the Atlantic just off the mainland. et al. (1993) and Weill (1992) show mixed
results for the United States United States, officially United States of America, republic (2005 est. pop. 295,734,000), 3,539,227 sq mi (9,166,598 sq km), North America. The United States is the world's third largest country in population and the fourth largest country in area. , while Melville (2001) based on panel of
31 industries in period from 1965-1991 found positive impacts of IT
investments for USA as whole.2.3. Industry based research Industry based research display mixed results. Navarrette and Pick (2002) in their study on Mexican banking industry for period 1982 to 1992 tested correlation between the industry's information technology spending and three performance measures: profits, return on assets Return on assets (ROA) Indicator of profitability. Determined by dividing net income for the past 12 months by total average assets. Result is shown as a percentage. ROA can be decomposed into return on sales (net income/sales) multiplied by asset utilization (sales/assets). and return on equity. Using methodological frame of longitudinal correlations of the industry consolidated data over eleven-year period and graphical analysis of time series they came out with a positive association between IT expenditure and industry's net profits and return on assets. Hence, the productivity paradox was rejected. Alpar and Kim (1990) examined selected U.S. firms in the banking industry and through use of production function showed a significant cost reduction from IT. However, these results are contrary to major part of other industry level studies that support the productivity paradox or show mixed results. Among them: Strassman (1997b), Turner (1985), Harris and Katz (1991), Loveman (1988), Barua et al. (1995). Strassman (1997b) used data for U.S. corporations, including a sample of the 16 large U.S. banks with revenues over $5 billion, to examine the relationships between IT expenditures and productivity as well as return on equity. Even though the banks in the sample expanded substantially their IT investments in the period 1989-1996 and invested more in IT than any other economic sector, there were not corresponding gains in productivity or return on equity of banks. Turner (1985) used a sample of 58 mutual savings banks Mutual savings bank A state-chartered savings bank which is owned by its depositors and managed by a fiduciary board of trustees. in the U.S., examined the relationship between IT expenses or IT use and banks' performance and found no relationship. Harris and Katz (1991) tested the relationship of IT investment and performance ratios for selected companies in the U.S. life insurance industry. For a data set from year 1983-1986 they found that firm performance is weakly weak·ly adj. weak·li·er, weak·li·est Delicate in constitution; frail or sickly. adv. 1. With little physical strength or force. 2. With little strength of character. linked to the level of IT investment intensity. Loveman (1988) and Barua et al. (1995) examined non-financial sectors. For the U.S. manufacturing sector, Loveman (1988) concluded that the contribution of IT capital investment to output is almost zero. Barua et al. (1995) tested the impact of IT in two non-financial industrial sectors and found that IT capital investment had an impact on five intermediate variables of capacity utilization Capacity Utilization measures the rate at which a firm makes use of their capital productive capacities, such as factories and machinery. Capacity Utilization generally rises when the economy is healthy and falls when demand softens. , inventory turnover, quality, relative product prices and new products. 2.4. Firm level of research Another pattern one sees in studies rejecting the paradox is prevalence of firm unit of analysis (Navarette and Pick, 2002). Banker and Kauffman (1988), Haynes (1990) and Brady and Targett (1995) conducted some studies that focused at the firm level. Firm based research and international research mainly rejects the paradox, as we saw previously. On the other hand projects with mixed results and supporting the paradox tend to be at the national and industry levels (with some exceptions). Before moving further to methodology and empirical investigation let us examine categorization of IT (and consequentially con·se·quen·tial adj. 1. Following as an effect, result, or conclusion; consequent. 2. Having important consequences; significant: investments in IT). Which independent variables should we include in our model? Vast majority of authors (Turban et al., 2001a, 2001b; Beynon-Davies 2002; Fitzgerald and Dennis, 1999) agree that the most applicable classification of IT into its' segments is following: hardware, software and telecommunications. Andersen and Segars (2001) used similar classification when examining relationship among IT and firm performance. It is obvious, however, that for those investments to become active, knowledge is essential. Learning organization (Senge, 1995) and organizational learning Organizational learning is an area of knowledge within organizational theory that studies models and theories about the way an organization learns and adapts. In Organizational development (OD), learning is a characteristic of an adaptive organization, i.e. (Argyris and Schon 1978) are not new concepts. Nevertheless, they are becoming more and more discussed and implemented in today's business Today's Business is a show on CNBC that aired in the early morning, 5 to 7AM ET timeslot, hosted by Liz Claman and Bob Sellers, and it was replaced by Wake Up Call on Feb 4, 2002. society. A pattern that one might notice is that those two concepts are becoming more and more related to information technologies (Andreu and Ciborra, 1996; Robey et al., 2000). That is why we decided to add investments in research as another independent variable. Another question that arises looking at previous research is what is the most appropriate and representative measure of productivity to use in our research. Some authors (Andersen and Segars 2001; Navarette and Pick 2002) used measures associated with profits (ROA ROA See: Return on assets ROA See: Right of accumulation ROA See return on assets (ROA). , ROI, ROE, net profits etc.). What needs to be noted is that profit is a category that is can be very much influenced by the ability of firm's accountants to be 'creative' for reasons of taxation. Ideally we should use financial and non-financial measures according to Freeman's Stakeholder stakeholder n. a person having in his/her possession (holding) money or property in which he/she has no interest, right or title, awaiting the outcome of a dispute between two or more claimants to the money or property. Approach (1984, 1994) and Balanced Scorecard Balanced Scorecard A performance metric used in strategic management to identify and improve various internal functions and their resulting external outcomes. The balanced scorecard attempts to measure and provide feedback to organizations in order to assist in implementing (Kaplan and Norton, 1992, 1993, 1996, 1996a) to evaluate success from all shareholders' point of view. However, due to the fact that those measures are just starting to be used in Slovenian firms, the most appropriate (and accessible in National Accounts) would be value added. Next we will deal with brief description of methodology utilized in our research by presenting a research model developed and data utilized in our research. 3. METHODOLOGY OF RESEARCH 3.1. Research model Methodological frame of our research has been consisted out of method of EDA (1) (Electronic Design Automation) Using the computer to design, lay out, verify and simulate the performance of electronic circuits on a chip or printed circuit board. (Exploratory Data Analysis) developed by Hartwig and Dearing (1979) accompanied with confirmatory analysis--multiple regressions. Using Stem-and-Leaf Plots and descriptive statistics we determined location, spread and shape of each individual variable in the model. Afterwards af·ter·ward also af·ter·wards adv. At a later time; subsequently. afterwards or afterward Adverb later [Old English æfterweard] Adv. 1. we looked into relationships among those individual variables using correlation coefficients (relationships among independent variables--to test fulfillment of assumptions of OLS such as absence of multicolinearity) and multiple regression analysis to test relationships among independent variables and dependent variable. Since our data are panel data we first conducted Ordinary Least Squares regression (OLS) followed by Fixed Effects Model (Least Squares with Group Dummy Variables This article is not about "dummy variables" as that term is usually understood in mathematics. See free variables and bound variables. In regression analysis, a dummy variable ) and Random Effects Model. Using Hausman test we determined that the specification of model is appropriate and Random Effects Model provides better estimations of regression coefficients Regression coefficient Term yielded by regression analysis that indicates the sensitivity of the dependent variable to a particular independent variable. See: Parameter. regression coefficient than Fixed Effects Model. Research model is presented in Figure 1 and research questions are developed out of this scheme. Reasons to include above given variables were described in previous section. The main question that comes to our mind is--can productivity paradox be observed in Slovenian economy? Out of this main research question we develop 4 hypotheses (Table 1) that will be a subject of our multiple regression analysis (based on panel data) after conducting exploratory data analysis (EDA)--Stem-and-Leaf plots and correlation coefficients. Stem-and-Leaf plots developed by Tukey (1977) are interesting tool to display individual data since they have ability to retain all of the observed values in rank order and at the same time convey well the shape of the distribution. [FIGURE 1 OMITTED] 3.2. Research data Data for our research have been gathered from 3 different sources: Statistical Office of Republic of Slovenia (SURS SURS State Universities Retirement System (State of Illinois) SURS Surveillance Utilization Review System SURS Stevens Untitled Rock Show SURS Surveillance Squadron SURS Standard Umbilical Retractor System SURS Surveillance Review Subsystem ), Agency of the Republic of Slovenia for Public Legal Records and Related Services (AJPES) and Chamber of Commerce and Industry of Slovenia (GZS GZS Gospodarska Zbornica Slovenije GZS Gesellschaft Fuer Zahlungssysteme (German credit card processor) ). Our panel involved time series from 1996-2000 and used data at level of industry (industries from A to O, according to Slovenian version of Standard Industrial Classification--SKD). This allowed us to perform analysis on 75 observations of ratio data (without any missing values In statistics, missing values are a common occurrence. Several statistical methods have been developed to deal with this problem. Missing values mean that no data value is stored for the variable in the current observation. ). We performed countrywide research that gives additional value to our research vs. partial industry research or sample based research. Prior to analyzing, values for investment in hardware, software, telecommunications and research, they all have been recalculated to fixed prices with base year 1995 to eliminate inflation effect. We used consumer price indices for that transformation. 4. ANALYSIS To understand data we decided to use EDA approach and afterwards to test hypotheses based on multiple regression confirmatory analysis. EDA approach allowed us to look deeper into our data and understand them better at individual level as well as relationship among four independent and one dependent variable. 4.1. Exploratory data analysis We used SPSS A statistical package from SPSS, Inc., Chicago (www.spss.com) that runs on PCs, most mainframes and minis and is used extensively in marketing research. It provides over 50 statistical processes, including regression analysis, correlation and analysis of variance. to explore data to decide whether to use linear or non-linear function. Results of exploratory analysis are best shown in Tables 3 and 4 and Figures 1 and 2. We present some descriptive statistics to present shape, location and spread of single variables. All values are in 1000 Slovenian tolars The tolar was the currency of Slovenia from 1991 until December 31, 2006. It was subdivided into 100 stotinov. The ISO 4217 currency code for the Slovenian tolar was SIT. The name tolar comes from Thaler, and is cognate with dollar. (SIT) in fixed prices with base year 1995. On average Slovenian industries invested in hardware (HW) almost 2 billions SIT per industry per year. Maximum was achieved in 1999 when industry D (manufacturing industry) invested above 7 billions of SIT into hardware equipment. Mode and median are well below average what indicates positive skewness Skewness A statistical term used to describe a situation's asymmetry in relation to a normal distribution. Notes: A positive skew describes a distribution favoring the right tail, whereas a negative skew describes a distribution favoring the left tail. of distribution of HW variable. Average investment in telecommunications (TCM (1) (Trellis-Coded Modulation/Viterbi Decoding) A technique that adds forward error correction to a modulation scheme by adding an additional bit to each baud. TCM is used with QAM modulation, for example. ) has been roughly 1.7 billion SIT per industry per year. All together Slovenian companies This is a list of Slovenian corporations:
one of the voices of cattle. Usually refers to the arrogant call of the bull used to announce territorial rights. Abnormalities of the voice include hoarseness as in rabies, or continuous repetition as in nervous acetonemia. See also low, moo. average value. Investments in software (SW) have been much lower that those in hardware and telecommunications. In our opinion this can be contributed also to effect of pirate software in past. Nowadays, companies give much more attention to obtaining legal licenses for software applications. This and the fact that more and more companies are starting to use Enterprise Resource Planning See ERP. (application, business) Enterprise Resource Planning - (ERP) Any software system designed to support and automate the business processes of medium and large businesses. (ERP (Enterprise Resource Planning) An integrated information system that serves all departments within an enterprise. Evolving out of the manufacturing industry, ERP implies the use of packaged software rather than proprietary software written by or for one customer. ) systems can result also in increased software investments in future. On the other hand, use of Web services (1) Loosely, any online service delivered over the Web. Such usage appears in articles from non-technical sources, but not in IT-oriented publications, because definition #2 below describes the correct use of the term. might very well lead to decreased software investments while providing higher value added. Average software investment accounts for 340 million SIT per industry per year. Slovenian industries invested on average almost 1 billion SIT in research projects (RES) per year in observed period. In all 4 cases (investments in hardware, telecommunications, software and research) positive skewness can be spotted based on Stem-and-Leaf plots as well as from high discrepancies among mean, median and mode. Before we move further, let us look at the portions that all four types of IT investments have in overall sum (Figure 1). What can be seen quickly is that investments in hardware (HW) represent 40% of all IT investments in studied period, followed by 34% of investments in telecommunication (TCM), 19% investments went into studies and research projects (RES) and only 7% was invested into software. To our belief this latest number could be also expression of problem of pirate software but it is very difficult to prove this statement statistically. This remains a challenge for future research. To deal with shape of distributions of single data--what needs to be observed is that investment in hardware (HW), investment in telecommunications (TCM), investment in software (SW), investment in research projects (RES) and value added per industry (VA) all show a positive (right) skewness. This means that linear function would not be appropriate function to use. (Our speculation was confirmed by conducting such an analysis, which resulted in extremely high adjusted [R.sup.2] (0.99228) and at the same time almost none regression coefficients were statistically significant (except for hardware investments). This also pointed out problem of multicolinearity.) However, we have a reason to believe that it is in the nature of our independent variables to have very high and significant correlation coefficients (Table 3). Simple reasoning leads us to conclusion that PC without software is worthless (and vice versa VICE VERSA. On the contrary; on opposite sides. ). In times of Internet the same goes for investment in telecommunications. If those 3 variables could be compared to engine that drives our vehicle on the highways of knowledge-based economy, investment in research is the fuel. So we have all the reason to believe that high and statistically significant correlation is the natural state of being for those 4 independent variables. Positive skewness and famous Cobb-Douglas production function lead us to test a new form of production function that would be more suitable for our purposes. We logaritmized values for all 4 independent variables and value added as dependent variable and got much better results. Looking at (or to be more exact--into) distributions of single data we realized that linear function is not appropriate--that is why we decided to utilize log-log functional form (with (2) and without time lag (1) between independent and dependent variables): (1) log VA = [[beta].sub.1] * log HW + [[beta].sub.2] * log TCM + [[beta].sub.3] * log SW + [[beta].sub.4] log RES + [[beta].sub.0] (2) log VA = [[beta].sub.1] * log H[W.sub.-1] + [[beta].sub.2] * log TC[M.sub.-1] + [[beta].sub.3] * log S[W.sub.-1] + [[beta].sub.4] * log RE[S.sub.-1] + [[beta].sub.0] where following denotations mean VA ... value added per industry per year, fixed prices (1995) HW ... investments in hardware TCM ... investments in telecommunication equipment SW ... investments in software RES ... investments in research Using Limdep tool suitable for panel data) we analyzed an·a·lyze tr.v. an·a·lyzed, an·a·lyz·ing, an·a·lyz·es 1. To examine methodically by separating into parts and studying their interrelations. 2. Chemistry To make a chemical analysis of. 3. data using all three functional forms in three steps: OLS, Fixed Effects Model and Random Effects Model accompanied with Hausman test to realize which provides better estimations of regression coefficients (Fixed vs. Random Effects Model). This tool allows us to upgrade OLS with specifics of panel data and calculates two different models: Fixed effects model and Random effects model. Using Hausman test we confirmed that Random effects model provides better estimation of regression coefficients (probability value of Fixed vs. Random effects Random effects can refer to:
Let us now interpret results of regression analysis In statistics, a mathematical method of modeling the relationships among three or more variables. It is used to predict the value of one variable given the values of the others. For example, a model might estimate sales based on age and gender. based on linear functional form (see Table 5). Using Hausman test we determined that Fixed Effects Model provides the best estimations of regression coefficients. However, [R.sup.2] is extremely high (0.99) and at the same time there is just one statistically significant regression coefficient in the model. We can confirm that in linear model a problem of multicolinearity has been detected. By implementing log-log functional form (1) we got two significant coefficients and much more 'normal' [R.sup.2] as can be seen in Table 5. While there is no doubt that, regardless of our belief of benefits of IT and knowledge investments, there are also other factors that have significant impact on value added (like work, other types of capital etc.). We can claim that investment in hardware in Slovenia resulted in rise of value added. On average, 1% rise in hardware investment resulted in 0.9% rise in value added per industry. The same (although in much lesser extent) goes for investment in telecommunication equipment. 1% rise in investment into telecommunication equipment resulted averagely in 0.16 * [10.sup.-8] % rise in value added. In our case, investment in software and investment in research showed no statistically significant correlation to value added. In case of software investments this could be influenced by (according to our estimations) relatively high rate of pirate software, which results in the fact that all 'investments' in software are not registered. What it matters to us is that hypotheses [H.sub.1] and [H.sub.2] were confirmed, while we cannot accept [H.sub.3] and [H.sub.4]. However, first two hypotheses have higher weight since investment into software and telecommunications represents above three quarters of overall IT investments in Slovenia for period 1996-2000. This allows us to discard productivity paradox in the field of IT investment in Slovenia for that period (based on industry data for whole economy). Another question that came to our mind was how time lags influence results of analysis. We conducted same type of multiple regression on panel data for 1-year lag (2) and 2-year lag. This means that we anticipated that there could be a 1- or 2- year time lag from time of hardware, software, telecommunication or research investment to time when it gives results in increased productivity measured by value added. When 2-year lag was introduced, number of observations has decreased to such an extent (from 75 to 45) that all of the regression coefficients became statistically insignificant. Degrees of freedom were too scarce to be able to use this kind of functional form. Results for log-log functional form with 1-year lag (2) can be seen in Table 5 and are very similar to those without lag. Again, first two hypotheses were accepted. Regression coefficient for investments in hardware became slightly higher and slightly more statistically significant than in our basic model without time lags. We believe that this can be contributed to the fact that activation time for investment in IT is shorter than 1 year. We can claim that investment in hardware in Slovenia resulted in rise of value added after 1-year lag. On average, 1% rise in hardware investment resulted in 1.18% rise in value added per industry. 1% rise in investment in telecommunication equipment resulted averagely in 0.108 * [10.sup.-8]% rise in value added. Random effects model proved to be better than Fixed Effects model (probability of Hausman test is higher than 0.05) and its statistical power measured by [R.sup.2] augmented up to 0.77. In Table 5 we summarize sum·ma·rize intr. & tr.v. sum·ma·rized, sum·ma·riz·ing, sum·ma·riz·es To make a summary or make a summary of. sum the most important results. First, in linear form, problem of multicolinearity was detected. Therefore, that kind of functional form was found to be inappropriate. (Reason for that is right skewness of all variables included). Second, in both log-log models (with and without time lag), statistically significant positive influence of two most important independent variables (investments in hardware and investments in telecommunications) on value added was confirmed. 5. CONCLUSIONS Based on panel data for all 15 Slovenian industries in period from 1996-2000 we came to following realizations regarding our research question: 1. First of all, linear function is not appropriate form to estimate production function in our case. Reason for this is positive skewness of distribution of all individual data. Instead of this we used log-log functional form which proved to be much more useful. This corresponds also to functional form of well-known Cobb-Douglas production function. We used examined also effect of 1-year time lag between investments in hardware, telecommunication equipment, software and research as independent variables and value added per industry per year as dependent variable. 2. Second, investments into hardware proved to have statistically significant and positive impact on value added per industry per year. This has been confirmed in all functional forms utilized (linear, log-log without lag and with 1 year lag). Hence, we can accept our first hypothesis that investments in hardware have positive impact on value added as measure of productivity. This is the first part of our research that negates productivity paradox in Slovenian case for period 1996-2000 based on industry data. It is also very important to know that vast majority of IT investment are directed into hardware. This gives additional value and weight to our conclusion. 3. Third, investments in telecommunications proved to have statistically positive impact on value added in both log-log functional forms (although in much lesser extent than investments in hardware). Again, IT investments in telecommunications represent important part of overall IT investments (34%) and together with investments in hardware compose com·pose v. com·posed, com·pos·ing, com·pos·es v.tr. 1. To make up the constituent parts of; constitute or form: nearly three quarters of all IT investments in Slovenia in period from 1996 till 2000. 4. Fourth, investment in software and research did not prove to have any statistically significant impact on value added. We cannot accept hypothesis 3 and 4. However, those investments represent below one quarter of total IT investments in Slovenia in observed period, so their weight is much lower than the weight of investments in hardware and telecommunications equipment. Ergo Latin, therefore; hence; because. ergo (air-go) conj. Latin for therefore, often used in legal writings. Its most famous use was in "Cogito, ergo sum:" "I think, therefore I am" principle by French philosopher Rene Descartes (1596-1650). , the productivity paradox tested on panel data for all 15 Slovenian industries (SIC: A-O) in period from 1996-2000 can be rejected. Additional weight to our statement gives the fact that most (74%) of IT and knowledge investment in distributed into first 2 variables, which have statistically positive impact on productivity measured by value added per industry. Directions for future research to test productivity paradox in Slovenian case could be to include some measures of labor and capital inputs to estimate a production function (such as Cobb-Douglas) or to expand time series. All those future research depends heavily on the availability and accessibility of data. Time series was limited due to the fact that SKD SKD Skilled SKD Semi Knocked Down SKD Samarkand, Uzbekistan (Airport Code) SKD Sigma Kappa Delta (National English Honor Society) SKD Shikimate Dehydrogenase SKD Srpsko Kulturno Drustvo (Slovenian SIC) was not implemented before 1995 and at the time when we conducted our research data for year 2000 were the latest possible. Logical next step in our research would be also to expand research to micro and mezzo mez·zo n. pl. mez·zos A mezzo-soprano. mezzo Adverb Music moderately; quite: mezzo-forte Noun pl -zos level and to drill down into managerial implications of observed results. Differentiation between public and private sector could be of significant importance as well. What would be also interesting to know for Slovenia is how do investments in computer literacy Understanding computers and related systems. It includes a working vocabulary of computer and information system components, the fundamental principles of computer processing and a perspective for how non-technical people interact with technical people. education, programming and other types of IT education impact productivity and whether productivity paradox can be found there or not. Either way, (in context of previously described limitations) our present research in Slovenia rejects it.
TABLE 1: RESEARCH HYPOTHESES
No. Hypothesis text
[H.sub.1] Hardware investments are positively correlated with value
added per industry.
[H.sub.2] Telecommunication investments are positively correlated
with value added per industry.
[H.sub.3] Software investments are positively correlated with value
added per industry.
[H.sub.4] Investments for research are positively correlated with
value added per industry.
TABLE 2: CONSUMER PRICE INDICES, BASE YEAR 1995.
Year 1996 1997 1998 1999 2000
Consumer Price Indices 109,9 119,1 128,5 136,4 148,5
Source: Statistical Yearbook of the Republic of Slovenia 2001.
TABLE 3: DESCRIPTIVE STATISTICS, 1000 SIT FIXED PRICES 1995.
HW TCM SW
Mean 1980451.21 1716792.96 340984.10
Median 1148363.42 569398.44 150318.18
Mode 1668.48 (a) 191.25 (a) .00
Std. Deviation 1946572.45 4557954.91 468264.89
Minimum 1668.48 191.25 .00
Maximum 7076565.25 27551305.72 2063934.01
RES VA
Mean 977299.51 145185.48
Median 123253.20 109175.00
Mode .00 409.00
Std. Deviation 2082868.58 145483.95
Minimum .00 409.00
Maximum 9911016.79 692199.00
(a) Multiple modes exist. The smallest value is shown
TABLE 4: PEARSON CORRELATION COEFFICIENTS AMONG INDEPENDENT VARIABLES
HW TCM SW RES
HW 1 0.23 * 0.57 ** 0.29 *
TCM 0.23 * 1 0.38 ** 0.07
SW 0.57 ** 0.38 ** 1 0.15
RES 0.29 * 0.07 0.15 1
* Correlation is significant at the 0.05 level (2-tailed).
** Correlation is significant at the 0.01 level (2-tailed).
TABLE 5: RESULTS OF PANEL DATA ANALYSIS--FIXED EFFECTS MODEL FOR LINEAR
AND RANDOM EFFECTS FOR LOG-LOG FUNCTIONAL FORM WITH AND WITHOUT 1-YEAR
TIME LAG.
Linear functional Log-log model, no
form, Fixed Effects time lag, Random
Model Efects Model
[R.sup.2] 0.99 0.53
[B.sub.1] 0.51 *** 0.9 **
[B.sub.2] -0.46E-03 0.16E-08 ***
[B.sub.3] -0.63E-01 0.36E-01
[B.sub.4] -0.18E-01 0.13E-09
[B.sub.0] / -1.33
Hausman 114.49 (0.000) *** 7.05 (0.133)
test (prob.
value)
Log-log model, 1 year
lag, Random Effects
Model
[R.sup.2] 0.77
[B.sub.1] 1.18 ***
[B.sub.2] 0.11E-08***
[B.sub.3] -0.13E-01
[B.sub.4] -1.20E-11
[B.sub.0] -0.12E-10
Hausman 8.05 (0.896)
test (prob.
value)
* Significant at P < 0.05 (2-tailed).
** Significant at P < 0.005 (2-tailed).
*** Significant at P < 0.001(2-tailed).
FIGURE 1: PERCENTAGES OF 4 TYPES OF IT INVESTMENTS IN SLOVENIA,
PERIOD 1996-2000, FIXED PRICES.
HW 40%
TCM 34%
RES 19%
SW 7%
Note: Table made from pie chart.
REFERENCES [1] Alpar, P. and Kim M., "A Microeconomic mi·cro·ec·o·nom·ics n. (used with a sing. verb) The study of the operations of the components of a national economy, such as individual firms, households, and consumers. Approach to the Measurement of Information Technology Value", Journal of Management Information Systems The Journal of Management Information Systems (JMIS) is an academic journal that publishes original peer-reviewed research articles in the areas of Information Systems and Information Technology. , Vol. VII (3), 1990, 55-69. [2] Andersen, T.J. and Segars, A.H., "The impact of IT on decision structure and firm performance: evidence from textile and apparel industry", Information and management, No. 39, 2001, 85-100. [3] Andreu, R. and Ciborra, C., "Organizational Learning and Core Capabilities Development: The Role of IT", Journal of Strategic Information Systems, No. 5, 1996, 111-127. [4] Argyris, C. and Schon, D.A., Organizational Learning: A Theory of Action Perspective. Reading, MA: Addison-Wesley, 1978. [5] Barua, A., Kriebel, C.H. and Mukhopadhay, T., "Information Technologies and Business Value: An Analytic and Empirical Investigation", Information Systems Research, Vol. VI (1), 1995, 3-23. [6] Banker, R.D. and Kauffman, R.D., "Strategic contributions of Information Technology: An Empirical Study of ATM Networks", Proceedings of the Ninth International Conference on Information Systems International Conference on Information Systems (ICIS) is an annual international conference for academics and research-oriented practitioners in the area of Information Systems. , Minneapolis, Minessota, 1988. [7] Benyon-Davies, P., Information Systems--An Introduction to Informatics Same as information technology and information systems. The term is more widely used in Europe. in Organisations. Palgrave, 2002. [8] Brady, T. and Target, D., "Strategic Information Systems in the Banking Sector: Holy Grail Holy Grail: see Grail, Holy. A very desired object or outcome that borders on a sacred quest. There are several Holy Grails in the computer business. or Poison Chalice", Technology Analysis and Strategic Management, Vol. 7(4), 1995, 387-406. [9] Brynolfsson, E. and Hitt, L.M., "Beyond the Productivity Paradox. New York New York, state, United States New York, Middle Atlantic state of the United States. It is bordered by Vermont, Massachusetts, Connecticut, and the Atlantic Ocean (E), New Jersey and Pennsylvania (S), Lakes Erie and Ontario and the Canadian province of : Association for Computing Machinery See ACM. Association for Computing Machinery - Association for Computing ", Communications of the ACM (publication) Communications of the ACM - (CACM) A monthly publication by the Association for Computing Machinery sent to all members. CACM is an influential publication that keeps computer science professionals up to date on developments. , Vol. 41(8), 1998, 49-55. [10] Dewan, S. and Kraemer, K.L., "International Dimensions of the Productivity Paradox", Communications of the ACM, Vol.41 (8), 1998, 56-62. [11] Dewan, S. and Kraemer, K.L., "Information Technology and Productivity: Evidence from Country Level Data", Management Science, Vol. 46 (6), 2000. [12] Dos Santos, B.L., Peffers, K. and Mauer, D.C.: The Impact of Information Technology Investment on Announcements on the Market Value of the Firm", Information Systems Research, Vol. 4(1), 1993, 1-22. [13] Dos Santos, B.L. and Sussman, L.: "Improving the Return on IT Investment: The Productivity Paradox", International Journal of Information Management, No. 20, 2000, 429-440. [14] Fitzgerald, J. and Dennis, A., Business Data Communications data communications, application of telecommunications technology to the problem of transmitting data, especially to, from, or between computers. In popular usage, it is said that data communications make it possible for one computer to "talk" with another. and Networking. John Wiley John Wiley may refer to:
[15] Freeman E.R., Strategic Management--A Stakeholder Approach, London: Pitman, 1984. [16] Freeman E.R., "Politics of Stakeholder Theory As originally detailed by R. Edward Freeman (1984), stakeholder theory identifies and models the groups which are stakeholders of a corporation, and both describes and recommends methods by which management can give due regard to the interests of those groups. : Some Future Directions", Business Ethics business ethics, the study and evaluation of decision making by businesses according to moral concepts and judgments. Ethical questions range from practical, narrowly defined issues, such as a company's obligation to be honest with its customers, to broader social Quaterly, No. 4, 1994, 409-422. [17] Greene, W.H., Econometric e·con·o·met·rics n. (used with a sing. verb) Application of mathematical and statistical techniques to economics in the study of problems, the analysis of data, and the development and testing of theories and models. Analysis, 2nd edition, New York: MacMillan, 1993. [18] Harris, S.E. and Katz, J.L., "Organizational Performance Organizational performance comprises the actual output or results of an organization as measured against its intended outputs (or goals and objectives). Specialists in many fields are concerned with organizational performance including strategic planners, operations, and Information technology Investments Intensity in the Insurance Industry", Organization Science, Vol. 15(3), 1991, 333-344. [19] Hartwig, F. and Dearing, B.E., Exploratory Data Analysis, Sage Publications This article or section needs sources or references that appear in reliable, third-party publications. Alone, primary sources and sources affiliated with the subject of this article are not sufficient for an accurate encyclopedia article. , 1979. [20] Haynes, R.M., "The ATM at the Age Twenty: A Productivity Paradox", National Productivity Review, Vol.9 (3), 1990, 273-280. [21] Kaplan, R.S. and Norton, D.P, "Balanced Scorecard--Measures That Drive Performance", Harvard Business Review Harvard Business Review is a general management magazine published since 1922 by Harvard Business School Publishing, owned by the Harvard Business School. A monthly research-based magazine written for business practitioners, it claims a high ranking business readership and , 1992, 71-79. [22] Kaplan, R.S. and Norton, D.P., "Putting the Balanced Scorecard to Work", Harvard Business Review, 1992, 134-147. [23] Kaplan, R.S. and Norton, D.P., "Using the Balanced Scorecard as Strategic Management System", Harvard Business Review, 1996, 75-85. [24] Kaplan, R.S. and Norton, D.P., The Balanced scorecard, Boston: Harvard Business School Harvard Business School, officially named the Harvard Business School: George F. Baker Foundation, and also known as HBS, is one of the graduate schools of Harvard University. . 1996a. [25] Kraemer, K.L. and Dedrick, J., Asia Computer Report 1996, Centre for Research on Information Technology and Organizations, University of California The University of California has a combined student body of more than 191,000 students, over 1,340,000 living alumni, and a combined systemwide and campus endowment of just over $7.3 billion (8th largest in the United States). , http://www.crito.uci.edu. [26] Kraemer, K.L. and Dedrick, J., Asia E-Commerce Report 2000, Centre for Research on Information Technology and Organizations, University of California, http://www.crito.uci.edu. [27] Kraemer, K.L. and Dedrick, J., "Information Technology and Productivity: Results and Policy Implications of Cross-country Studies", Pohjola, M. (ed.): Information Technology, Productivity and Economic Growth: Implications for Economic Development. Oxford University Press, 2001. [28] Kraemer, K.L., Tallon, P.P. and Dunkle, D., "Performance Benchmark for Information Systems in Corporations", Centre for Research on Information Technology and Organizations, University of California, 2001, http://www.crito.uci.edu. [29] Loveman, G.W., "An Assessment of the Productivity Impact of Information Technologies", Sloan School of Management, Massachusetts Institute of Technology Massachusetts Institute of Technology, at Cambridge; coeducational; chartered 1861, opened 1865 in Boston, moved 1916. It has long been recognized as an outstanding technological institute and its Sloan School of Management has notable programs in business, , Management in the 1990s, Working Paper No. 8, 1988, 88-154. [30] Melville, N., "Impact of IT Investment: An Industry Analysis", Working Paper, Center for Research on Information Technology and Organizations, University of California, 2001, http://www.crito.uci.edu. [31] Navarette, C.J. and Pick, J.B., "Information Technology Expenditure and Industry Performance: The Case of the Mexican Banking Industry", Journal of Global Information Technology Management, Vol. 5 (2), 2002, 7-28. [32] Plice, R, "A Contribution to the Empirics of IT Returns", Working Paper, Centre for Research on Information Technology and Organizations (CRITO), University of California, 2001, http://www.crito.uci.edu. [33] Roach, S., "America's Technology Dilemma: A Profile of the Information Economy", New York: Morgan Stanley  To comply with Wikipedia's , the introduction of this article needs a complete rewrite. . Economics
Newsletter Series, April 22, 1987.[34] Robey, D., Boudreau, M. and Rose, G.M., "Information technology and organizational learning: a review and assessment of research", Accounting, Management and Information Technologies, No.10, 2000, 125-155. [35] Senge, P.M., The Fifth Discipline, Sloan MIT MIT - Massachusetts Institute of Technology , 1995. [36] Schrage, M.: "The Real Problem with Computers", Harward Business Review, Vol. 75 (5), 1997, 178-188. [37] Shao, B.B.M. and Lin, W.T., "Technical Analysis of Information Technology Investments: A Two-stage Empirical Investigation", Information & Management, No. 39, 2002, 391-401. [38] Strassman, P.A., "Computers have yet to make Companies More Productive", Computerworld, September 15, 1997a. [39] Strassman, P.A., "Information Productivity", Information Economics Press, 1999. [40] Strassman, P.A., "Facts and Fantasies about Productivity: Report", 1997b, http://www.strassman.com/pubs/fnf/factnfantasy.shtml. [41] Stratopoulos, T. and Dehning, B., "Does Successful Investment in Information technology solve the Productivity Paradox? ", Information and Management, No. 38, 2002, 103-117. [42] Turban, E., McLean, E. and Wetherbe, J., Information Technoloqy for Management, 2nd edition. New York: John Wiley and Sons, 2001a. [43] Turban, E., Rainer, R.K. and Potter, R.E., Introduction to Information Technology, New York: John Wiley and Sons, 2001 b. [44] Tukey, J.W, Exploratory Data Analysis, Reading, MA: Addison-Wesley, 1977. [45] Weill, P., "The Relationship Between Investment in Information Technology and Firm Performance: A Study of The Valve Manufacturing Sector', Information Systems Research, Vol.3 (4), 1992, 307-332. [46] Wilcocks, L. and Lester, S., "Beyond the IT Productivity Paradox", European Management Journal, Vol. 14 (3), 1996, 279-290. Dr. Vlado Dimovski earned his DBA in Management and Finance at Cleveland State University Cleveland State University, at Cleveland, Ohio; coeducational; founded 1964, incorporating Fenn College (est. 1923). The Cleveland-Marshall School of law was incorporated in 1969. in 1994. Currently he is a Minister of Labor, Family and Social Affairs of Republic of Slovenia and an Associate Professor at the University of Ljubljana The University of Ljubljana (in Slovenian, Univerza v Ljubljani; in Latin, Universitas Labacensis) is the first and the largest university in Slovenia; with 56,000 enrolled students, it ranks among the biggest universities in the world. , Faculty of Economics. Miha Skerlavaj earned his B.S. degree in Economics at University of Ljubljana. In 2001 he received faculty Preseren award for his final thesis "e-Economy: Comparison between USA and EU". Currently he is a Teaching Assistant at the University of Ljubljana, Faculty of Economics. |
|
||||||||||||||||||||
Printer friendly
Cite/link
Email
Feedback
Reader Opinion