Information technology productivity has been studied on a macroeconomic level, a specific application level, as well as at a firm level. None of these studies has firmly determined whether there is an absence of productivity gains despite increased information technology expenditures (often known as the Productivity Paradox). This study reviews a specific year, 1995, using publicly available InformationWeek data from their IW 500 and analyzes productivity of the firm using both market and

more »

... ial based measures stratified by overall industry type. This study is to determine whether a Productivity Paradox is observed after eliminating the overall industry impact. Many researchers have studied at the macroeconomic level and found differing results. Strassman (1999) and Berndt and Morrison (1991) found little or no gains from IT expenditures at the macro level. Clemons and Weber (1990) and Morrison (1996) did find gains. Peslak © 2003 JITI 78 An example of a macroeconomic study was prepared by Dewan and Kraemer (1998) . The authors examine the Productivity Paradox from another perspective, by country. Up until this study all analyses had been performed on U.S. companies or U.S. multinational companies. Dewan and Kraemer examine information technology productivity across international boundaries and study data from 17 developed countries that invested significant dollars in information technology. They first review the U.S. literature and note how early studies showed no positive impact on productivity (the investment or Productivity Paradox period). More recent studies in the 1990s, however, have shown positive correlation between information technology investments and productivity (the payoff era). They also note the work of Sichel, Roach, and Strassmann who still argue that information technology productivity is not increasing. They suggest that these authors research is not "rigorous", but do note that caution still must be exercised and further research is still needed. With regard to their 17 country international study, they culled international data from IDC, Penn World tables, the International Labor Organization, and the International Monetary Fund. Their key data was gross domestic product, information technology capital, number of workers, and non-information technology capital. They use a traditional Cobb-Douglas production function and used ordinary least squares regression analysis to analyze the data. The results of the study show a positive correlation between change in information technology capital per worker vs. gross domestic product per worker, controlling for changes in non-information technology capital per worker and the number of workers. The average return on information technology capital in this study was 70.6 percent across all 17 countries while the U.S. return was 59 percent. This is consistent with the Paradox lost study by Brynjolfsson and Hitt (1996) . Daniel Sichel, the Federal Reserve Board economist, in Computers and Aggregate Economic Growth: An Update (1999), an article in Business Economics, attempted to review the Productivity Paradox at a macro level. He suggests there are two possible reasons for the sluggish productivity growth at the macro level. The first is a time lag effect. It simply may take time to reap the benefits of the past computer investment. The second possible reason is that the recent contribution from computers is the result of transitory factors. The transitory factors suggested by Sichel are that in recent years, computer hardware prices have dropped precipitously, resulting in more purchases and wider dissemination of technology. He suggests this rapid price decline, dissemination, and contributions to productivity may be an anomaly and may be short-lived. He studies Bureau of Labor Statistics and other government data, and estimates the growth rates of output over the time frame and the contribution from computer hardware. This analysis suggests that in the 1970s, 1980s, and early 1990s, the output growth contribution from computer hardware was extremely small, relative to other factors. Though it has significantly increased in the last few years, it still represents a small portion of output growth. He makes these estimations of the impact of output of computer capital by using a neoclassical approach. This has as its key assumption that business makes the best investment decisions, and all types of capital earn the same return at the margin. Application level studies Positive effects were found in specific application level studies such as Mukhopadhyay, Lerch, and Mongol (1997) and Mukhodpadhyay, Kekre, and Kalathur (1995). Offsetting these studies A Firm Level Study of Information Technology Productivity