|
Article Excerpt
1. Introduction
The swift development of information and communication technologies (ICT), as well as the declining prices for its use, have considerably enhanced the diffusion of ICT during the last few years. As a consequence, the impact of ICT on productivity has become a broadly discussed topic in management sciences and economics. Several studies find empirical evidence for positive productivity effects of ICT at the firm level (Brynjolfsson and Hitt 1996, Lichtenberg 1995, Greenan and Mairesse 2000, Licht and Moch 1999). Although it seems reasonable that ICT also has an indirect effect on labor productivity by enabling firms' reorganization of workplaces, researchers have only recently become interested in the joint effects of workplace organization and ICT on labor productivity. Studies like that of Black and Lynch (2001), Brynjolfsson and Hitt (2000), and Bresnahan et al. (2002) find empirical evidence that ICT and workplace reorganization do, in fact, have positive and significant effects on labor productivity.
A flaw of these studies is that they assume a unidirectional relationship between labor productivity and workplace reorganization. They take the view that workplace reorganization affects labor productivity, but ignore a potential reverse causality, because a main reason for firms to reorganize workplaces is to increase labor productivity. Such a simultaneity renders the economic interpretation of the results presented in earlier studies questionable and casts doubt on any recommendations for management practices based on them. (1)
The merits of this paper are twofold. First, it takes the potential simultaneity between labor productivity and firms' decisions to reorganize workplaces into account by estimating an endogenous switching regression model for a sample of 411 firms from the German business-related services sector. Second, it allows for complementarities in firms' organizational design; e.g., it allows workplace reorganization to change any parameter of the production function. Hence, we apply a general and flexible econometric methodology.
In our model, firms are assumed to reorganize workplaces if the productivity gains arising from the reorganization exceed the associated reorganization costs. The reorganization decision defines two labor productivity equations--or "regimes"--one which involves firms with workplace reorganization, and another regime including firms without such a change in human resources management. Besides taking a potential simultaneity into account, the switching regression model also allows workplace reorganization to change the entire set of partial productivity elasticities instead of a priori restricting workplace reorganization to act as a productivity shift parameter in the productivity equations, as earlier studies do. Our estimation results indeed indicate that it is worthwhile to allow for a more flexible effect of workplace reorganization on labor productivity.
Moreover, this paper provides evidence for two distinct types of workplace reorganization: enhancement of group work, established by 39% of the firms in our sample, and flattening of hierarchy levels, introduced by 28% of the analyzed firms.
Our estimation results clearly indicate that labor productivity and workplace reorganization are simultaneously determined. We find that the individual output elasticities of ICT investment, non-ICT investment, and labor do not significantly differ between firms with and without workplace reorganization and that there are insignificant differences in the returns to scale between the two regimes. The point estimates of the partial output elasticities of labor and non-ICT investment are, however, larger for the set of firms that conducted a workplace reorganization, but the coefficients do not differ significantly from each other.
We conduct a counterfactual analysis related to the questions: (i) what would have been the effect of workplace reorganization on productivity for a firm without changes in human resources practises if it had changed the organization of workplaces, and (ii) what would have happened to the productivity of a firm that changed the organization of workplaces if it had not changed it? We visualize the joint differences in the point estimates by plotting the entire labor productivity distributions of firms with workplace reorganization and of firms without workplace reorganization, using Kernel density estimation. Our results indicate that the firms in our sample, on average, reached the right decision: Only those firms that reorganized workplaces actually gained from the reinforcement of group work or the flattening of hierarchies, while firms that did not introduce changes in workplace organization would not have realized gains in productivity.
This paper is organized as follows: [section]2 briefly reviews the existing literature. Section 3 presents the theoretical framework as well as the empirical model. Section 4 introduces the data set; [section]5 presents and interprets estimation results, and [section]6 concludes.
2. Earlier Research
Until recently, two main strands of literature have dealt with the relation between ICT investment, organizational change, and productivity. One branch concentrates on the impact of ICT investment on organizational change. For instance, Leavitt and Whisler, as cited by Crowston and Malone (1988, p. 1051), already predicted in 1958 that "the use of information and communication technology would lead to the demise of middle management" and that the number of hierarchy levels in organizations will decrease if, for example, computers are used increasingly often to perform the functions of middle management. During the 1970s and 1980s, there was a broad discussion about the effects of ICT on workplace organization, with ICT being loosely defined as something in-between a new payroll system and a new personal computer. Due to binding data restrictions, few empirical analyses of the relationship between workplace organization and ICT exist for that time period.
The other branch of the literature deals mainly with the impact of workplace organization or human resources management on labor productivity (Black and Lynch 1996, Eriksson 2003, Huselid 1995, Ichniowski et al. 1997, Milgrom and Roberts 1990).
Studies on the effects of ICT and organizational change on firms' productivity emerged only recently. It seems plausible that the implementation of a new information and communication system alone is not sufficient to cause positive productivity effects. The implementation of a new software system such as SAP often requires a restructuring of the firm to use the new system efficiently. Thus, it appears likely that workplace reorganization has to be changed accordingly to make workflow more efficient or, to put it differently, that ICT is enabling organizational change, as pointed out recently by Brynjolfsson and Hitt (2000). Related evidence is provided by Black and Lynch (2001), who analyze the productivity effects of several workplace practices, ICT, and human capital, using cross-sectional and panel data estimation on a sample of about 600 firms of the U.S. manufacturing industry. Their results indicate that workplace reorganization has positive and significant effects on labor productivity. Bresnahan et al. (2002) also find empirical evidence that ICT and workplace reorganization as well as new products and services positively affects the demand for skilled labor and firms' labor productivity. Their analysis is based on a data set of 300 large U.S. firms from manufacturing industries and services.
3. Theoretical Background
3.1. Complementarities in Firm Strategies
It is likely that firms with organizational changes not only differ from other firms with respect to their organizational form, but also in various other respects such as skill mix or investment strategies....
|
