4.1 - Choice of Variables
4.1.1 - Modularization and Outsourcing
In Section 2, we examined the possible effects of modularization on information service productivity. We showed that modularization enables firms to outsource parts of their products in an efficient way. Thus, the degree of outsourcing may be a measure of the degree of modularization, although the degree of modularization itself is not observable. Consequently, in our empirical analysis, we use the degree of outsourcing, which is the ratio of outsourcing to total sales in information service activities, as a variable for modularization and resulting outsourcing.
There are three kinds of modularization:
See Fujimoto (2002). (a) modularization of product architecture (modularization in development process), (b) modularization of production process, and (c) modularization of supplier relations. Outsourcing is a result of (c), which is based on (b) and (a). However, outsourcing may not be motivated solely by modularization, and other factors may prompt firms to outsource their business activities. We will explicitly consider the latter possibility in our later interpretation of the empirical results obtained in this paper.
4.1.2 - Scale of Development Organization and the Number of System Engineers
The second determinant we examined in Section 2 is the scale of software development. Ideally, if a measure of the average quality-adjusted scale of software development in a firm is available, this is the variable that should be used for the empirical analysis. However, since the average scale of software development is not directly observable, we need a proxy for this variable.
Software development is skilled-labour intensive, and thus the scale of development is likely to be highly correlated with the size of the skilled workforce. In the context of information service industries, and especially the software industry, the skilled workforce includes system engineers, programmers, and research scientists, about which we have data from theSurvey of Selected Service Industries, Volume of Information Service Industries. System engineers play a pivotal role in software development, especially in custom software one, which dominates Japanese software development.
Tailoring software to the demands of customers is of utmost importance. System engineers in the worker classification of the Survey are those responsible for receiving customers’ needs and interpreting them into schemes that are programmable. They are responsible for deciding on the most suitable development languages and organizing the programmers to get the best out of them. Based on the importance of system engineers to Japanese software development, we use their number as a proxy for the scale of software development.
We tried number of programmers and number of research...
4.1.3 - “Profit-driven R&D Investment” Hypothesis and the Profit-to-Cost Ratio
Finally, it has often been argued that the higher a firm’s profits, the higher the level of R&D activities (and thus TFP), especially when a firm is liquidity constrained. This “profit-driven R&D investment” may be relevant in the information service industries in Japan since relatively small firms in our samples may have been liquidity-constrained in the period 1991-1998. To control for this possibility, we consider the ratio of gross profits to operating costs of information service activities as one determinant of TFP.
4.2 - Models and Estimation
Taking account of the arguments above, we formalize the level of TFP in information service industries as follows:
Here i denotes the i-th firm and t denotes the year. Equation (1) implies that the TFP of the i-th firm’s information service activities in year t, TFPi,t, is determined by a constant α0, a stochastic trend factor α1it (explained later), microeconomic variables xij,t discussed in the previous sections, time dummy dk,t representing macroeconomic conditions (which is 1 if t = k and 0 if otherwise), and disturbances εi,t. In this formulation, the stochastic trend factor varies across firms, representing firm heterogeneity.
As explained in the previous section, we consider the following microeconomic determinants xij,t.
See Appendix for the way these variables are const...
1) OUT: the ratio of outsourcing expenditure to total sales in information service activities, which is used as an index of modularization.
2) lnSE_number: the logarithm of the number of system engineers, which is used as an index of the scale of development.
3) PROFIT: the ratio of gross profits to operating expenses in information service activities.
We use a standard growth accounting procedure to find firms’ TFP growth, and we base our analysis on the following “growth” or first-difference formulation (2).
where ΔYt = Yt + 1 – Yt and ei,t = εi,t + 1 – εi,t. Equation (2) depicts how firm i’s TFP growth is determined.
We considered two additional factors. As explained before, information services are categorized in several subgroups such as custom software and data base services
We classify firms into product subgroups in such a.... It is quite likely that TFP growth is similar within subgroups, but different between them. In addition, TFP growth might be dependent on firm-specific idiosyncratic factors that are not observable. Taking these two factors into account, we assume that the (stochastic) term α1,i in TFP growth regression (2) is the sum of a constant α1, product-subgroup dummies Σγihyih where yih = 1 if firm i belongs to subgroup h and = 0 otherwise, and an unobservable idiosyncratic random variable ui
Then equation (2) can be written as an ordinary one-way error component regression model (3) with differences in explanatory variables Δxij,t (d-OUT, d-SE, d-PROFIT, where d- denotes difference) as well as product subgroup dummies yih and differences in time dummies Δdk,t.
To estimate (3), we take account of possible simultaneity explicitly between TFP growth and explanatory variables. It is likely that explanatory variables, in particular the outsourcing-to-sales ratio and the profit-to-cost ratio, are endogenous and thus they may be correlated with errors in equation (3). To deal with this issue, we employ an instrumental variables method for panel data. We use GLS estimators of the random effects model (Baltagi and Chang, 2000). As instruments, we use the first lags of all explanatory variables and current government investment and expenditure.
We tried other sets of instruments, but found that... The results are reported in the first and the second columns of Table 4. The first column shows the result of the regression analysis ignoring product subgroup heterogeneity, and the second incorporates the heterogeneity. To save space, we omit the results of time dummies, which are statistically significant and quite similar for all regression equations.
4 - Estimation Results: 1991-1998
dependent variable: TFP growth IV IV Panel OLS Panel OLS d-OUT – 1,1175*** (0,2239) – 1,1245*** (0,2233) – 0,8343*** (0,0672) – 0,8387*** (0,0672) d-lnSE_number – 0,2724*** (0,0435) – 0,2706*** (0,0434) – 0,0591*** (0,0101) – 0,0593*** (0,0102) d-PROFIT 0,0562*** (0,0064) 0,0561*** (0,0065) 0,0761*** (0,0030) 0,0760*** (0,0030) Constant – 0,0413 (0,0302) – 0,0263 (0,0183) – 0,1574*** (0,0362) 0,1481*** (0,0367) Time Dummy Yes Yes Yes Yes Product-Subgroup Dummy No Yes No Yes Adj. R-squared 0,1706 0,1823 0,0907 0,1051 # of Firms 1086 1086 1106 1106 # of Firm-Year’s 5346 5346 6117 6117 Notes: Standard deviations are in parenthesis. “***”, “**”, and “*” denote significance at 1%, 5%, and 10%, respectively. IV denotes the instrumental-variable method for panel data and instruments are the first lags of explanatory variables and government investment and expenditure. Panel OLS reports the results of the random-effect model (see the text). The number of firms each year varies because of the unbalanced nature of our panel.
To check for robustness, we also report the results of Panel OLS. Since random effect models are accepted by Hausman’s specification tests, we only report the results of random effect models.
Furthermore, we examined the panel AR(1) method assuming... In Table 4, all coefficients (except for constant terms in IV) are statistically significant at the 1% level.
4.3 - Results and Discussion
Let us consider the estimation results in Table 4. We have fairly consistent and significant results: the coefficients of the d-OUT, d-SE, and d‑PROFIT are significant at the 1% level, where “d-” means the first difference. The sign of d-OUT and d-SE is negative, while that of d-PROFIT is positive. Time dummies (not reported here) are all significant at the 1% level.
4.3.1 - Outsourcing and Remnants of Traditional Relationship
A remarkable result found in Table 4 is that the outsourcing-to-sales ratio (OUT) has a significant negative effect on TFP. This result is quite robust in all specifications. Moreover, close examination of data shows that it is not a simplistic case that less productive firms are outsourcing their business to more productive firms. Outsourcing firms are often major ICT vendors, which supposedly have high productivity.
The robust negative effect of outsourcing is striking. As in the Section 2, modularization behind outsourcing should improve rather than reduce productivity. Thus, a robust negative result of outsourcing on TFP suggests that outsourcing in Japanese information service industries has a different origin from modularization, which hinders productivity.
In fact, in-depth analysis of the industrial structure of the Japanese software industry suggests that the negative effect of outsourcing stems from the remnants of traditional subcontracting practices found in this industry. In the traditional relationship, there are main contractors (often major ICT vendors) on the one side, who get contracts from customers which are often large corporations and central and local governments. On the other side of the relationship, there are sub-contractors (medium-to-small firms) that depend mostly on the main contractors to allocate business to them.
Main contractors outsource their business not to promote efficiency in software development, but to make sub-contractors act as “buffers” against economic fluctuations to reduce the costs of adjustments necessitated by such fluctuations. It is sometimes argued that this cost consideration of main contractors leads to “over-outsourcing” to sub-contractors in the sense that programming expertise is not properly retained within these main contractors. If this is the case, they are violating the dictum of “do not outsource the core of your competence.”
For sub-contractors, their poor financial position makes them unable to take advantage of outsourcing. These sub-contractors have little human-capital or physical-capital investment.
This inefficient subcontracting system cannot survive, however, if new efficient firms enter the market. Unfortunately, buyers’ (consisting of local and central governments and large corporations) preferences to use “established vendors” makes the entry of new, especially small, firms difficult and thus this industrial structure can exist for a long time.
4.3.2 - Mythical Man-Month and Communication Problems
In Section 2, we argued that scale of development may affect productivity positively in some cases, and negatively in others, depending on the smoothness of communication between customers and development team members and among development team members themselves. The results of our study strongly suggest diseconomies of scale in software development: the scale of development organization affects firms’ productivity negatively.
Thus, the phenomenon of the mythical man-month dominates Japanese information service industries. Effective communication to reduce costs between system engineers and their customers, and among system engineers themselves, is lacking, and this is likely to be one of major obstacles to Japanese firms improving the efficiency of their information service provision.
In this respect, central and local governments should play a major role in reversing this tendency. Central and local governments are the major customers of information service firms. However, governments are often shown to be incompetent in articulating their needs when they place orders with information service firms. As a result of their lack of eloquence, communications between governments as customers and system engineers in supplier firms are not smooth, and in some cases, this leads to last-minute specification changes that in software development are very costly.