Thursday, April 3, 2008

Production Sharing, Industrial Adjustments and Wage: an Indonesian Case

Imelda Maidir

executive summary

Indonesia provides a stark example that the social outcomes of the adjustment process attendant to increased economic integration and globalization leave much to be desired. On the one hand, Indonesia appears to have been a spectacular success in terms of the dramatic shift in the structure of its exports with probably the East Asian country currently most dependent on technology intensive products, primarily semiconductors. On the other hand, the country’s manufacturing sector has in fact failed to generate employment and growth for the whole economy, with one of the lowest shares of manufacturing to total employment and output in the region. Indeed, the share of manufacturing to total employment declined in the 1990s at the same time that the share of manufactured products dominated Indonesian exports. This explains in part the failure of Indonesia to have a large decline in poverty incidence, considering that the manufacturing sector tends to provide more remunerative wages than the informal sector in the services industry, especially retailing, where a substantial increase in employment occurred.

In short, ensuring smooth industrial adjustment and at the same time engendering favorable social outcomes is clearly not an easy task; indeed, this invariably involves active engagement of the stakeholders in the labor market; i.e., labor, management and the government, in terms of refining the regulatory environment in the workplace, improvement in the investment climate and economic governance, and stronger cooperative mechanisms between labor and management toward productivity improvement and productivity sharing. Moreover, ensuring inclusive growth in the face of relentless competition domestically and abroad may necessitate affirmative government interventions in order for the less endowed to be able to reap the benefits of globalization and growth.

This study, in particular, addresses the impact of greater economic integration and market competition that increase the pressures for industrial upgrading to wage. It poses the following questions: First, what lessons can be learned for the design of policies for furthering increased worker welfare from the experiences of industry adjustment and firm adjustment in Indonesia. Second, how the industrial relations environment and labor laws and institutions can be improved in order to have smoother industrial adjustment and upgrading process while at the same time facilitating increase in workers welfare. Lastly, what overall trade and development strategies contribute well to increase in wages consistent with the rise in labor productivity.

There has been a growing debate regarding the extent to which workers have benefited from recent trends in the global economy. Although increased globalization in the form of production sharing and foreign direct investment, global sourcing activities of MNCs, has led to strong output growth in some economies, the extent to which these gains have been passed on to workers in the form of improved wages and earnings is questionable; rather globalization has been blamed for the deteriorating position of low-skilled workers globally.

Given the contending views, the impact of export-oriented MNCs, as a reliable measure of production sharing, on real wages in developing countries manufacturing remains very much an empirical issue. However, one threat to validity in establishing empirical regression model is misspecification model, owing to omitted of irrelevant variables to explain the relationship of the phenomenon. Literature review can be done to summarize the contending theory around the explanatory variable that associated with the phenomenon. Nevertheless, there remains subjectivity in the reviewing the literature (Stanley, 1989). Besides, through literature review we might find difficulty in understanding the magnitude of a certain relation.

Through meta-regression analysis we attempt to estimate the empirical magnitude of the association between the global sourcing of MNCs and wage in a more objective and structural way to complement our literature review in determining the explanatory variable that associate with the phenomenon of interest. Empirical results on the impact of the export-oriented MNCs on wage in several researches were varied. The source of the variations could be the choice of measurements, study design, characteristics of the sample or number of observation. Through meta-analysis, we attempt to identify the source of variation and its impact on the regression model. These results, however, solely develop support for hypothesis that merit further testing. Therefore, it is necessary to conduct an econometrics analysis to ensure the robustness of the results. The methodology is to estimate a fully specified inter-industry wage growth regression which incorporates foreign ownership with separately export orientation.

There is a potential endogeneity problem in estimating equation, with exporting history (EXPORTt-1) being the endogenous variable. EXPORTt-1 is likely endogenous because there is strong persistence in the variable correlates with εit. Previous studies (e.g. Roberts and Tybout 1997; Campa 2004) found a very strong effect from the previous years’ exporting status on the current decision to export. Another important statistical issue regarding estimation of model is “sample censoring”. The dependent variables, AVGWAGE, can only be calculated for pairs of plants that remained as exporters in t and t-1. Therefore, the sample is likely to be truncated, particularly during the peak of the crisis, and estimating equation on the selected sample may lead to biased estimates.

The main data set for the quantitative analysis is the annual manufacturing surveys of medium- and large-scale establishments (Statistik Industry, or SI), panel data, from 1980 to 2004.

As major problems of panel design, the survey obtains an initial representative sample of respondents at set intervals over an extended time. As such, panel establishments may drop out of the survey, and the repeated observations may influence their behavior.

This study adopts firm-level survey, and in-depth interview, as the data collection strategy to answer research-policy questions. Special attention is given to the factors that cannot be represented by some variables extracted from the secondary data. The interview, in particular, asked the following topics: industrial upgrading and transfer mechanism to workers; and the problems that company faces in industrial adjustments related to wage restraining effects. Determining the counterfactual, however, may be at the heart of core problem in this project design. It is, however, quite tricky to net out the impact of global sourcing activities of MNCs from the counterfactual conditions that are likely to be influenced by contemporaneous events, selection bias, and contamination. Hence the findings should be treated carefully, since they can offer no grounds for establishing reliability of generality of findings

No comments: