Characteristics of Indonesia's large and medium scale manufacturing industries: An exploratory analysis

The manufacturing sector has contributed significantly to the Indonesian economy. This sector contributed 27 per cent on average to Indonesia's GDP between 2000 and 2015, with more than 40 per cent of its value-added was contributed by large and medium scale industries. Despite the important roles of large and medium scale industries in Indonesia's economy, this sector has three major problematic characteristics that are explored by three different research papers in this thesis. These characteristics are inconsistent growth either for output or labour productivity growth, a steady increase in wage inequality, and relatively low labour absorption and labour mobility. The first paper in this thesis aims to observe how productivity growth measured by Total Factor Productivity (TFP) growth plays a role in the production process of Indonesia's manufacturing sector. This paper estimates TFP growth and its decompositions based on varying parameter stochastic frontier analysis (VSFA) framework as this approach enables me to consider firm heterogeneity explicitly. By using datasets from Indonesia's Yearly Large and Medium Manufacturing Industries Survey over the period 2002-2014, VSFA reveals that a constant parameter stochastic frontier (SFA) overestimated mean technical efficiency (TE) and the TE rank under VSFA is more consistent than under SFA. Hence, it is logical to assume that firms should not have a constant production function response, so TFP growth is measured based on the results of VSFA. The mean TFP growth was estimated at 4.3 per cent and was mostly decomposed by technological progress. Moreover, it was estimated that labour efficiency in this sector was relatively low, at 46 per cent. This implies that technological progress had not been absorbed well by workers. The second research paper analyses how wage inequality affects firm productivity. By implementing several econometric approaches, which are panel fixed effects, dynamic panel data estimation - Generalised Method of Moments (GMM) approach and instrumental variable estimations - two-stage least squares regression (2SLS), it is found that wage inequality significantly affects firm productivity in an inverted U-shaped relationship. This implies that when wage inequality in Indonesia's large and medium scale industries from 2000 to 2015 was relatively low, it increased firm productivity. However, if wage disparity was more than the threshold, it reduced productivity. This means that the findings support the argument of the 'tournament' model (Lazear and Rosen, 1981) rather than the 'fairness' model (Akerlof and Yellen, 1988). To increase firm productivity, relatively low wage inequality is needed to motivate workers. The last paper examines how manufacturing jobs and labour mobility, which are measured by geographical mobility and job mobility, affect wage inequality. The findings show that manufacturing jobs and occupational mobility have an inverted U-shaped relationship with wage inequality. This implies that a relatively low level of job absorption and occupational mobility among workers will increase wage inequality. However, when this is beyond the threshold, it will reduce wage disparity. On the other hand, spatial mobility significantly affects wage inequality in a U-shaped relationship. This means that relatively low geographical mobility in the labour supply will reduce inequality. Once it is more than the threshold, it will increase wage inequality. These findings are robust across many dimensions: different types of wage inequality measurement - conditional and unconditional wage disparity; the use of different techniques such as using OLS, FE, and dynamic panel models with and without lagged independent variables, instrumental variables (IV) technique; and the use of different levels of data - industrial group and regional level data