Industrial concentration and price-cost margin of the Indonesian food and beverages sector

Abstract

This article investigates trends in industrial concentration and its relationship with the price-cost margin in 54 subsectors of the Indonesian food and beverages sector in the period 1995 to 2006. This study uses firm-level survey data provided by the Indonesian Bureau of Central Statistics (BPS), classified at the five-digit International Standard Industrial Classification (ISIC) Level. The results show a significant increase in industrial concentration in 1995 to 1999, which coincided with the period of the economic crisis in Indonesia. After 1999, the industrial concentration exhibits a slightly decreasing long-term trend. Furthermore, the industrial concentration for all subsectors tends to converge to the same value in the long run. Additionally, results show that higher industrial concentration yields a higher price-cost margin. Finally, the introduction of the competition law in 1999 has slightly lowered industrial concentration and the price-cost margin.

Keywords:

• industrial concentration
• price-cost margin
• competition law
• food and beverages sector

JEL Classification::

• D43
• L16
• K21
• C33
• L11

Acknowledgements

The Authors are indebted to the Directorate General of Higher Education, Republic of Indonesia for the funding and an anonymous referee as well as Ronald Coase Institute Workshop for helpful comments and suggestions.

Notes

¹ We define industrial concentration as in Bain (1951).

² Considering stocks, this article calculates industrial concentration based on sales data because sales seem to explain more about the market share than the output.

³ CR4 is commonly used to classify the market into some categories of oligopoly (Shepherd, 1999), but it cannot capture the distribution of the market share for all firms in the market. However, HHI can capture the distribution of the firm's market share in a market, but it is rather difficult to classify the oligopoly categories from the HHI. In spite of this, Besanko et al. (2004) gave a classification of market structure based on HHI. Furthermore, Liebenberg and Kamerschen (2008) also discusses the importance of using both HHI and concentration ratio.

⁴ BPS defines medium-sized firms as those firms employing more than 20 workers.

⁵ Domowitz et al's (1988) article which has a connection to this article has been challenged by Kamerschen and Park (1993a, b). In the same year, Domowitz et al. (1993) replied that they did not view the Comment by Kamerschen and Park (1993a, b) as a substantive criticism of Domowitz et al. (1988) paper, and the issue remains open.

⁶ Actual codes used are not the ISIC ones, but comparable, coming from Klasifikasi Baku Lapangan Usaha Indonesia (KBLI).

⁷ Although the data on demand growth and COR are too heterogeneous, this may not undermine the accuracy of the estimation given that these two variables appear only as instruments in the model.

⁸ The lowest demand growth is found on subsector 15 316 (Nuts), which dropped about −94.21% in 2002 compared to 2001. According to the data survey, this may be caused by firms exiting the market, as well as the decrease in the demand for this product in 2002. The highest demand growth was experienced by subsector 15 134 (dried fruit and vegetables), which increased about 696% in 1998 compared to 1997. The significant increase of the demand growth in 1998 is caused by the considerable increase of sales in 1998, after the sales drop caused by the crisis, starting in mid-1997. The high variation of the COR seems obvious because of the different capitals used in each subsector. The subsector with the highest COR is 15 143 (coconut oil), with a value of 3.937, whereas the subsector with the lowest COR is 15 510 (liquors), with a COR of 5.27 × 10⁻⁶.

⁹ This model is different from the model used by Bird (1999), who also investigated the trend of the industrial concentration using the models of Khemani (1980) and Brozen (1982).

¹⁰ Derivation of this model can be found in Barro and Sala-I-Martin (1992) and Sala-I-Martin (1996).

¹¹ The derivation of the relationship between industrial concentration and PCM can be found in Cowling and Waterson (1976) and Clarke and Davies (1982).

¹² Some subsectors, such as 15 145 and 15 149, are not reported in 2006.

¹³ In the CR4 model, the lowest and the highest fixed-effects parameters are −0.248 and 0.333, respectively. Almost similar to the CR4 model, in the HHI model, the lowest and the highest fixed-effects parameters are −0.253 and 0.200, respectively.

¹⁴ Based on the parameters of the fixed effects of the model with CR4, some subsectors, such as 15 492 (ice cubes and other ice products), 15 421 (granulated sugar), 15 141 (crude cooking oil), 15 540 (soft drinks) and 15 496 (crackers and other chips), have the highest PCM in the period of estimation. However, some subsectors, such as 15 329 (other starch nec), 15 321 (wheat flour), 15 530 (beer), 15 134 (dried food and vegetables) and 15 312 (other grain milling), have the lowest PCM in the period of estimation. Also, based on the parameters of the fixed effects of the model with HHI, some subsectors, such as 15 421 (granulated sugar), 15 510 (liquors), 15 540 (soft drinks), 15 139 (other fruit and vegetables) and 15 141 (crude cooking oil), have the highest PCM in the period of estimation. However, some subsectors, such as 15 329 (other starch nec), 15 134 (dried food, vegetables), 15 125 (processed fish), 15 321 (wheat flour) and 15 423 (other sugars), have the lowest PCM in the period of estimation.

¹⁵ We slightly abuse notation here. Although we use the partial derivative sign, we evaluate the effect of the competition law as the dummy variable changes from zero to one.