Analysing location choices of small and large enterprises of electronics-manufacturing industry in Korea

ABSTRACT

This paper measures regional economic and spatial effects on locations of small and large enterprises, by using binary logit models. The econometric results confirm that the location of small firm depends on the regional labour force availability and the population growth of its neighbouring regions. In addition, the large enterprise tends to agglomerate into regions to be less-regulated or to be near to high-speed railroad stations. Finally, industrial clusters of the small and large enterprises would be likely to be geographically close each other in order to share intermediate inputs as a co-agglomeration pattern.

KEYWORDS:

• Electronic manufacturing sector
• industrial location
• large firm
• small and medium sized enterprise
• regional economics

JEL CLASSIFICATION:

• R30
• L25

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The LE is defined as a firm that has more than 300 employees.

² Korea comprises 226 cities and counties, 65 of which are contained in the Seoul Metropolitan Area (SMA).

³ Local Moran’s I is as shown in $I_i=\frac{z_i-\bar{z}}{{\sigma }^2}\sum _j^n\left[w_{ij}\left(z_j-\bar{z}\right)\right]$ (Anselin 1995). where $\bar{z}$ is the average number of employees in the electronics-manufacturing sector; $z_i$ is the number of the employees at location $i$; $z_j$ is the number of employees at all other locations; ${\sigma }^2$ is the variance in the number of employees; $w_{ij}$ is the spatial weight matrix calculated as the inverse of the distance $d_{ij}$ between regions $i$ and $j$; $n$ is the number of areal units.

⁴ The binary logit model is used to predict probabilities of occurrence of an event using a logistic function, relaxing key assumptions of normality and homoscedasticity of error terms. The dependent variable has qualitative outcomes from a mutually exclusive and collectively finite set of discrete alternatives, while the independent variables are linearly associated to the log odds (Judge et al. 1982; Horowitz and Savin 2001; Cramer 2003; Garrow 2010). The model has been widely applied to decision issues such as households’ tenure choice, preferences of traveller’s personalized travel behaviour, farmers’ agricultural adaption to climate change and female labour force participation.

⁵ In the pre-test stage, we found there were no impacts of highway variables on clusters of the SMEs and the LEs.

⁶ planning acts (RPA) Korean government legislated Seoul Metropolitan Area Readjustment Planning Acts in 1982, which targeted to disperse industries and population concentrated around Seoul, the capital city.