Testing and implementing a new approach to estimating interregional output multipliers using input–output data for South Korean regions

ABSTRACT

Flegg's location quotient (FLQ) is a useful tool for estimating intraregional output multipliers. This paper uses it as one component when estimating interregional multipliers. Using statistical information criteria and official data for 16 South Korean regions, it is found that the best approach is to combine the FLQ with a simple trade model. The paper explains how the proposed procedure can be implemented for both multiple and individual regions, and also how a region-specific value for the unknown parameter $\delta$ in the FLQ formula can be determined. Finally, an illustrative case study of one of the regions is carried out.

KEYWORDS:

• interregional multipliers
• Flegg's location quotient (FLQ)
• gravity model
• information criteria

ACKNOWLEDGEMENTS

The authors thank the editor and the three anonymous referees for numerous very helpful comments and suggestions.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1 For a more detailed discussion of FLQ's properties, see, for example, Bonfiglio and Chelli (2008) and Flegg and Webber (1997, 2000). We have chosen not to discuss the augmented FLQ (AFLQ), which takes regional specialization into account, but tends to produce similar results to the FLQ (Bonfiglio & Chelli, 2008; Flegg & Webber, 2000; Flegg et al., 2016; Kowalewksi, 2015).

2 Zhao and Choi (2015) also investigated this topic using data for the same two regions; for a critique of their study, see Flegg and Tohmo (2019).

3 The output and employment data in Table 1 exhibit a strong linear relationship (r = 0.921). However, this aggregate relationship masks much variability at the sectoral level.

4 The correlation analysis mirrors that in Flegg and Tohmo (2019, p. 11).

5 As expected, the intraregional share of inputs (Table 1) and the intraregional multiplier (Table 2) are strongly correlated (r = 0.881). North Gyeongsang is an exception: its intraregional share is close to the mean for all regions, yet it has the highest multiplier.

6 These numbers differ slightly from the corresponding values in Table 2 owing to differences in the method of calculation.

7 Boero, Edwards, and Rivera (2017) pursue a very different and much more complex approach, using US county-level data on supply and demand, along with measures of transport costs, to estimate trade flows. Although this interesting new procedure seems to yield reasonably accurate results, the authors note (p. 12) that it is computationally burdensome, especially where the focus is on a single county.

8 For an interesting and ambitious application of gravity models using Scottish data, and a comparison with other approaches, see Riddington, Gibson, and Anderson (2006). For some comments on this study, see Flegg and Tohmo (2013, pp. 707−8).