Abstract
This chapter examines whether and how the productivity of a firm outsourcing to offshore suppliers differs from those not active in cross-border outsourcing based on Japanese firm-level data . We also compare them with firms outsourcing within the same country and with firms sourcing within the firm boundary . Before reporting our empirical findings, we briefly summarize predictions from theoretical models in international economics.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Although Helpman et al. (2004) analyzed the choice between exporting and horizontal FDI to serve a foreign market, the fundamental argument can be applied to our comparison of cross-border outsourcing versus vertical FDI . We have assumed away here, but will discuss the issue of incomplete contract in the Chap. 8. Among excellent surveys, the Coase Lecture by Helpman (2014) overviewed the firm heterogeneity trade theory, and Antràs and Yeaple (2014) reviewed FDI literature, but Antràs and Rossi-Hansberg (2009) placed emphasis on firm boundary .
- 2.
Defever and Tubal (2013) found that FO firms are more productive than FDI firms . However, as their sample is limited to FDI firms , their finding indicates additional entry costs of FO.
- 3.
The profit shown in this diagram is that from cross-border sourcing only. Firms are assumed to source also from domestic suppliers and earn by selling their products.
- 4.
We for the moment assume that the productivity is exogenously given for a firm and firms are self-selected into a different globalization mode (foreign outsourcing or FDI) depending on its productivity level. However, the productivity of a firm may change as a result of its globalization mode choice through such mechanism as learning or reorganization. Although it is difficult to exactly test the causality direction within our limited data, we will discuss productivity dynamics in the final section of this chapter.
- 5.
We treat domestic outsourcing and domestic intra-firm sourcing combined as domestic sourcing , partly because our research focuses on cross-border outsourcing , and also due to the limitation of our datasets.
- 6.
The productivity dispersion is from a part of Table 2 in Katayama et al. (2011).
- 7.
Which material cost data should be used is critically related to the outsourcing concept, as we discussed pertinent data reported by manufacturing census in Chap. 4.
- 8.
As repeatedly discussed in previous literature, OLS estimates of production function are inconsistent because unobserved shocks to a firm distort the firm’s choice of factor inputs. To bypass this problem, instrumental variable estimation is proposed, but we cannot apply it to our research of cross-section data . The econometric problems of using investment or material expenditure based on the assumption about decision timing as an instrument variable are recently pointed out too.
- 9.
We have confirmed that all of our results are not qualitatively affected even if we replace the labor share two-thirds by alternative three-fourths.
- 10.
Kamal et al. (2015, Table 3.14) compared value-added per employees across firms with different engagement in contract manufacturing services.
- 11.
We show below the estimates at 2005, but the results are basically the same in 2000. See Ito et al. (2011a). Capital stock is calculated by the perpetual inventory method with the data on book value of fixed tangible assets and investment.
- 12.
Theoretically, FDI firms , which we should compare with foreign outsourcing firms in the sourcing decision, must be vertical FDI firms, but no direct distinction between horizontal and vertical FDI is possible in the MITI survey , as explained in Chap. 4. We will report the results with a primitive distinction of these two FDI types later.
- 13.
In the MITI survey , no data on capital depreciation are unfortunately collected. For many firms in the survey, the data on capital itself are unavailable either.
- 14.
Tomiura (2007a) also used domestic market share as another proxy for productivity, since it increases the firm’s productivity in most of the standard theoretical models.
- 15.
While our dataset also identifies FDI firms and exporters on a comparable basis, their dataset has an advantage in distinguishing intra-firm sourcing in domestic sourcing .
- 16.
From a different Japanese firm-level data set, Head and Ries (2003) reported that domestic firms (non-FDI non-exporters ) are 19–33% smaller, and firms active both in exporting and FDI are 105–332% larger than non-FDI exporters in size.
- 17.
We set each bin to include approximately equal number of firms. Each bin cannot include exactly the same number of firms, as the firms with exactly the same number of employees are assigned to the same bin. The analysis of exporters by Bernard et al. (2003) assigned plants into 500 bins, but the size of their bins are approximately the same as ours because their sample size is more than twice as large.
- 18.
Four industries with severely limited numbers of firms are not divided into within-industry bins.
- 19.
We merge ordnance into general machinery. The definition of domestic firms is the same as before in this section (firms that are active in none of the three globalization modes).
- 20.
As an example of similar micro-data study, Okubo and Tomiura (2012) analyzed how the industrial relocation policy affects not only the mean productivity but also the shapes of productivity distributions in Japan based on plant-level data from manufacturing census.
- 21.
We allocate the domestic firms in the top and bottom 1% to the top and bottom intervals, respectively. The productivity range between these two thresholds is equally divided into 16 intervals. As a result, the total number of intervals is 18, exactly the same as in Bernard et al. (2003) to facilitate comparisons.
- 22.
Antràs (2016a) correctly recognized this issue in referring to this overlap.
- 23.
Although no import ban or direct quota was introduced by the U.S. or EU side, voluntary export restraint for passenger automobiles to the U.S. was carried out by the Japanese side.
- 24.
Head and Ries (2003) also reported that Japanese firms investing in low-income countries are weakly less productive than those investing in high-income countries from a limited sample of 1070 publicly listed firms.
- 25.
All explanatory variables are in logarithm. Before taking logarithm, we add one to the R&D intensity and to the computer-usage intensity because many firms, especially small firms, conduct no R&D or had no computers in 1998.
- 26.
To keep consistency with the firm heterogeneity trade model, we assume that productivity is firm-specific rather than being determined by host country factors.
- 27.
If a firm invests both in Asia and ROW , a firm is defined to invest in Asia (ROW) if the number of overseas affiliates located in Asia (ROW) exceeds that in ROW (Asia, respectively).
- 28.
As some firms outsource to multiple regions, we introduce the following seven disjoint dummies into our regressions based on the firms’ combinations of their outsourcing destinations: (1) all three regions, (2) both China and North , (3) both China and other Asia, (4) both North and other Asia, (5) North only, (6) China only, and (7) other Asia only.
- 29.
Their instrumental variables assigned to the outsourcing variable include the firm’s age, and purchase/sales ratio, and import/purchase ratio.
- 30.
They find that firms outsourcing to all regions in the world or to both China and ASEAN are significantly more productive.
- 31.
Kohler and Smolka (2014) also reported that the productivity of firms outsourcing within FDI multinational enterprises tend to be more productive than those outsourcing through arm’s-length transaction in Spanish firm-level data .
- 32.
We should be careful in comparing productivity levels over time, as firms substitute away from expensive domestic inputs toward inexpensive foreign inputs by cross-border outsourcing . Unless we consider this price effect, the impact of outsourcing on productivity might be overstated.
- 33.
De Loecker and Warzynski (2012) is a rare example of studies using quantity-based productivity for the comparison of exporters versus non-exporters .
- 34.
As a useful study, though not on outsourcing , De Loecker (2007) identified the causal effect of exporting on productivity at the firm level by analyzing Slovenia’s accession to the EU .
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2018 Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Tomiura, E. (2018). Productivity Premium of Cross-Border Outsourcing Firms. In: Cross-Border Outsourcing and Boundaries of Japanese Firms. Advances in Japanese Business and Economics, vol 18. Springer, Singapore. https://doi.org/10.1007/978-981-13-0035-6_6
Download citation
DOI: https://doi.org/10.1007/978-981-13-0035-6_6
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-13-0034-9
Online ISBN: 978-981-13-0035-6
eBook Packages: Economics and FinanceEconomics and Finance (R0)