The impact of OFDI in global cities on innovation by Indian multinationals

ABSTRACT

We investigate the impact of Outward Foreign Direct Investment (OFDI) on the innovation performance of Indian multinationals by analysing their choice of entry mode (cross-border acquisitions versus greenfield investments) and their international location decisions (i.e. choice to locate in a global city). We rely on an augmented fractional logit estimator for 170 foreign investments in high- and medium-high tech manufacturing sectors in the period 2003 to 2011 and find that: (i) compared to greenfield investments, acquisitions generate more technological opportunities; (ii) location in a global city has a negative impact on the Indian companies’ innovation performance; (iii) the positive effect of acquisition decreases if the investment is focused on a global city. These findings should be informative for both firms and government authorities involved in developing going-global strategies.

KEYWORDS:

• Outward foreign direct investment
• entry mode
• Global cities
• innovation
• India

JEL Codes:

• O1
• O3
• F23

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ The definition is based on three main characteristics: high level of interconnectedness to local and global markets; a socio-cultural cosmopolitan environment; and (iii) a large endowment of advanced producer services. For more details see Section 3.1

² The Eurostat-OECD classification based on NACE Rev. 2 is available at http://ec.europa.eu/eurostat

³ We exclude greenfield investments in Sales or Retail activities (corresponding to 37.6% of the total) since these are less likely to be driven by asset-seeking motivations (Amendolagine, Cozza, and Rabellotti 2015)

⁴ Minority acquisitions are not included for two main reasons: first because they may have limited impact on the acquiring firm’s performance, thus introducing spurious effects on the main relation (for a similar approach see Woodcock, Beamish, and Makino 1994) and second because comparing the effect of greenfield investments and acquisitions on innovation performance, minority acquisitions do not offer an appropriate counterpart. Additionally, in line with Shaver (1998) we do not include joint ventures ‘because of the difficulty in comparing performance between this mode and acquisitions or greenfield entries’ (575).

⁵ High income countries are defined according to the World Bank classification.

⁶ We consider patent application date rather patent grant date since the former better captures the timing of the firm’s knowledge generation (Quintana-Garcia and Benavides-Velasco 2008; Stiebale 2016).

⁷ Top global cities are defined based on Beaverstock, Smith, and Taylor (1999); Table 2 provides the list of top global cities attracting Indian investments.

⁸ This variable can be considered to proxy also for firm size (Cantwell and Santangelo 2000).

⁹ Orbis classes very large companies are those satisfying at least one of the following criteria: operating revenue of more than/equal to €100 million; total assets of more than/equal to €200 million; 1,000 or more employees; and listed company; and defines large companies as those with an operating revenue of more than/equal to €10 million; total assets of more than/equal to €20 million; and more than 150 employees.

¹⁰ Based on the Eurostat-OECD classification, the high-tech industries include 21 Pharmaceuticals and 26 Computer & Electronics (2-digit NACE Rev. 2).

¹¹ We count the number of subsidiaries established by Indian MNEs in the same destination country in the year of the investment.

¹² 0.055 is equal to 6.5% of the average number of patents granted by investors after the deal (see Table 5 Column 1).

¹³ Following Desyllas et al. (2018), we test whether the structural moderating effect of a global city destination is driven by the nonlinearity of our model (Berry, DeMeritt, and Esarey 2010; Bowen 2012). The results are statistically insignificant, which suggests that the total moderating effects are due to the inclusion of the interaction term.

¹⁴ As a robustness check, we ran a model including the major world cities identified by Institute for Urban Studies of the Mori Memorial Foundation Global Power City Index (http://www.mori-m-foundation.or.jp/english/ius2/gpci2/). The results (available upon request) were largely consistent with our main results.

¹⁵ The moderating effects of global city reported in Table 7 are calculated from Table 5 Columns 4 and 5.

¹⁶ In a robustness check, we included investor NACE 2-digit fixed effects in the estimates, which led to a 19% smaller sample. The results (available upon request) were largely consistent with the main results.

¹⁷ Further information on the ‘Make in India’ initiative is available at https://www.makeinindia.com.