Diaspora-owned firms and social responsibility

ABSTRACT

A causal relationship between diaspora populations and bilateral foreign direct investment has been established empirically, but the question of which elements of diaspora difference are responsible for this relationship, and what this implies for development, remains unanswered. A growing literature in economic sociology and business suggests that diaspora investors are motivated by patriotism and other social and emotional factors, endowing them with unique potential as a force for international development. This literature argues that diaspora-owned firms are more socially responsible than other foreign firms, and engage in a range of economic development-promoting behaviors when investing in the homeland: hiring more local labor, paying higher wages, and making more contributions to charity. I argue that diaspora-owned firms enjoy competitive advantages in the homeland based on access and attention to information. I evaluate these theories at the firm level, using data from an original survey of 174 foreign-owned firms in the post-conflict country of Georgia. Across a range of self-reported behaviors and priorities, I find no evidence that diaspora-owned firms are more likely to engage in a specific set of socially responsible, pro-development behaviors than are other foreign firms, and some evidence that they are less likely to do so. I argue that diaspora investors are uniquely capable, but not uniquely philanthropic, when doing business in their homelands.

KEYWORDS:

• Diaspora
• foreign direct investment
• international business
• social responsibility
• migrants
• development

ACKNOWLEDGMENTS

Special thanks to the Rohr Chair of International Relations for fieldwork funding, to the Georgian Foundation for Strategic and International Studies and the Caucasus Research Resources Center for in-country support, and to the wonderful Capital and Conflict: Georgia team: Maya Baramidze, Amiran Chanchibadze, Anna Gogokhia, David Kokiashvili, Alexandre Kukhianidze, Anastasia Laitadze, Giorgi Mekerishvili, Maia Mestvirishvili, Anna Sekowska-Livny, and Giorgi Tarkhan-Mouravi. Thanks also to Lawrence Broz, Miles Kahler, Quan Li, Edmund Malesky, Craig McIntosh, and Kaare Strom for advice and comments and on drafts of this paper and to Liesl Riddle for her overall mentorship and guidance on the project. Mistakes and weaknesses that remain are entirely my own.

Notes

¹The African Diaspora Marketplace project launched by USAID in 2009 is one example. See also Foreign Service Institute (2010).

²See de Haan (1999) and Faist (2008) for reviews of this literature. Bakewell (2008) also documents a strong anti-migration bias among Western governments and development NGOs.

³For example, Chirwa (1997) had noted the positive effects of labor emigration on rural sending communities in Malawi, and the negative effects of the forced repatriation of those migrants.

⁴Some excellent, primarily qualitative, scholarly work on the development impacts of remittances had preceded this, such as Oberai and Singh (1980), but it was not until after the World Bank began to collect data on the volume of these flows globally that quantitative work on the topic also took off.

⁵The work of Michael Clemens (2011) and others has also highlighted the positive effects of emigration on migrants themselves, pointing out that a singular focus on the effects of emigration on the sending economy ignores the benefits to the very agents creating the phenomenon.

⁶Cited in Carroll (1999).

⁷For example, normative stakeholder theory argues that communities are among the stakeholders to whom firms are accountable, and that the interests of the broader community must be taken into account even when they directly conflict with the interests of shareowners (Freeman, 1984; Donaldson and Preston, 1995).

⁸Others, such as Jia and Wang (2012), look at which firms engage in philanthropy.

⁹This measure includes firm's personnel practices and community involvement, among others.

¹⁰See Chatterji et al. (2009) for an analysis of perceptions-based measures of social responsibility.

¹¹The best-known such measure, from KLD Analytics, covers a sample of the world's largest MNCs, most of them from the S&P 500.

¹²See Blowfield and Frynas (2005) for a discussion of the role of CSR in development.

¹³For a range of definitions, see Garriga and Melé (2004).

¹⁴The literature on CSR and development stresses the need for firms to respond to the specific needs of local stakeholders, which vary greatly from community to community (see Newell and Frynas, 2007; Moon, 2007)

¹⁵For examples of these arguments, see Debass and Ardovino (2009) and Terrazas (2010).

¹⁶See Seawright and Gerring (2008) for a discussion of the use of ‘extreme cases’ and ‘typical cases’ in hypothesis testing.

¹⁷Because of a paucity of cross-national data on diaspora investment, it is difficult to assess the volume of diaspora investment in a given country ex ante.

¹⁸Collier and Hoeffler (2004) find that diaspora populations in rich countries increase the risk of civil conflict, presumably because they represent a potential source of rebel financing.

¹⁹This relationship is most likely to be strained in the case of firms based in Russia. Neither dropping Russian firms from the analysis nor including a dummy variable for Russian origin changes the substantive results.

²¹The survey was taken through three rounds of reverse translation and refinement to ensure equivalence between the English and Georgian versions.

²²Some firms on the Ministry of Finance list also turned out to be registered before the year 2000, but had been re-registered after that date.

²³Some firms on the Ministry of Finance list also turned out to be registered before the year 2000, but had been re-registered after that date.

²⁴I treat any firm with at least one diaspora owner as ‘diaspora-owned’. However, all the empirical results that follow are robust to treating only majority-diaspora-owned firms as diaspora-owned.

²⁵All firms in the sample are privately owned, i.e. none are government-owned or publicly traded.

²⁶Two of these supplemental firms were identified by the State Office for the Diaspora; five more were located via snowball sampling. To create the snowball supplement, enumerators asked respondents at the end of the interviews if they knew of any diaspora-owned firms that we might be able to contact.

²⁷Not all firms have home countries listed on the original sampling lists. For those firms that responded to the survey but did not have a home country listed, I treat the location of the firm headquarters as the home country. Some respondents did not report a headquarters location outside of Georgia – the home country of those firms remains as missing data.

²⁸In practice, the West refers to Western Europe, the US, Canada and Australia.

²⁹Seventy per cent of firms in the sample have 50 employees or less. The median number of employees is 12.

³⁰I also checked the balance in a composite of real estate and construction – the balance here is quite even. I checked because these sectors were particularly hard-hit during the downturn.

³¹It is worth noting that very few firms in the sample entered after the 2008 conflict with Russia: of the 18 that entered during this period, 5 were diaspora-owned and 12 non-diaspora-owned (1 unknown). Most of the firms in the sample entered Georgia as Greenfield investments, rather than mergers, partnerships, or acquisitions of Georgian firms: 76 per cent of diaspora-owned and 78 per cent of non-diaspora-owned firms.

³²This is implemented using the ice and mim commands in Stata 10. I create 10 imputed datasets for analysis.

³³The coefficients of interest are similar in analyses using list-wise deletion.

³⁴These percentages refer only to the 161 firms that completed an owners/managers survey. Those firms where only a front-desk survey was completed are omitted from analysis (and imputation) because none of the social responsibility questions are included on the front desk survey.

³⁵King et al. 2004.

³⁶The questions referred to in this section were developed collaboratively with Professor Maia Mestvirishvili of Tbilisi State University.

³⁷The Cronbach's alpha for these two variables is 0.8.

³⁸Substantively, a respondent from a diaspora-owned firm is 2.8 times as likely as an otherwise identical non-diaspora-owned firm to respond with a higher category response across any given cut-point on a seven point scale from ‘strongly disagree’ to ‘strongly agree’.

³⁹The odds ratio for the diaspora dummy in Model 2 is 1.23.

⁴⁰Analysis was also run including an interaction term between diaspora identify and firm size, but no statistically significant results were found.

⁴¹An exciting first step in this research agenda was taken by Vaaler (2012), who finds that remittances from geographically concentrated diasporas have a larger impact on the availability of venture capital in the homeland than do remittances from geographically dispersed diasporas.