A multilevel analysis of the moderating role of trade union strength in the relationship between privatization and corporate gov

INTRODUCTION

For almost three decades now, governments in both developed and developing nations have been privatizing state-owned enterprises (SOEs) (Ramamurti, 2000). According to World Bank estimates, by the year 2004 approximately 6,540; 145,324; and 78, 490 SOEs in Africa, Asia, and South America respectively had been privatized in one form or another (World Bank, 2004). Brazil alone privatized over $80 billion in public assets within the same period (Doh, 2000). Definitions of privatization vary from broad to narrow. Broadly privatization refers to "any measure that increases the role of the private sector in the economy--for example through deregulation, which permits private entry into markets previously reserved for SOEs; economic liberalization, which exposes them to greater competition (e.g., through lower tariffs or fewer restrictions on foreign investment); or institution building, which improves the function of private firms and markets" (Ramamurti, 2000: 526). A narrow definition which is adopted in this study refers to the transfer of some or all of ownership and/or control from State to private entities (DeCastro and Uhlenbruck, 1997; Ramamurti, 2000). Transfer of ownership and control suggests that privatization transactions take various modes (Zahra, Irelnad, Gutierrez, and Hitt, 2000). SOEs may be sold outright (Doh, 2000), cooperatively ventured (Carlin and Aghion, 1996), or vertically integrated (i.e., government-owned professionally-managed) (George and Prabhu, 2000).

Governments privatize SOEs for reasons of economic transformation (e.g., socialism to capitalism) (Kikeri and Nellis, 2005), response to competitive pressures in domestic and international markets (Ramamurti, 2000), attraction of foreign investment (Megginson and Netter, 2001; Nellis, 1998), and technological transfer (Doh, 2000; George and Prahub, 2000). Nonetheless, privatization of SOEs affects groups and individuals that have a stake in the outcomes of the SOEs (i.e., stakeholders) (George and Prabhu, 2000). It alters the risks of stakeholders particularly employees and unions. Generally, majority of employees in developing countries belong to trade unions. So, trade unions are likely to oppose or support privatization if it is perceived to affect their membership (Kikeri, 1998). In addition to changing control and decision-making in post-privatization, unions may view privatized SOEs as lost assets (Doh, 2000). Second, fear of employment lost and reduced earnings which arise from uncertainty about the behavior of the new owner leads to agitations against privatization (Debrah, 2004). Indeed, some trade unions successfully resisted privatization by completely reversing privatization plans (Divano, 1995). Other unions were effective in changing the form of privatization from sale-to-foreigner to sale-to-natives (Kikeri and Nellis, 2005). Collectively these activities of unions suggest that the form of governance adopted for an SOE may vary according to the strength of unions in a country.

As a consequence of the negative effects (Aghion and Carlin, 1996; Mekonnen and Mamman, 2004) and the haphazard approach previously adopted (Debrah, 2004), governments have become concerned about how they privatized SOEs (Ramamurti, 2000). The methods chosen for SOEs therefore vary. From a transaction cost theory perspective (Williamson, 1985), methods adopted for privatized SOEs are referred to as modes of governance (i.e., the coordinating mechanisms adopted to regulate transactions of firms). They range from internal through hybrid to external (Williamson, 1985; 1991). How governments (particularly those in developing countries) privatize SOEs has become important in recent times because of negative effects which are associated with the process rather than the content of privatization. In other words, how governments decide to govern transactions of SOEs determines macroeconomic and microeconomic outcomes of privatization (Megginson and Netter, 2001). It also facilitates effective competition in the global market. What factors therefore determine the modes of governance of SOEs? Conceptual studies suggest factors at country and firm levels (i.e., multilevel factors) (Ramamurti, 2000). However, I am not aware of empirical studies on multilevel determinants of SOE governance or the moderating role of trade union strength.

The purpose of this study therefore was to examine multilevel factors that influence modes of governance of SOEs, and the moderation of trade union strength. I adopted a transaction cost theory (TCT) perspective. So, my approach differs from Ramamurti (2000) in two ways. First, the determinants in this study derive from transaction cost theory. Second, I view trade union strength as a moderator of the relationship between TCT determinants and mode of governance. However, consistent with Ramamurti (2000), I view firm size as a determinant of mode of governance at the firm level.

This study therefore contributes to the literature in three major ways. First, it provides empirically validation of additional factors likely to influence governance modes of SOEs. These TCT determinants supplement our knowledge of other factors identified by previous conceptual studies (see Ramamurti, 2000). Second, the study seems to provide robust support from multiple data sources. This approach seems different from survey studies that tend to obtain data from single sources. Third, this study seems to be the first to empirically test for the effect of trade unions. Previous studies have often suggested but not tested this effect (Kikeri and Nellis, 2005; Nellis, 1998; Ramamurti, 2000).

In response to the empirical gap--lack of empirical studies on Africa, seeming neglect of multilevel effects--and the importance of contingencies (e.g., trade unions) influencing privatization, I examined multilevel determinants of privatization using transaction cost and stakeholder theories. Next, I briefly review TCT as it relates to privatization and labor. Figure 1 shows the model. The model suggests that the mode of governance likely to be adopted for privatized SOEs depends on factors at firm and country levels. In addition, trade union strength is proposed as a moderator of the relationship between country- and firm-level determinants and governance mode. From a multilevel theory perspective (Kozlowski and Klein, 2000), the model suggest intra-level and cross-level interactions. Next, I review the literature on privatization and generate hypotheses from transaction cost theory.

[FIGURE 1 OMITTED]

LITERATURE REVIEW

The literature on privatization which is profuse (see Megginson and Netter, 2001; Academy of Management Review, 2000, vol. 25 number 3) suggests welfare economics theory as the dominant theory of privatization (Megginson and Netter, 2001) even though other theories (agency, property rights, stakeholder, and transaction cost) have been used (Dharwadkar, George, and Brandes, 2000; Doh, 2000; George and Prabhu, 2000). In addition to the country-benefits discussed above, privatization has also influenced performance of some industries (e.g., railways, industry, transportation) (Cook, 1999) and firms in both developed (Megginson and Netter, 2001) and developing (La Porta and Lopez-de-Silanes, 1999; Boubakri and Cosset, 1998) countries. For example, privatization has led to improvement in profitability, efficiency, output, and investments of SOEs (Megginson, Nash, and van Randeborgh, 1994). In Africa, increased capacity utilization through new investments, introduction of technology, and expansion of markets to privatized firms, fiscal and macroeconomic effects have also been found as positive outcomes (Davis et al., 2000; Kikeri and Nellis, 2005). Further, proceeds from privatization have been used to pay national debt which indirectly improves national welfare (Kikeri and Nellis, 2005). For example, in Egypt, the use of privatization proceeds "contributed to the strengthening and stabilization of the economy" (Kikeri and Nellis, 2005:14). Other observed benefits of privatization in Africa include improvement in welfare, income distribution, and employment (Kikeri and Nellis, 2005; Megginson and Netter, 2001).

Nonetheless, privatization has led to changes in the terms and conditions of service, labor force reductions, reduced wage earnings, job insecurity, and suboptimal performance post-privatization (Debra, 2005; Dinavo, 1995; Kikeri, 1998; Nellis, 1998). Economic welfare, productivity gains and employment levels have often not been raised to the expected levels (Kikeri, 1998). National, sector, and firm employment reductions have occurred more frequently following privatization (D'Souza and Megginson, 2000). Indeed, studies show that privatization-promotion institutions such as the World Bank or UNDP often provide funding or financial assistance to countries to mitigate the negative effects of privatization so as to minimize resistance from employees groups or unions (Berthelemy, Kauffmann, Valfort, and Wegner, 2004).

There is a profusion of studies, cross-sectional and case-studies, on developed countries. However, except for case studies, relatively few empirical studies exist on privatization in Africa (Megginson and Netter, 2001). For example, "Jones et al., (1999) show that Nigeria has been one of the most frequent sellers of SOEs, using public share offerings, although they were very small" (Megginson and Netter, 2001). Almost all the studies adopt single-level analysis by focusing on either country- or industry- or firm-level; none has empirically tested multilevel determinants (Ramamurti, 2000). In addition, I am not aware of studies that have examined moderators of privatization even though there is evidence that privatization transactions vary as a function of stakeholder motivations (Cuervo and Villanonga, 2000).

TRANSACTION COST THEORY

Print Get the Mag Weekly Updates