What is the Shape of the Multinationality-Performance Relationship?

Multinational Business Review, Spring 2004 by Thomas, Douglas E, Eden, Lorraine

ABSTRACT: Previous theoretical explanations and empirical analyses of the multinationality-performance relationship have produced mixed arguments and results. Linear and inverted U-shaped relationships have been theorized and confirmed empirically. Recent research has theorized that there is a three-stage, sigmoid relationship between multinationality and performance. We contribute to the debate by showing that the impact of multinationality depends on the time dimension incorporated in the performance measure; that is, the net benefits from multinationality are likely to be higher in the longer term. The results from our sample of US manufacturing multinationals indicate that there is a three-stage, sigmoid multinationality-performance relationship.

INTRODUCTION

Over the past three decades, international business and management scholars have sought to understand how international diversification through foreign direct investment (FDI) affects firm performance. However, previous research on the multinationality-performance relationship has produced mixed arguments and results (e.g., Buhner, 1987; Kotabe, Srinivasan, & Aulakh, 2002; Tallman & Li, 1996). The most recent research has focused on explaining the shape of the relationship: linear, U- or inverted-U, or sigmoid; however, theoretical and empirical gaps continue to bedevil researchers.

This paper provides a fresh perspective on the multinationality-performance debate. We first review recent literature on the debate, finding five different hypotheses about the shape of the multinationality-performance relationship. We argue that there are three partial explanations for the confusing results in the literature (Annavarjula & Beldona, 2000). First, the term 'multinationality' has different meanings. Conflicting results are partly due to different understandings and proxy measures. Second, the theoretical benefits and costs of multinationality to an individual firm, and how they are reflected in firm performance, are not well understood. Third, the temporal dimension of the multinationality-performance has generally been ignored. For example, the differences between short-run and long-run performance have not been fully explored.

In the following sections, we develop a theoretical explanation of the multinationality construct, arguing it has three dimensions: foreign production presence, foreign market penetration and country scope. Next, we provide a clear outline of the benefits and costs of multinationality to firms in general. Finally, we argue that the impact of multinationality depends on the temporal dimension incorporated in the performance measure, that is, the net benefits from multinationality are likely to be higher in the longer term. We test our model on a sample of US manufacturing multinationals over the 1990-94 period, and present the somewhat surprising results. Finally, we offer some conclusions and guidance for future research.

LITERATURE REVIEW

The literature provides several theoretical explanations of the nature of the multinationality-performance relationship. Five general models - positive and linear, positive but with diminishing returns, inverted-U shaped, and sigmoid - explaining the relationship have been presented by previous researchers. Empirically, these models have been tested to varying degrees. We review these models briefly below.

Positive and Linear Model

One stream of research on international diversification has hypothesized and found support for a positive, linear relationship (e.g., Grant, 1987; Grant, Jammine, & Thomas, 1988; Man, Lee, & Suk, 1998; Kim, Hwang, & Burgers, 1989; Tallman & Li, 1996). The theory and empirical results indicate that as firms expand internationally, there is a positive, linear impact on firm performance. More recent research finds that this positive, linear relationship is moderated by other factors (e.g., the firm's R&D and marketing capabilities (Kotabe et al., 2002)).

Positive But Diminishing Returns Model

From this perspective, multinationality has an initial positive impact on performance; however, over time, diminishing marginal returns set in. The slope of the curve is still positive (i.e., multinationality continues to have a positive impact on firm performance) but at a decreased rate (i.e., the benefits are not as great as they were initially) (Gomes & Ramaswamy, 1999). Very little empirical research has specifically tested this relationship.

U-shaped Relationship

Other researchers have theorized and found evidence for a U-Shaped relationship between multinationality and firm performance (Ruigrok & Wagner, 2003). Using organizational learning theory, Ruigrok and Warner (2003) find that firms initially experience negative performance when expanding internationally. Over time, however, as firms learn from their international experience, their performance becomes positive. Very little empirical research has examined this relationship.