Endogenous Transaction Costs and Division of Labor

Download this Paper

Xiaokai Yang
Department of Economics Monash University and
Center for International Development at Harvard University

and Yimin Zhao
Global Transfer Pricing Services, KPMG

Abstract
The paper develops a general equilibrium model with alternating offer bargaining and endogenous specialization to investigate interplays between network effects of division of labor and endogenous transaction costs caused by strategic interactions. It shows that endogenous transaction costs may have adverse effects on the development of division of labor and therefore welfare. The endogenous transaction cost may be eliminated in a super game where players care about their reputation.

1. Introduction
In the emerging literature of general equilibrium models of high development economics (see Fujita and Krugman, 1995 and Krugman and Venables, 1996) and in the growing literature of endogenous specialization (see Yang and Ng, 1998 for a recent survey and references there), effects of transaction costs on the equilibrium network size of division of labor, the extent of the market, and productivity are explored by formulating the trade off between increasing returns and transaction costs. In these models, however, transaction costs are exogenous in the sense that they are not caused by interest conflicts between players in economic arena.

North and Thomas (1970) distinguish the endogenous transaction costs, caused by moral hazard, adverse selection, and other types of opportunism, from exogenous transaction costs. Williamson (1975, 1985) also draw the distinction between the endogenous transaction costs caused by opportunism (holding up, cheating, and not credible commitment) and tangible exogenous transaction costs. At the same time, an extensive literature of formal models of endogenous transaction costs has been developed. . Game theory has been extensively used in the literature to investigate endogenous transaction costs caused by strategic interactions. However, none of these models can analyse the effects of endogenous transaction costs on the development of division of labor that classic economists (such as Adam Smith) focused on. The current paper intends to fill the gap by developing a model with endogenous structure of division of labor and endogenous transaction costs.

As elabrated by Smith (1776), the division of labor and specialization are the mainspring of economic progress, the division of labor is dependent on the extent of the market (chapter 3 of book I), and the extent of the market is determined by transaction costs (pp. 31-32). Allyn Young (1928) also noted that division of labor has positive network effects since "not only division of labor is dependent on the extent of the market, but also the extent of the market is determined by the level of division of labor." Trading between different specialists, however, involves transaction costs which in turn deters the development of division of labor. The transaction costs include not only tangible exogenous transportation costs, but also endogenous transaction costs caused by strategic interactions. Since network effect of division of labor is a general equilibrium phenomenon, the level of endogenous and exogenous transaction cost, each player's decision in choosing a pattern of specialization, and the network size of division of labor are interdependent. If all players choose autarky, there is no transactions and related cost, while players' decisions of the level of specialization are dependent on the equilibrium level of endogenous and exogenous transaction costs. In addition, all players' participation decisions of the division of labor are dependent on the equilibrium network size of division of labor, while the equilibrium network size of division of labor and related trade are dependent on all players' decisions of participation. Hence, we need a general equilibrium model to investigate the interactions among all of the interdependent variables.

The present paper specifies a general equilibrium model of alternating offer bargaining game (Rubinstein 1982, 1985) in which players compete for a greater share of the gains from division of labor to study the impact of endogenous transaction costs, caused by opportunism, on the equilibrium level of division of labor. The players' choices of patterns of specialization are endogenised in the model. The main finding is that the players' strategic competition for the first mover advantage results in endogenous transaction costs, which generate Pareto inefficient level of division of labor. If the game is repeatedly played and both players care about their reputation, it can be shown that cooperative outcome may emerge from noncooperative behaviour, where endogenous transaction costs can be eliminated.

The paper is organised as follows. Section 2 specifies the basic model. Section 3 solves for the Nash bargaining equilibrium as the benchmark case, where the equilibrium level of division of labor is Pareto efficient. Section 4 analyses the endogenous transaction costs associated with Pareto inefficient level of division of labor by introducing the game of competing for a greater share of the gains from division of labor, of which the alternating offer bargaining game is a subgame. Section 5 examines how the consideration of reputation can partly eliminate the endogenous transaction costs in a supergame. Section 6 concludes the paper.

Inframarginal Economics Society¯¸ www.inframarginal.com