Two key elements of most popular models of Korean development are the active role of the government and the prevalence of the conglomerate form of industrial organization. On careful scrutiny, however, these elements turn out to be neither necessary nor sufficient conditions for development. Furthermore, both the role of government and that of conglomerates have changed over time as the Korean economy transformed itself from a poor agricultural economy of the 1960s to a prospering industrial economy in the 1990s. Both the government and conglomerates are certain to change further as the Korean economy goes through a structural change, which is probably occurring at the present time and expected to continue well into in the next decade. The point is that the true picture of the Korean economy is a dynamic one, and therefore one should pay attention to the evolution of different players in the game if one is to understand the dynamics of the Korean economy.

In explaining the dominant position of the conglomerate system in the Korean economy, many people have focused on the socio-cultural aspect of the Korean society. The logic of this methodology runs something like this: "The prevalence of conglomerates is a phenomenon that can be found in many countries, advanced and developing. So the unique role Korean conglomerates or "Chaebols" played in Korean development must be due to their uniquely Korean characteristics. The most important of these is the family ownership and management. Therefore, this is the key to explaining the success of the Chaebol system."

I find the socio-cultural approach not so enlightening because it does not help me much in thinking about the future of the Chaebol system. What I find useful is an approach which is partly based on neo-classical economics. The logic of this alternative approach can be outlined as follows: "The prevalence and dominance of the conglomerate system must be due to some kind of efficiency associated with it. The distinguishing characteristic of the conglomerate system is the combination of unrelated businesses. Therefore, the reason for the success of the conglomerate system in Korea should be found in the environment that makes the combination of unrelated businesses efficient.

In neo-classical economics, the cost advantage of a firm which produces multiple products is called economies of scope. For instance, if a firm which is producing cars also produces trucks, its cost of producing trucks would be lower than that of a firm which is not producing cars. A Korean conglomerate typically produces products as diverse as ice cream and computers. It is difficult to believe that there are technological economies of scope between ice cream and computers. Observing, however, that conglomerates keep producing such unrelated products as ice cream and computers, one cannot but think that there must be some kind of cost efficiency in producing both of them. I submit that the cost savings of Chaebols are primarily due to economies of scope in lobbying.

If one wants to do any worthwhile business in Korea, one first needs to make the acquaintance of officials of the relevant government agencies, officers of banks, congressmen, and officers of law enforcement agencies. If one wants to do a really profitable business, one has to be connected to the power center around presidency because business opportunities which generate big rents are distributed at the top and loans of big amounts have to be approved at the top. Once the connection is established, it is easier to do business in any field. Thus as one adds more businesses, be it a construction company or a steel company, the per-company cost of maintaining the connection goes down. That is, there exist economies of scope in lobbying.

All this is perhaps well understood by Koreans. The primary reason why I propose the hypothesis that economies of scope in lobbying underlie the prevalence of conglomerates is that it is a testable hypothesis. For instance, if this theory is right, the share of conglomerates in a country will be greater as the importance of lobbying is greater. More importantly, assuming that Korean society becomes a more transparent one leaving little room for lobbying as a key element of success, one can make the prediction that the cost advantage of the Chaebol system will decline and therefore the Chaebol system will not stay as the dominant form of industrial organization in Korea.

The demise of the Chaebol system can take many different routes. Some marginal Chaebols will dissolve because they will no longer be profitable. Big Chaebols will regroup into business groups specializing in closely related businesses. The predicted reorganization of the Chaebol system will incur substantial social costs during the transition, for the reorganization can not be done without bankruptcies and mergers. Paradoxically, the government sector and some Chaebols may become larger during the transitional period because someone has to take over the assets of failing Chaebols at least temporarily.