Japan in the Process of Institutional Change

Masahiko AOKI (Professor, Stanford University, and President, RIETI)

Institutions As Shared Beliefs Regarding How the Game is Played

How should Japan's current economic and social condition be characterized from a historical perspective? Those who talk about the "lost decade" might say that the first year of the 21st century was not much different from the last decade.

I do not share that view. My view is that we have entered an era of institutional change, which started around 1993 when the bubble burst and the one-party rule by LDP ended. This process accelerated last year, yet its destination is not yet clear.

It has now been widely recognized among economists, including neoclassical economists, that "institutions matter" in order to understand differences in performance among various economies. Nevertheless, consensus is yet to be reached among economists regarding what "institutions" really mean. Policy implications of the proposition, therefore, differ among different individuals.

Those who think that economic institutions should converge to the Anglo-American market system through the global integration of financial markets have been focusing on legal arrangements for property rights protection and contact enforcement, as well as informational transparency and disclosure. Actually, this kind of approach was quite influential, for example, in the early stage of transition of Russia to the market economy, but has failed to produce intended results. Is this simply a failure of politics in spite of good advice from economists?

I think that the premises of these economists' advice were problematical. Laws and organizations that are designed rationally and intentionally were mistakenly identified as "institutions." We need to conceptualize "institutions" so as to understand differences in economic performance.

Certainly, statutory law tends to influence people's behavior. Yet, it may produce unintended results, depending on factors such as its enforcement mechanism, the structure of the state, and social norms.

The lifetime employment system, the main bank system, for example, can be considered "institutions," but not in the sense that they were stipulated by codified contract or law. It used to be generally expected that anyone who graduated from college and joined a large company could maintain the employment relation until retirement, except in some emergency cases, and such shared expectations constrained the behavior of individual employers as well as employees. Arbitrary dismissal by employees became illegal in accordance with precedent, but that was a result of admittance of general perception, and not vice versa.

Thus, institutions might be conceptualized as follows. Institutions are a set of self-enforcing rules that have emerged from the game-theoretic (strategic) interaction of people in political, economic, social, and organizational domains, and have become taken for granted by all parties involved (see M. Aoki, Towards a Comparative Institutional Analysis, MIT Press, 2001). But then, why are there some particular expectations, and not others, that are generally shared by everyone and maintained in a stable fashion?

Expectations or beliefs that are shared by everyone as an effective rule of the game in various social and economic domains reflect an equilibrium state that has been attained in each of those domains. That is why it is not in the interest of any individual economic agent to deviate from such a rule. Furthermore, an equilibrium attained in one social domain is complementary with equilibria attained in other domains. This relationship is called "institutional complementarity."

Experiments with Various Business Models

In those terms we can understand why institutions tend to show rigidities and robustness, why they cannot easily be changed by law, and why unintended results are often obtained. Conversely, an "institutional crisis" can happen if, by chance, those expectations that have generally been taken for granted become problematical.

In this sense, Japan entered the era of institutional change sometime in the early 1990s. Since then, no one has been able to take for granted LDP one-party rule, bailing-out by main banks, or lifetime employment.

Then it should be clear why it is not so easy for a new stable institution to emerge. It will require not only the demise of the old rules of the game, but also the formation and convergence of new expectations in a mutually consistent fashion. Currently in Japan, people are surely anticipating changes in the rules of the game, but have not been able to see new rules clearly. And quite a few of them are wishing for their familiar rules to remain unchanged.

If institutions are not just codified laws, then their transformation will require the simultaneous occurrence of three things: First, in the private sector, various types of business models in a broad sense should be experimented with and effective arrangements should be evolved and chosen by competition, whether it is in industry, in education, in the medical field, etc. This is because no one can design the best institution (governance rules for organizations and markets) in advance.

Second, the government should abolish regulations that hinder entrepreneurial competition for experimentation, and instead introduce the kind of laws, regulations, and tax schemes that promote competition. New organizational models or new market rules might well hurt the interests of those groups that have benefited from old rules. In Japan, where labor immobility was the rule, interest groups tried to protect themselves by forming business associations that made new entry and new experimentation extremely difficult. That, in turn, lowered mobility among occupations and among organizations. In other words, the lifetime employment system and the formation of business associations are institutionally complementary.

Here the word "business associations" is being used in a broad sense to include such interest groups as farmers associations, doctors associations, the society of national universities, etc. Those groups put pressure on relevant bureaucracies through "zoku" politicians to protect their vested interests. Bureaucrats try to protect their legitimacy by transmitting those interests to the administrative processes. In this way, their mutually interdependent, collusive structure has been formed.

This kind of politico-economic institution may be called "bureau pluralism" (M. Aoki, ibid). In such an institution, there is no assurance of achieving global optimization of social welfare even if policies are locally optimized within each bureaucratic domain. The reason why this institution used to work to some degree, for example, in income equalization, is because there existed overseas models of future industrial structures for the Japanese economy.

The IT revolution and market globalization have made traditional business boundaries and organizational architecture largely obsolete, and along with it the effectiveness of Japan's "bureau pluralism" has become in serious doubt. Therefore, the third factor needed for Japan to overcome the current institutional crisis is political leadership to break down the compartmentalized protection of vested interests.

New Signs for Public Decision Making

Whether the Koizumi government satisfies this condition remains to be seen. But its high popularity rate probably means that the general public may be feeling the necessity of such a condition intuitively. Some characterize the Koizumi phenomenon as populism, but it is certainly different from the kind of populism that makes promises to rent-seeking interest groups without any discipline, since the Koizumi administration seems to be pointing to a small government.

Results are not yet visible in terms of abolition and restructuring of special public corporations. However, we can see some signs of new processes of public decision making to lower the walls of existing compartmentalization, beginning last year. One example is the recent episode regarding the introduction of the consolidated taxation system. If it were two years ago that the Taxation Bureau of the Ministry of Finance insisted that there was not enough time to prepare a bill, it could not have been implemented. In reality, however, its introduction beginning this April has been decided under the leadership exercised by the Council on Economic and Fiscal Policy. Many people are now beginning to doubt the legitimacy of the LDP tax council, which controlled tax policy prior to any decision by the cabinet.

Discussions on university reform, triggered by the issue of university-industry linkages, are gradually spilling over the closed domain constituted of the Society of National Universities and the Ministry of Education, Culture, Sports, Science and Technology. A year ago this could not have happened either. For another example, the introduction of safeguards (emergency import restrictions) against Chinese agricultural imports was avoided in spite of the strong pressure from agricultural "zoku" politicians. This is almost the first instance that their pressure was overridden by the prime minister.

It takes time to move from the occurrence of an institutional crisis to the emergence and stabilization of a new institution as a de facto system of rules. Even in the U.S., it took more than ten years until business model innovations were well established after Reagan's tax reforms and the Japanese and German industrial challenges inspired earnest effort for economic revitalization.

We do not have to accept the pessimistic notion of a lost decade, as we are searching for unknown and unprecedented rules for ourselves. Not just politicians and bureaucrats, but all citizens are participants in this searching process.