1. Introduction

At G7/G8 summits and on other solemn occasions, clarion calls for greater transparency of government methods of operation and institutional conditions and practices have rung out repeatedly. Even louder resolutions have been issued by NGOs who, of course, jealously protect the intransparency of their own internal processes. By focusing on the characteristics of particular entities or functions, these calls have tended to treat transparency as a static, self-referential property, as in optics, rather than as a dynamic system of political, cultural, and technological relationships, as in the social sciences.

In the latter, transparency is an interactive relation between a subject and indirect objects: elements of X are transparent to elements of Y and/or Z in dealing with each other. This relation is not necessarily mutual nor transitive and inclusive. Hence the fact that X is transparent to Y does not imply that X is transparent to Z, or that Y and Z are transparent to each other and to X, although degrees of transparency may exist in more than one direction. In fact, behind changes in the direction and length of the vectors of transparency lie changes in the distribution and extent of economic and political power and competence in society. All parties generally know what the prevailing transparency relations are and seek to change them to their own advantage. This makes transparency a fighting word in the social sciences even though there are public goods aspects as well.

Transparency is a characteristic valued in others. It is therefore valuable to withhold or not to give away without adequate compensation. In lieu of providing such compensation, force is often used to require transparency of some for the benefit of others. Principals, in turn, may require their agents, such as institutions and governments, to be transparent and accountable lest they exercise undelegated powers over them. Transparency of technologies, as of agents, also empowers principals because technologies that become transparent because they are based on publicly available knowledge, open access, and free use may have characteristics of public goods. Because the principle of exclusion does not apply to such goods, transparent high-powered technologies may also lend themselves to uncontrolled, indeed anarchic, abuse. They may be turned by individual private principals against others, or by small groups of terrorist principals against domestic or foreign societies and their vital institutions.

The latest surge in IT has begun to change the direction of transparency very much in favor of terrorists, particularly "private" organizations of terrorists that lack demonstrable support from any established government. Such terrorists now are able to gain anonymous access to advanced system-information resources and to avail themselves of relay and self-propagation mechanisms that allow them to impose significant civilian defense costs on the leading countries, in particular the United States. One of the foremost tasks of G8 cooperation would therefore be to contribute to the timely anticipation of the growing vulnerability to expert decentralized saboteurs who operate from multiple pseudo- or borrowed-host locations, particularly from within the United States, so as to deny retaliation any identifiable target.

2. Loss of Innocence About Transparency

A good place to start thinking about the transmissions of transparency and their uses to others is with children. In literature (e.g., Carlyle, 1843, Past & Present, ii.i), transparency has been identified as a child-like quality because of the presumed absence of dissembling and guile. This identification hints correctly that transparency may be a liability or involve either lack or forfeiture of an endowment or asset. Thus it could involve deprivation of the involuntary giver of transparency, like Athens being laid open to Sparta through the forced razing of its Long Wall in 404 B.C., while bestowing a control asset on the receiver, Sparta, to whom Athens then was accessible and more transparent.

In a broader examination, however, children would not be found highly transparent. For one, they are masters of "cheap talk," testing the limits of their own imagination and what might work with or on their parents through a blizzard of confusing and half-hearted propositions designed to elicit a response. Since conditional predictability is the fruit by which transparency must be recognized in the social sciences, parents would have to know at least as well as the children themselves what the latter will dream up next, and what they mean and want if the children were highly transparent to them. Not all parents can "read" their children quite so well.

Defining transparency by the property of conditional predictabilty implies that transparency must exist ex ante and yield tolerably accurate forecasts that are not systematically biased, and not just explanations after the fact. This definition does not allow for a wide gap between ex ante and ex post transparency. According to one interpretation (Enoch, Stella, and Khamis, 1997), ex ante transparency of agents refers to "clear" rules whose application to individual cases may nevertheless be so uncertain and "ambiguous" as to require public justification after fact. However, arbitrary and surprising acts, whether by agents or principals, do not become part of a transparent scheme of things just because they are explained or rationalized after the fact. Transparency can not make up ex post for a lack of it ex ante.

If conditional predictability thus is the hallmark of transparency, individuals are classsified as transparent to an interested party only if, given any specified set of circumstances and stimuli, that party can predict with a high degree of confidence how these individuals will act. By this definition, children are inclined to transparency in one respect: Unlike adults, they may not fully realize that their behavior is conditionally predictable so that they may be manipulated by outside control of their conditions. Though sometimes good at "playing games" and manipulating adults, they may not know or care enough to change their behavior and to take effective countermeasures even when they start to become aware of their transparency. Children may not yet be impressed with what Dixit and Nalebuff (1991, p. 185) emphasize is "the importance of avoiding patterns that can be exploited," but they are quite good at surprising even themselves, which, as the authors point out, is the surest way to surprise others.

Hence even if children react little to any recognition of vulnerabilities imposed by transparency, their conditional predictability may still not be high. Caregivers are advised to keep an eye on them because the pathways and steadiness of reasoning and logic, interests and motives, and commitments and abilities are still unclear and evolving in children. Their perceptions of conditions and their choice of behavior remain difficult to predict. Even the modus operandi of certain children may not be transparent if it could be so inconsistent that it cannot reliably be inferred and validated from repeated observation. In fact, when many of the important conditions and what follows from them are subject to change, learning, and innovation, conditional predictability may cover only what is left out: routine challenges and responses. The value of a subject's offering what limited area of transparency it can supply may be small if that area covers little that is of use to an interested party.

3. Privacy and Other Precious Information Asymmetries

Adults may be able to supply more valuable transparency precisely because they are more "rational" and "settled." However, application of that same steady logic and consistent pursuit of self-interest that could make the actions of adults conditionally predictable much in the way economists solve constrained intertemporal optimization problems by driving them to a mathematical conclusion also puts these same adults on red alert. It warns them that it is quite possibly dangerous and potentially ruinous to be transparent to others who are not transparent in return, and that even reciprocal transparency may provide unequal benefits to participating parties. "Many times in a game with imperfect information a player has information that might be valuable to an opponent and detrimental to the player if it were to be revealed. In such cases that player has a strong incentive to keep that information secret" (Gardner, 1995, p. 74). For this reason people choose to hide their tracks when stalking an objective and protect their privacy and poker face. When they must submit to questions about their status and concerns or comply with generally-accepted accounting standards or disclosure requirements in their business dealings, they try to confine their exposure to their lawyer, banker, or tax accountant, unless public disclosure is mandatory. When baring their bodies, minds, and souls to their doctors, psychiatrists, and priests, patients and parishioners expect their confidences to be respected. There are protections against forced self-incrimination in civilized society, and, unlike show-trials, trials are public to make the system of justice transparent.

As a general rule, individual actors thus clearly have an interest in being least transparent to others lest they lay themselves open to ready evaluation, manipulation, disrespect, and being taken advantage of. When someone is told, "I have finally figured you out," he or she should worry about being imposed upon and suffering a loss. Religious or not, many of us, in our basic sense of self, pride ourselves in being unique in some ways and endowed with free will. We acquire a personality that is not conclusively defined by heredity and genetic disposition nor even fully accessible to our conscious self. Hence we cannot fully be grasped, judged or summed up by anyone, including ourselves except "god." If an earthly "big brother" on the other hand were watching us day and night, as in George Orwell's 1984, we would feel hugely diminished and violated, with freedom and humanity denied. Not that monitoring just our outward behavior would make us entirely transparent since we would know that any private information revealed through our patterns of behavior might be held against us. But the pain of being on our guard day and night and distorting our behavior would be a heavy price to pay for denying transparency to a hated monitor who is trying to control our mind and affections.

Far from being a childlike, praiseworthy, disarming, and rather charming quality, transparency of principals, more often than not, is the product of being at the bottom of a hierarchy and subjected to collectivization, degradation, or at least some amount of tutelage and force. A high level of transparency of principals tends to be voluntary only in those limited areas of equality where individuals are willing to trade greater mutual transparency with those who concern them by making full and complete disclosures to each other, particularly in areas of perceived strength. Building trust and thereby reducing information and transaction costs and lowering risk premiums may be the reward, not to mention any comfort in sharing, empathy, and intimacy derived from the social part of our nature which economic paradigms generally fail to grasp. Horizontal transparency pacts may also occur between organizations and governments, for instance in the area of arms control verification or contract and treaty compliance. "Unilateral disarmament" also may have its rewards in some areas. Thus agents may opt for truthfully and completely revealing their goals, procedures, and track record to acquire a good reputation and credibility with principals and with other agents. However, the relation between a subject and the different parties that benefit from its transparency remains typically adversarial because their interests conflict. Accepted conventions or transparency codes are designed to arbitrate and to dull the edge of conflicts about moving in on any of the information asymmetries that are "pervasive in markets and life" (Gardner, 1995, p. 239).

Technological developments also may help break down information asymmetries by force of new technical capabilities even if there is no change in the cultural consensus about the proper limits of transparency. Such developments now contribute to making technologically unsophisticated individuals potentially more transparent without adequate compensation, i.e., against their will. Because of the IT revolution in the delivery of payment, ordering and communication services and in a range of tests, we leave ever more records of where we go and stay, what we buy and rent, and how we are diagnosed, priced, rated, and evaluated. These records can readily be retrieved, aggregated, and disseminated to generate a variety of profiles for interested parties all over the world. National law that inhibits profiting from unauthorized use of those types of information that should by law be treated as confidential provides only limited comfort. Targets for prosecution are first of all difficult to identify and then injunctions tend to prove ineffective against rapidly multiplying and self-propagating foreign and domestic targets. Thus opportunities for abuse are growing exponentially and without effective checks. Although it may be right for G8 summits to call for greater transparency of public-sector agents, it is noteworthy that they have shown no comparable concern for protecting legitimate principals against the growing technologically-driven encroachment on their rightful sphere of privacy.

4. Transparency of Agents and of their Rules and Procedures

In an ideal democratic society, the many principals would not be individually transparent to their collective agents. Rather these few key agents should be transparent to their many individual principals, and transparency should not all far in the opposite direction to preclude descending toward Orwellian control. The secret ballot, for instance, is a process and institution under which election procedures and the determination of election outcomes and their consequences are perfectly transparent and binding on politicians while the actual ballot choices of individual voters are not transparent. Voters may have identified themselves voluntarily as members of particular parties, interest groups, or lobbies, but even then it is not possible for them to "prove" for whom they actually voted because secrecy of the ballot is mandatory for their own protection as a group.

In dealings between agents, such as international financial institutions and member governments or member government institutions, calls for transparency have become common. A good example is the "Code of Good Practices on Transparency in Monetary and Financial Policies" recently adopted by the Interim Committee (now known as the Financial Stability Forum) of the International Monetary Fund (1999). The code includes an admirably comprehensive definition of transparency that includes all the ingredients and some of the incentives necessary for contingent predictability of institutional agents. Thus, for these agencies, transparency is defined as "an environment in which the objectives of policy, its legal, institutional, and economic framework, policy decisions and their rationale, data [related to the proper exercise of agencies' functions], and the terms of agencies' accountability, are provided to the public on an understandable, accessible and timely basis" (pp. 1-2).

The Code also shows an awareness that "giving away" transparency can be problematic if those to whom it is given choose to benefit improperly from this gift by playing upon exploitable aspects of agency behavior that have become conditionally predictable:

[T]he rationale for limiting some types of disclosure arises because it could adversely affect the decision-making process and the effectiveness of policies… For example, extensive disclosure requirements about internal policy discussion on money and exchange market operations might disrupt markets, constrain the free flow of discussion by policymakers, or prevent the adoption of contingency plans… Additional concerns could be posed by some aspects of transparency: … Moral hazard, market discipline, and financial market stability considerations may justify limiting both the content and the timing of the disclosure of some corrective actions and emergency lending decisions, and information pertaining to market and firm-specific conditions. In order to maintain access to sensitive information from market participants, there is also a need to safeguard the confidentiality and privacy of information on individual firms (commonly referred to as ‘commercial confidentiality')" (p. 2).

The same document notes that giving more transparency may require a quid quo pro of some kind to limit the redistribution of the means of control. It does so by asserting, "the benefits of transparency may be fostered by appropriate policies to promote transparency of markets in general, for institutions that are being supervised, and for self-regulatory organizations" (p. 3). Similarly, the Report of the Working Group on Transparency in Emerging Markets Finance (Institute for International Finance, 1999) balances its support for the IMF's transparency code as applied to financial agencies by noting "the importance of transparency of the private corporate sector in emerging market economies, to permit proper risk assessment" (p. 17). By trading greater transparency in one direction for increased transparency in the reverse direction, a compensating degree of reciprocity may be obtained. Put differently, the "free" provision of transparency, in situations where transparency can technically be withheld, is easier if it is reciprocated or shared by those to whom it is provided in some way. The IMF has internalized this insight by applying to its own conduct some of the principles of transparency it has promulgated for other agencies that may be parties to its programs and consultations.

The Code thus shows sophisticated awareness of the asset characteristics of transparency that make it a tradable private good or endowment in some respects and a public good, possibly open to over-exploitation, in other respects. It is less persuasive in crediting greater unilateral transparency to "all interested individuals and institutions (p. 3)" with greater efficiency of markets, more effective policies, greater policy consistency, and other good things (pp. 1-3) because these claims are not substantively hedged or supported. Still the Code provides a welcome contrast with recent, far less carefully hedged, G7/G8 calls for "greater transparency." The Group's ringing appeals appear implicitly to have equated a lack of public transparency with a litany of obvious evils -- from cronyism, insider dealings and organized crime to money laundering and pervasive government corruption -- thereby putting financial transparency in company with motherhood and apple pie. The "Corruption Perceptions Index" put out by a German NGO, Transparency International, makes the same link. Intransparency is a euphemism for corruption also when the latter is described, as in Kopits and Craig (1998, p. 15), as a situation in which the lines of demarcation between public and private spheres are not transparent.

Claims that assured benefits will flow from some particular increments of transparency also create dubious associations. Thus the insistence in a recent IMF (2000, p. 4) study that "greater transparency of economic and financial developments, through the publication of economic statistics, …, is essential to help strengthen/establish market discipline, and assure that asset prices and financial flows adjust less rapidly to adverse information" is based on faith, not evidence. There are bubbles, sunspots, and sunspot switching models, including some with empirical support showing the importance of extrinsic changes in beliefs (e.g., Salyer and Sheffrin, 1998). Sunspot disturbances represent shocks which influence equilibrium outcomes, but which do not affect fundamentals. Models that accord an important role to such non-fundamental disturbances do not provide much room for a smoothing effect from the unrestricted flow of statistical and other information. A recent analysis of the stock market valuation of Japanese corporations has concluded, for instance, "the Japanese stock market does not benefit from such information [about the future profitability of firms' investment]" (Ogawa and Kitasaka, 1998, pp. 206-207). Paraphrasing Georgantzas (2000, p. 14), it is probably true of financial valuations, as in deterministic chaos, that minute changes can lead to large deviations in behavior, but the dynamics of when and how they do so are, for practical purposes, unpredictable ex ante and poorly explained even after the fact. The latter shows that statistical transparency, though an intermediate good of some value, can not yield the final transparency product we most care about in this area. Perhaps egged on by international and national financial institutions and government agencies that are in the business of gathering and releasing economic statistics, G7/G8 summits have tended to promise too much by playing up statistical transparency as the key to reform and the avoidance of crises in international capital markets.

5. Imbalances of Transparency

Instead of assuming uncritically that the more transparency there is in any one sector the better, attention should focus on large imbalances or asymmetries of transparency and its new technological foundations. Two-way mirrors allow witnesses in a separate room to pick out criminals from a group without being seen to obtain any positive identification without undue fear and stress. The latest phase of the IT revolution may have turned tables: Now potential terrorists can look in at their leisure without being identified. There are important implications of this new direction of transparency for public security and defense planning.

The growing reliance on open or penetrable Internet communications technology may have begun to change the direction of transparency very much in favor of terrorists, particularly "private" organizations of terrorists that lack demonstrable support from any established government. Such terrorists now are able to gain anonymous access to advanced system-information resources and to avail themselves of relay and self-propagation mechanisms that allow them to impose significant civilian defense costs on the leading countries, in particular the United States. One of the foremost tasks of G8 cooperation would therefore be to contribute to the timely anticipation of the growing vulnerability to expert decentralized saboteurs who operate from multiple pseudo-locations, preferably from within the United States, so as to deny retaliation any identifiable target.

Instead of taking the lead in directing more resources to that goal, the U.S. government, with its reformulated antimissile defense program, for domestic political reasons, appears wedded to an obsolete concept of IT threats. Implementation of this "Maginot-Line" program would be wasteful for the United States because the costs are huge and completion far enough away for the currently envisaged constellation and delivery of threats to have been superseded by something quite different against which antiballistic missile defenses are inapplicable. A recent news account (New York Times, June 30, 2000, p. A 10) gives an impression of some of the costs and time schedules:

The scope of the proposed [national missile defense] project is both challenging and enormous, growing in four steps from an initial 20 missile interceptors in 2005 to a much larger system by 2011. By the time it is complete, the shield would require at least 2 launching sites, 3 command centers, 5 communication relay stations, 15 radars, 29 satellites, 250 underground silos and 250 missile interceptors. It would be based in Hawaii, Alaska, California, Colorado, North Dakota, Massachusetts, Greenland, Britain and possibly Maine. Two radars [radar stations] would be set up in Asia, possibly in Japan and South Korea. Building it would cost at least $60 billion and running it would take at least 1,455 people … The first phase alone is supposed to take eight years, the government says, though most experts consider 12 to 16 years a safer pace and less risky.

It is entirely conceivable that "rogue" states notionally would support acts of terrorism against the United States and other countries. However, concrete acts of government-sponsored terrorism might follow only if stealth and indirection of preparing for and executing such hostile designs could be such that rogue-state governments could not be identified as sponsors or as providing bases of operation. Launching increasingly miniaturized but nevertheless hugely capital- and coordination-intensive detectable and trackable land- or even sea-based missiles from identifiable platforms and locations against the United States would be a national suicide mission that no government in the world rationally could approve. It would be an extremely expensive and technically complex proposition requiring decades of concerted preparation that could hardly be sustained by suicidal intent. Terrorists could not produce or operate such offensive systems without a significant degree of government coordination and complicity, leaving governments nowhere to hide. Multi-point cyber and bio attacks launched preferably from within the United States, and using unwitting hosts and Internet access and replication technologies, are far more rational and thus likely. As long as there is a good chance that the perpetrators, or their hidden sponsors or contact organizations, can not positively be identified and punished, they can settle for diffuse destruction with low expected payoff. The IT summit should take heed that the nature of threats changes when attack-related new technologies become transparent from below and inexpensive while their potential abusers may know enough to make themselves invisible.

6. Summing Up What Has Changed

Transparency, like power, is distributive and not something to be maximized in democratic societies but to be finely balanced for their common good. Any change in the balance, whether driven by new technology, political restructuring, or cultural change, has pervasive consequences for welfare, freedom from intimidation, and security. Abrupt changes in political or technical operating regimes and in their transparency requirements and consequences thus have redistribution effects and create exposure to unaccustomed risks. As ordinary citizens find their protections against transparency eroded through new technology, the technology itself offers much greater reach and stealth to potential terrorists at low costs. By starting cyber-gang attacks, spontaneously associating hackers in a frenzy of destruction in which participants rush to play their recognized, and not pre-planned or pre-assigned, part, terror systems may acquire self-organizing features while being bent on disorganizing. Such self-organizing systems (see Georgantzas, 2000) are inherently nontransparent ex ante.

Perhaps the previous application to national defense shows how the distribution and capabilities of intransparency have changed. In its proposed missile defense program, the United States is continuing to develop privileged technology and complex applications that are not transparent to others, including the Congress, and whose performance characteristics and rationales are difficult to establish, let alone to scrutinize publicly. In effect, some part of the U.S. government, at the bidding of a politically but not technically competent agent, the U.S. Congress, is to own and operate that program which will be directed against remote offensive capabilities attributed to other countries' governments which also operate in secrecy. Each government expects its citizens to show trust and deference in critical matters of defense. The setting therefore follows the cold-war pattern of a government-to-government arms race where governments hold all the cards because they are not transparent to the citizens who are asked to cheer. The fact that no one else is likely to be willing to enter a contest whose winner is known with certainty beforehand does not means that "deterrence is successful." It could mean that a different type of contest looms that can inflict deadly harm and great damage by disorganizing networks and established systems with much more limited means and ends using technology that initially was a U.S. gift to the world.

The future technology that may be used to damage the United States and other G7 countries will have many inherently transparent, widely and anonymously accessible and replicable elements, with legitimate and illegitimate, harmless and hostile, foreign and domestic instigators not readily or definitively distinguishable. Threats thus will be privatized, dispersed, proxied and carried out by electronic infiltration into and from the territory of the party to be attacked with all or most explicitly military or particularly difficult targets bypassed. Attacks will have only diffuse political goals and need not have a high expected payoff. Rather, attackers may be satisfied to cause significant harm, civilian disruption, confusion, and humiliation not directly related and at best marginally and incidentally helpful to any substantive "military-political" issues they most care about. While it is hard for governments, at the highest levels, to deal with something largely outside their area of privilege and competence, the IT summit should try to urge greater joint preparations against threats from this slyly non-governmental sphere. These may be the supreme threats that will have emerged by the time the missile defense system is ready to start its pointless watch over threats that passed away in the President-Reagan era and will not arise again.

References

Enoch, Charles, Stella, Peter, and Khamis, May (1997) Transparency and Ambiguity in Central Bank Safety Net Operations, International Monetary Fund Working Paper WP/97/138-EAWP/97/138.

Dixit, Avinash K., and Nalebuff, Barry J. (1991) Thinking Strategically: The Competitive Edge in Business, Politics, and Everyday Life, New York: W.W. Norton.

Gardner, Roy (1995) Games for Business and Economics, New York: John Wiley.

Georgantzas, Nicholas C. (2000), "Self-Organizing Systems (SOS)," Draft, Graduate School of Business, Fordham University at Lincoln Center, New York.

Institute of International Finance (1999) Report of the Working Group on Transparency in Emerging Markets Finance, Washington, D.C.: IIF, March.

International Monetary Fund (1999) Code of Good Practices on Transparency in Monetary and Financial Policies: Declaration of Principles, adopted by the Interim Committee on September 26. http://www.imf.org/external/np/mae/mft/code/index.htm

International Monetary Fund (2000), Recovery from the Asian Crisis and the Role of the IMF, Issues Briefs for 2000 by IMF Staff. http://www.imf.org/external/np/exr/ib/2000/062300.htm

Kopits, George, and Craig, Jon (1998) "Transparency in Government Operations," Occasional Paper No. 158, Washington, D.C.: International Monetary Fund.

Ogawa, Kazuo, and Kitasaka, Shin-Ichi (1998) "Market Valuation and the q Theory of Investment," The Japanese Economy, 50(2), June, 191-211.

Salyer, Kevin D. and Sheffrin, Steven M. (1998), "Spotting Sunspots: Some Evidence in Support of Models with Self-Fulfilling Prophecies," Journal of Monetary Economics, 42(3), November, 511-523.