The costs of managers’ self-protection and high-level indecision interacted and fed off each other. At lower levels, the Soviet officials that left these issues hanging in the air were demonstrating indifference to heightened transaction costs. While they left matters unresolved, officials and managers below them continued to avoid responsibility where necessary, to work around the rules where possible, and to take out the insurance that seemed to be recommended—to give time and effort to lobbying Moscow for change. So, decision costs and trade costs compounded each other. The officials that repeatedly delayed effective resolutions tolerated this negative spiral.
As the short term drifted into the long term, a mitigating factor was the human capacity to normalize changes in the environment. At first, the new secrecy law was frightening. Officials did not know how it would work and how it might affect them. It took several years to learn how to work around it and insure against it. Eventually, however, it became the new “normal.”
From this point of view, procrastination had its merit. In postponing resolution, senior officials simply waited for managers to find a workable accommodation to the new regime. Eventually, they legislated in such a way as to formalize this accommodation. There is no evidence that this happened by design, but it suggests that wait-and-see was a sensible strategy from the private perspective of those involved.
The effects of the secrecy law of June 1947 offer a natural fit to the concept of the secrecy/capacity tradeoff. Figure 4.2 extends the diagram introduced in Chapter 1 (Figure 1.2). As before, the vertical axis measures state capacity, and the horizontal axis measures the degree of secrecy (which in turn depends on the mix of regulation and penalization).
The outcome of this figure will be, as before, to show state capacity with an effective maximum at the degree of secrecy S*. But the figure’s starting point is the notional maximum of state capacity, shown as the level of the horizontal dotted line level marked C. Think of notional state capacity as what might be achieved by selfless, fearless officials, true “servants of the people” who seek neither personal advantage nor self-protection. The ideal limit is not achievable, because real servants of the people are human and therefore to some degree self-interested. I don’t mean that all public figure 4.2. The secrecy/capacity tradeoff: Secrecy adds to state capacity by cutting leakages, but subtracts from it by augmenting fear
Note: On the horizontal scale, higher degrees of secrecy (marked by increasing regulation and penalization) lie to the right. On the vertical scale, C marks the level of notional state capacity, as defined in the text. At each degree of secrecy, effective state capacity is notional capacity, less leakage costs, less fear costs. S* marks the degree of secrecy that delivers the maximum of effective state capacity. From Z, a point initially preferred by the dictator, an increase in secrecy moves the state to Z': leakage costs are further reduced, but fear costs increase by much more, so effective state capacity declines
servants are venal; clearly, in fact, some societies have been much better than others at approaching the upper limit, because of their social norms and their methods of selecting and training noncorrupt public servants. Nonetheless, the supply of saints is limited everywhere, and everyone has private needs and desires, so public administration is never perfect. Because of those private motivations, effective state capacity will always fall below notional capacity.
The figure simplifies the private motives of government personnel into two deductions from notional state capacity, arising from leakages and from fear respectively. Effective state capacity is what is left after accounting for the combined costs of leakages and fear.
Leakage costs are what is lost from state capacity when officials leak government information in order to share a benefit with private or foreign interests. This is the violation of which Kliueva and Roskin were accused. The reasoning behind the idea that state capacity is reduced by leakage of government information might be that the information leaked is intrinsically valuable, or that leakages deprive the government of its first-mover advantage in games of strategic rivalry. In the KR affair, as Stalin and Zhdanov saw it, Kliueva and Roskin sold information that belonged to the state and might have been of both commercial value and of value to national prestige, in exchange for personal fame and favors in the West.
The Politburo response was to intensify information regulation and to create new offences subject to criminal penalties, in the expectation that these would deter further leakages. In other words, the Stalin-Zhdanov model of regulation and penalization predicted that more secrecy would reduce leakages. Large at first, then diminishing as the degree of secrecy is increased, leakage costs are shown in Figure 4.2 by the vertical distance between notional state capacity (the horizontal dotted line) and the dashed curve. As the curve suggests, it might not be possible to eliminate all leakages and, in that case, each further increase in the intensity of secrecy would have a diminishing marginal effect. Still, we already know that the results were effective to some degree.
Fear costs are the other deduction from notional state capacity. These are shown in Figure 4.2 by the vertical distance between the dashed curve and the solid curve. When the state is transparent, there is no fear.
Suppose the regulation of official life becomes more complex and demanding, and penalties for violation become both more severe and also, because of growing complexity, less predictable. Then, fear will spread through officialdom. The human response to fear is to seek insurance, individually and in groups. But the process of building insurance consumes time and effort, which are diverted from the resources assigned to the officials’ state business. As secrecy becomes more intense, fears multiply. As fear grows, the costs of fear increase at an increasing rate, shrinking state capacity.
Effective state capacity is always less than notional capacity because of the deduction of leakage costs and fear costs. More secrecy narrows the gap at first, because large leakage costs can be easily suppressed while fear costs remain low. At first, effective state capacity is increased. It rises to a maximum where secrecy is set at S*. Beyond that point, fear costs take over because, while officials are already too frightened to leak much information, increasingly they are also too frightened to do their jobs. As the fear effect sets in and increasingly dominates their working lives, so effective state capacity declines slowly at first, then rapidly.[226]
226
In this perspective the Soviet political economy is viewed from above as a hierarchical command system. The rationality of the model belongs to the ruler. In the context of the historical literature on Stalinism (surveyed by Kotkin, Magnetic Mountain, 1-25), its spirit is closer to a totalitarian view than one that emphasizes plural interest groups or independently operating factions, but it is not designed to lend support to that view. It captures the facts that Stalin was clearly in charge, but also had to reckon with the behavior of others for whom unhesitating obedience would be just as dangerous as overt resistance. The dictator presented them with a Catch-22, from which they had to protect themselves at the same time as they attended to his orders. Richer models are possible, but they are not required here to understand the matter at hand.