Cuban economy, and socialist economies in general do not strike me as being high on transaction costs. After all, eliminating transaction cost was the main raison d'etre of planned economy (or vertical integration as institutionalists would say, cf. Oliver Williamson, _Markets and Hierarchies_).
To clarify a point or two, it makes sense to distinguish transaction costs from systemic inefficiency, the later being systemic feature that create disincentive to a more efficient utilization of resources.
One example of systemic inefficiency in central planning is unintended effect of regulating wages and prices. Such regulation was the key mechanism to avoid inflation and profiteering, especially during the rapid industrialization phase. It initially increased efficiency because it prevented food producers (mainly labor-intensive small farms) from profiteering from increased demand for food created by urban growth. Administrative regulation of food prices allowing keeping industrial wages low, and thus reduc4ed the cost of industrialization.
However, administrative regulation of prices and wages had an adverse effect on productivity further down the road. This regulation produced a pool of cheap labor, which created a disincentive for technological innovation, which almost invariably is pursued to save labor cost. Since cheap labor was institutionally guaranteed by fixed prices and fixed pay schedules, firms had no incentive to modernize they technologies - especially that such modernization often relied on imports, which posed a real cost to firms due to anti-import currency exchange policies.
As a result, centrally planned economies became increasingly less efficient, but that inefficiency was caused by disincentive to innovate rather that due to increase in transaction costs.
A transaction cost is a cost of operations that are auxiliary to the main economic activity of the firm yet they are necessary to successfully perform that main activity. A typical example of such cost is legal cost associated with securing and monitoring contractual obligations, product liability, regulatory requirements, etc. In centrally planned economies, such costs were rather low vis a vis market economies due to vertical integration, which means that most transaction among productive units were not contracts among independent units, but administrative assignment. Consequently, the cost of marketing research, contract monitoring and enforcement and liability were eliminated. The main transaction cost of such economies was the cost of the planning apparatus (statistical data collection, planning commissions) - which was quite low (this is basically the cost of running such institutions as the Bureau of the Census, Bureau of Economic Analyses, or Bureau of labor Statistics).
Transaction costs are typically a result of the complexity of the economic system as a while - the greater the complexity, the greater the transaction cost. It is generally the cost of maintaining the system's efficiency, but it may also produce inefficiency, especially in situations in which it arises in connection to complexity that is artificially maintained to preserve monopoly rights of certain interest groups. An example is the US health care system, where multiple private insurance firms add considerable transaction cost. This complexity is artificial, because simpler systems (e.g. single payer) exist and are proven effective yet they are not implemented due to political influence of insurance companies.
So to summarize, centrally planned economies were inefficient due to their structural features rather than transaction costs, and that inefficiency was particularly pronounced in the agricultural sector. Transaction costs are much higher in more developed economies due to the greater complexity of these economic systems, but there also a certain level of artificial complexity (especially in the US) that produces unnecessary (i..e. avoidable) transaction cost.
Wojtek