The paper by Pérez-López and Mesa-Lago provides a very eloquent account of the recent evolution of Cuban GDP statistics. It places a particular emphasis on Cuban GDP statistics since 2001 and explains the methodology used to estimate GDP as a result of a series of changes introduced in that year. The authors make several important contributions to the study of recent Cuban GDP statistics. For the sake of brevity, I will focus my comments on three crucial aspects of the paper: (1) GDP rebasing; (2) recent changes in the methodology to estimate (or calculate) GDP; and (3) the political context in which these changes have taken place.
As Pérez-López and Mesa-Lago indicate, recent changes in the methodology to estimate Cuban GDP statistics are divided into three chronological stages connected to GDP rebasing. Generally speaking, GDP rebasing can be carried out by implementing one or more of the following: (1) changing the relative weights of a given index or measure (i.e., changing or restructuring the relative weights of the specific finished goods and services used to estimate GDP); (2) changing the price reference period; and (3) changing the temporal reference period.
First, in the case of Cuba, as the authors demonstrate, GDP rebasing involved three types of changes: (1) changing the “base year” (used to adjust from nominal GDP to real GDP, and to calculate the GDP deflator) from 1981 to 1997; (2) changing the relative weights of the final goods and services included in the calculation of GDP to reflect the structural changes that have taken place in the composition of Cuba’s GDP (and in the Cuban economy) since the 1990s; and (3) changing the prices assigned to the final goods and services used to estimate (or calculate) GDP.
Second, as the authors show, 2001 was a turning point in the estimation (or calculation) of Cuban GDP statistics. In 2001, Cuba changed the base year (used to estimate “real”—or inflation-adjusted—GDP) from 1981 to 1997. The relative weights and prices of the finished goods and services used to produce GDP statistics were also changed. And services, such as banking, insurance, private housing rentals, tourism services, and the services offered by registered self-employed workers (RSEWs) or cuentapropistas were added to the calculation of GDP.
The year 2004 marked another turning point in the calculation of Cuban GDP statistics. In 2004, free social services were included in the value of GDP to reflect their importance and contribution to the Cuban economy. These included: community services (servicios comunitarios), personal services (provided by both the State and Non-State sectors), and exports of professional services (particularly “internationalist” health care workers). Finally, 2007 was another turning point, following the revision of the community services, and social services included in GDP. This was primarily done to reflect the value of these State-provided services in the calculation of Cuba’s GDP statistics, to bring Cuban GDP statistics closer to the SNA methodology recommended by the U.N. and other international organizations, and to reflect the evolving nature of Cuba’s economy in the post-classical socialist era.
As Pérez-López and Mesa-Lago indicate, these adjustments have significantly impacted the estimation (or calculation) of GDP statistics: (1) the change in the “base year” from 1981 to 1997 has resulted in annual average increase of 60% in the value of GDP, compared to its value during the 1996—2000; and (2) adding services to GDP statistics has increased its value by an average of 2 percentage points since 1996.
According to Pérez-López and Mesa-Lago it is not possible to conduct a robust analysis of the methodology used to estimate (or calculate) recent Cuban GDP statistics, and its evolution, for the 1996-2007 period due to the discontinuities in the data series provided by Cuba’s National Statistics Office (Oficina Nacional de Estadísticas—ONE). Their comparison of Cuban GDP statistics for the 1996-2000 period with GDP statistics for the 2001-2003 period, and GDP statistics for the 2004-2007 period generates the following observations:
- The values of GDP during the 2001-2003 period are notably lower than the values estimated for the 1996-2000 period, and the 2004-2007 period.
- GDP values for the 2004-2007 series (or period) are significantly higher than GDP values for the 1996-2000 series. On average, this difference ranges from 15% to 19%.
- GDP values during the 2004-2007 period are 6% to 7% below values estimated for the 2001-2003 period, but 7% to 11% above the values recorded during the 1996-2000 period.
Finally, the authors point out to the differences between ECLAC and Cuban authorities with respect to the methodology employed to estimate Cuban GDP statistics. They also assert that it appears that these differences have been resolved at the present time since ECLAC seems to have accepted Cuba’s (rather unique) adaptation of the SNA methodology to estimate GDP.
I agree with the general premise of Pérez-López and Mesa-Lago: the recent adaptation of the SNA methodology to estimate Cuba’s GDP statistics has resulted in several appreciable changes in official GDP figures (since 2001) compared to the data compiled during the 1996-2000 period. Furthermore, there has been very little clarification and explanation on the part of Cuban authorities with respect to the specific details about the methodology or approach used to calculate recent Cuban GDP statistics.
And third, I am inclined to believe that these “discontinuities” and “puzzles” are, rather than a deliberate attempt by Cuban authorities to cover or mask the true state of the Cuban economy, the result of several strategic considerations that extend beyond the realm of the “dismal science.” The official decision to limit (or not disclose) the information and methodology used to estimate Cuban GDP statistics, thereby constricting its transparency, has ideological and national security dimensions linked to the “Battle of Ideas” and recent changes in Cuba’s external environment.
From the Cuban perspective, Cuba remains a country under siege. As it well known, this sense of operating under external threat and pressure intensified after the U.S. invasion of Iraq in 2003, and growing U.S. support for Cuban opposition groups. While the rest of the world, with one or two exceptions, condemns the “let’s isolate Cuba to propel regime change approach” embraced by the U.S. for almost five decades, and Cuba has been able to expand its international linkages with the rest of the world, the Cuban leadership believes that its capacity to confront its powerful northern enemy rests—to a large degree—on its ability to manage and control information. The ability to effectively manage and control information in all fronts, from their perspective, represents a crucial element of the defense and national security of the country.
This strategy can be defined as the “national security or defense argument” for maintaining State control of all types of information, including economic statistics. An essential part of this strategy is to defend the uniqueness of the “Cuban model,” while embracing modern, widely-accepted practices for the estimation (or calculation) of national income, and other important economic, demographic, and social statistics. Defending the unique characteristics of the “Cuban model,” as reflected in official socioeconomic statistics, is an integral element of the “Battle of Ideas.”
In the case of Cuba, as in most other countries, national account statistics are also dependent on data quality. In this context, data quality spans the boundaries of other variables like information availability, which itself is dependent on the systematic (institutional) mechanisms in place to collect and validate data and statistics and the frequency and reliability with which data are collected. I would argue that in the case of Cuba, despite very profound “material necessities”—that is, the need for an improved data collection system, with modern equipment and technology—there is a deeply ingrained sense of duty and responsibility on the part of most individuals in charge of collecting data and compiling statistical reports. During a June 2009 visit, I had the opportunity to speak with several mid-level ONE officials, economists, statisticians, and field workers in various provinces and municipalities. While our exchanges were informal, and by no means constitute any form of empirical evidence of field research, I was highly impressed by their professionalism and familiarity with modern research and statistical methods despite operating in a very difficult environment with severe material limitations such as insufficient or antiquated supplies and equipment, limited broadband capabilities, etc. Most ONE employees that I met take their jobs very seriously and understand the importance of collecting and providing reliable data and statistics.
Nevertheless, as several of them pointed out in private, and as Pérez-López and Mesa-Lago set out in their paper, there are several issues concerning recent Cuban GDP statistics that merit more detailed explanations or clarification from Cuban authorities. By providing those of us interested in understanding and analyzing the Cuban economy with more detailed information about the methodology employed to calculate GDP statistics, Cuban authorities would hardly compromise national security. Instead, they would provide students of the Cuban economy with the necessary tools to study, understand, and analyze its characteristics and evolution in the rapidly-changing post (classical) socialist era.
To conclude, I believe that the work by Pérez-López and Mesa-Lago makes its greatest contribution to the study of the Cuban economy in its attempt to close the existing gaps surrounding recent Cuban GDP statistics created by the absence of official explanations and clarifications about the methodology used to estimate the final value of the goods and services produced by the Cuban economy.