Back to Index

Challenge to Cuba and Nicaraguan literacy claims

Tim Brown answers Hank Levin: "Professor Levin asks for some documentary evidence that Marxist ditatorships falsify literacy data. I suggest he look at Charles Stansifer, "The Nicaraguan National Literacy Campaign", American Universities Field Staff reports, Hanover, Vt.1981. On page 12, Stansifer reports without criticism that literacy rates in Nicaragua by Department [province or state] had increased by up to 62.61% from October 1979 to August 1980, thanks to that revolution's Lieracy Campaign. He reports that by August 1980 literacy had reached 95.85% in the Department of Esteli, 97.77% in Chinandega, 87.81% in Nueva Segovia, and 63.75% in Matagalpa. He fully accepts without wuestion that this miracle happened. These and comparable data were then routinely repeated throughout the Revolution by UN agencies, academics, and others.

Eleven years later, in 1991, the OAS and UN tested literacy rates in the rural areas of these same four Departments, Esteli, Chinandega, Nueva Segovia, Matagalpa, where 80% or more of their populations live. Literacy rates were found to be between 3% and 5% [illiteracy of 95% to 97%]. When Iooked at both Stansifer's data and these later data while researching my book The Real Contra War, I was able independently to substantiate the UN/OAS data but not Stansifer's, and could imagine only two possible explanations for the amazing difference between them. One, while about 90% the peasants in the Segovian mountains had, in fact, learned how to read and write during the initial nine months of the Revolution's literacy campaigns, they had forgotten how by 1991; or, two, the data reported by Stansifer were at the very least wildly exaggerated and more probably had simply been invented by the revolutionary government. Regardless, I also found that even though these data were wrong, they were then repeated almost ad nauseam in subsequent publications by opponents of US policy toywards the Sandinista Revolution, by sympathizers of that movement, and by international organizations and academics even though they were based purely on unsubstantiated claims made by the Sandinista government in its reports.

On Cuba, I'm happy finally to learn what methodology UNESCO used, "testing across countries using a common methodology and standardized statistical tests of the results." The only way to obtain entirely valid results by using this methodology is first fully to identify the entire universe to be tested [N], in the case of literacy that would be everyone in a country between the ages of 12 and 65 except the mentally incompetent, and then to test each individual in that universe. This is the only way to obtain statistically defensible results with almost no margin of error. Since this is usually prohibitively expensive and time-consuming when a universe is very large, the common approach is to use random sample surveying techniques to approximate a fully valid result within a mathematically quantifiable margin of error, in this case to approximate literacy levels within a stated percentage of probability of accuracy, say 3-5% one way or the other. To be valid, this also requires first obtaining a complete listing of everyone in the universe to be examined. Then one must assigning a number to each of them, determine how many need to be tested in order to draw statistically defensible conclusions, then generate a valid sampling set from among them using a random units table, then locating each person selected, and then testing only them. Unless the internal integrity and research independence of each of these steps is guaranteed, the results lose statistical integrity and more and more become simply approximations. The less freedom and independence the researcher has to identify, select, contact, and test all the subjects selected, the less defensible the results. It is still possible to manipulate the findings, but such manipulations become increasingly problematical the further they stray from the core data.

As far as I know the only complete lists of Cubans available are held by the government, especially its security services, and there is no way to determine whether those used to do the UNESCO study were partial or complete other than to depend on the government's honesty. No independently held lists exist, and, even if it did, making direct contact with Cuban subjects for such a study without the prior knowledge, consent, and indeed cooperation of the government, would be illegal. And working exclusively, or even primarily, through the government immediately subjects any such research project to Potempkin Village manipulation by a regime that is expert at such things. Perhaps I am reading too much into Professor Levin's short comment, but his description leads me to conclude that the UNESCO studies to which so many have made reference are probably not statistically defensible. That ia a conclusion based on an examination of the methodology reportedly used, not ideology.

This conclusion probably applies also to the results obtained in other Latin American countries as well, including some democratic ones, although to a lesser extent. This should not come as a surprise to anyone. The literacy rates in many Latin American countries, and especially of dictatorial regimes such as those of Somoza, Stroessner, Trujillo, Sandinista Nicaragua, and apparently in Paul Simon's China of today, have routinely been exaggerated by the ruling regime in order to enhance its prestige. What surprises me is that so many otherwise open-minded and objective professionals seem incapable of evn entertaining the possibility this could also be true of the Castro dictatorship. Could it be that ideology, interests, or just plain sympathies have not dimmed only my vision?

Ronald Hilton - 3/26/02