Nicaragua
Nicaragua's Poll Wars
Envío team
In recent months, national and international organizations have unleashed a rash of public opinion surveys in an attempt to predict the outcome of Nicaragua's upcoming elections. The polls generally fall into two categories. Those carried out by ITZTANI-INOP, Greenberg-Lake, ECO and the Manolo Morales Foundation show the FSLN with a clear advantage over UNO; while surveys carried out by Borge and Associates, CID-Gallup and Univision show UNO strongly in the lead or neck and neck with the FSLN.
Two fundamental criteria determine the validity of a public opinion poll—the representativity of the sample and the objectivity of the pollster. The sample must be representative of the population as a whole, including appropriate age, gender, occupational, urban/rural and regional distributions. The pollster's objectivity must pervade the formulation of the questionnaire, the organization and presentation of the field workers and the analysis and presentation of the results.
The following "Who's Who" of polling agencies touches on these issues of representativity and neutrality, as well as the polling history of the seven different pollsters who have carried out surveys in Nicaragua. In December, West Germany's Friedrich Ebert Foundation sponsored a seminar in Managua for all the different pollsters, with the goal of establishing a methodological debate. Each polling agency was
invited to present its sampling methodology and contribute to the discussion. CID-Gallup and the Manolo Morales Foundation did not attend. Univision attended but did not present its methodology. Says Raúl Obregón, director of ECO, "I could state my own opinion but I think that this fact is more revealing. To us, this is evidence that [these polling agencies] are not disposed to openly explain how they carried out their polls."
Only after February 25th will we know definitively which pollsters were accurate. It is important, however, to remember that polls can be used not only to reflect but also to affect public opinion. The international agencies of Borge and Associates, CID-Gallup and, possibly, Univision, have strong ties to the Nicaraguan opposition and/or the US government that interfere with their neutrality. UNO's strong lead
over the FSLN, "certified" by their surveys, could provide the basis for the US and UNO to "cry fraud" after a legitimate Sandinista victory in February.
The pollstersITZTANI-INOP: an independent research institution under the auspices of the parent organization INOP, the Nicaraguan Institute of Public Opinion. ITZTANI-INOP pioneered polling in Nicaragua, carrying out the first publicly disclosed survey in June 1988.
The institution began operating with technical, moral and financial support from several experienced pollsters including Jesuit priests Peter Marchetti and César Jérez of the Central American University (UCA) in Managua; the late Father Ignacio Martín-Baró, authority on public opinion polling in El Salvador; and Bill Bollinger and Daniel Lund of the US-based Inter-American Research Center. Several ITZTANI staff members had prior experience in public polling. In the past 19 months, they have now directed or administered five public opinion surveys, including the most recent poll designed by Greenberg-Lake.
In each ITZTANI poll, quotas for age groups and gender were set based on statistics from the Nicaraguan Institute of Statistics and Census (INEC). In Managua, additional quotas were set based on type of neighborhood (upper class, working class, etc.). A random sample was then selected based on these quotas.
Greenberg-Lake / The Analysis Group: a Washington-based political consulting firm with extensive polling experience in the United States and the Third World. It is said to be the second most prestigious polling agency in the United States next to Gallup. In the US, the firm has conducted polls for national and state candidates, including Senator Christopher Dodd and former Vice-President Walter Mondale, as well as for public and private organizations on public issues. In surveys before the elections this year in El Salvador and Paraguay, Greenberg-Lake worked with local university specialists and survey or marketing professionals to create an election survey team. In both cases, the firm conducted the first national polls carried out in each country. Working with ITZTANI, Greenberg-Lake recently did the same in Nicaragua. In both El Salvador and Paraguay, their results came within 3% of the actual election results.
Greenberg-Lake's is the first truly national random sample—including the Atlantic Coast and the rural population—and the first to be based on the 1989 voter registration lists. Unlike INEC's census statistics, which are extrapolated from a 1972 national census, these lists are up-to-date and clearly accurate. Quotas for age group, gender and geographic location were set from these lists, and only registered voters were polled. Greenberg-Lake conducted the poll through personal interviews but allowed voters the option of a secret ballot for making their choice for president.
Estudios y Consultas de Opinion (ECO): an independent research firm that has carried out numerous polls in Nicaragua since May 1988. Until its October poll, the research team primarily conducted economic, marketing and communications media studies for private clients. The firm's election polling has been contracted by the UCA. The professional team includes pollsters with experience in El Salvador, Mexico and Costa Rica.
ECO has the distinction of working with the largest samples of any of the polling organizations, in order to decrease the margin of error. ECO used INEC statistics to set quotas by age, gender and region and selected random houses by computer. The second poll included rural municipalities (1,025 of the total interviews), but ECO directors were unsatisfied with the representativity of the sample due to problems of access and chose not to include rural areas in their most recent poll.
Manolo Morales Foundation (FMM): associated with the Popular Social Christian Party (PPSC). The Center for Documentation and Social Analysis of the FMM has carried out three publicly-disclosed opinion polls in Nicaragua for the independent opposition weekly La Crónica. The director of FMM and La Crónica, Luís Humberto Guzman, is a member of the faction of the PPSC that chose to remain with UNO when the party left the coalition. The Foundation's surveys are believed to have been funded by West Germany's Friedrich Ebert Foundation, and field work was directed by José Gabriel Moya. FMM's poll results, published in La Crónica, have generally been in line with other reliable polls and are considered legitimate by these other pollsters.
Regarding its sampling methodology, La Crónica has generally stated only that its samples are random, filling quotas for age group and gender. In spite of repeated phone calls to both the Foundation and La Crónica, pollsters were unavailable for comment.
La Prensa / Borge and Associates: Opposition newspaper La Prensa is owned and published by UNO's presidential candidate Violeta Barrios de Chamorro. La Prensa's polls were carried out by Victor Borge of the firm Borge and Associates of Costa Rica. Borge is an experienced political consultant and pollster and has worked for numerous National Liberation Party candidates in Costa Rica including Luís Alberto Monje, Oscar Arias and Carlos Manuel Castillo. He is said to have helped guide Arias' come-from-behind presidential victory in 1985.
La Prensa's poll results are diametrically opposed to those of ITZTANI-INOP, Greenberg-Lake and ECO. Though Borge defends the plausibility of the results, even he admitted that the nature of some of the questions in the May/June poll's questionnaire (not publicly disclosed but obtained by analysts through a third party) could have affected responses to later questions and therefore biased the results.
La Prensa has refused to reveal important information about its polls, including the full questionnaire, sampling methods and detailed results. The polls have clearly political ends and cannot in any way be entrusted with the ethic of impartiality. The results of the first poll were timed to be released just prior to and during the July 19th tenth anniversary of the revolution and was an attempt to "rain on the FSLN's parade." The poll was also designed to encourage the numerous opposition parties to unite behind Violeta Chamorro as presidential candidate, which many did shortly after, and to create the impression that the FSLN cannot win a free and fair election by demonstrating overwhelming support for the opposition. In the same vein, La Prensa's second poll, which appears to have focused primarily on the economic crisis, was designed to show grave discontent with the Sandinista government.
Since La Prensa refuses to disclose the sampling methodology used, it is impossible to fully analyze the accuracy of its sample. However, some information made available with respect to the October poll reveals disproportional representation of people over 30 and of housewives, who tend toward the opposition.
Univision-News is the leading US Spanish-language television network in the US. The network has done extensive polling in Latin America with generally accurate results. Seven pre-election polls carried out by Bendixen and Schroth for Univision since 1984—in El Salvador, Peru, Guatemala, Honduras, Colombia, Chile and Panama—all fell within 4% of the actual election results.
Though Univision has not fully disclosed its methodology, experts attribute the discrepancy between the results of their Nicaragua poll and other reliable polls to Bendixen and Schroth's choice of sample. The sample includes departmental capitals plus eight additional municipalities that are intended to make the poll "national." But the municipalities were selected by geographic location and are not politically representative of the nation as a whole. This can be seen by comparing their 1984 voting records with that of the average for the region in which they are located. In four of the eight municipalities, the opposition took between 5.8% and 21.7% more votes than the regional average. The municipality of Teustepe, for example, had the highest percentage of opposition votes in the nation in the
1984 elections—at 21.7%, higher than the Region V average. When challenged on the issue of sampling methodology in the recent Friedrich Ebert Foundation seminar for polling agencies, Schroth stated that he would send additional information about their methodology that would clarify the situation.
At least some other pollsters trust Univision's reputation and do not believe that their choice of unrepresentative municipalities was intentional. Individuals within Univision, however, have close ties to the Nicaraguan opposition. Although this is not a guarantee of bias in polling, as can be seen with the Manolo Morales Foundation polls, it is worthy of note when the results are questionable. Carlos Briceño, correspondent for Univision's main cable station (Channel 23 in Miami), is also one of the key people behind UNO's effort to set up an independent television station in Nicaragua. Sergio Bendixen, of Bendixen and Schroth, is on the board of the Washington-based Bipartisan Committee to Guarantee Free Elections in Nicaragua, whose primary purpose is to support UNO.
Consultoría Interdisciplinaria en Desarrollo (CID): a Costa Rican affiliate of Gallup International Research Institutes (GIRl). CID-Gallup has conducted polls in El Salvador, Honduras and Costa Rica but does not have a reputation for accuracy. For example, CID polls in El Salvador before the March 20, 1988, municipal and National Assembly elections showed Duarte's Christian Democrats maintaining a wide margin over the right-wing ARENA party. In the elections, however, ARENA crushed the Christian Democrats.
The U.S. Information Agency (USIA) has used CID-Gallup to carry out polls in Central America that conveniently support US policy in the region. It is likely that CID's Nicaragua polls serve the same end. In March 1986, right-wing Costa Rican newspaper La Nación, to which CID founder and half-owner Charles Denton gives exclusive rights over his poll results, released the results of a CID poll "proving," on the eve of a congressional contra aid debate, that 69% of Costa Ricans supported military aid to the contras. It was later revealed that the poll had actually been conducted nine months earlier, on the heels of a Nicaragua-Costa Rica border incident. A more recent CID poll, in fact, had showed 42% of Costa Ricans opposing contra aid and only 39% supporting it. Reporters who have attempted to track down information on methodology used in CID surveys have gotten the run-around from both CID and the USIA
An article highly critical of CID-Gallup published in The Nation (May 7, 1988) led Andrew Kobat, director of the US Gallup organization, to publicly disassociate the US organization from CID's surveys. He wrote that Gallup "neither owns nor controls" CID and "is in no position to judge the quality of [its] work for USIA" He acknowledged that "the US government played up the Gallup connection to give the survey greater credibility."
CID's poll was conducted in Managua, seven other major cities and 20 other municipalities. Points were chosen in each of these areas in a radius of which no more than 10 interviews were conducted. Age, gender and urban-rural quotas were set based on INEC statistics.
Analysts have attributed CID-Gallup's strongly pro-Violeta Chamorro results to the questionnaire design (not made public, but obtained through a third party). In an analysis by political scientist Dr. Paul Oquist, who formerly worked as an expert in Research Methodology for the United Nations and is currently Director of Information in the FSLN's National Campaign Office, he writes, "This questionnaire is a textbook case on the subject of massive and systematic induction in polls. It is also classic with respect to its one-dimensionality, In the entire poll there is not a single question about the political theme of war and peace."
The Polls
1988
June 4-6
ITZTANI-INOP
Sample: 1200
Managua only Methodology fully disclosed
Key results (not election-related)
Principle political problem in Nicaragua:
war/US intervention/contra 50% poor gov’t/lack of freedom 13.7%
Opposition 2%
Lack of political project 2%
Other 11.8%
Don’t know/no answer 20.6%
Principle cause of war:
Reagan/imperialism/contra/USA 47%
FSLN 8%
Lack of democracy 6%
Struggle for power 8%
Communism/totalitarianism 2%
Other 13%
Dec 11-19
MANOLO MORALES FOUNDATION
Sample: 1800
Managua/Masaya/León/Granada/Chinandega/Matagalpa
Methodology not disclosed
Key results
Presidental preference:
Daniel Ortega 12.3%
Sergio Ramírez 10.7%
Tomás Borge 9.1%
Virgilio Godoy 5.7%
Enrique Bolaños 2.9%
Mauricio Díaz 2.3%
Alfredo César 1.6%
Erick Ramírez 1.1%
Cardenal Obando y Bravo 0.7%
Edén Pastora 0.5%
Other/undecided (not published)
Party preference:
FSLN 32.7%
Opposition 29.8%
Don’t know 29.5%
Won’t vote 8.0%
1989
March 28-April 6
MANOLO MORALES FOUNDATION
Sample: 2800
National, except Atlantic Coast
Methodology partly disclosed
Key results
Party for president:
FSLN 29.6%
An opposition party 36.2%
(the PLI leading with 11.8%)
undecided 25.1%
won’t vote 9.1%
Individual for president:
Ortega 15%
Other Sandinistas 16%
Violeta Chamorro 15%
Other opposition leaders 13%
Undecided/no answer (not published)
May 13-14
ITZTANI-INOP
Sample: 1500
Managua/Masaya/Granada
Methodology fully disclosed
Key results
Party for president:
FSLN 18%
an opposition party 20%
don't know/no answer 29%
won't vote 20%
Individual for president:
FSLN leaders 30% (Ortega 18%, others 12%)
opposition leaders 13%
don't know/no answer 36%
no one acceptable 19%
Would you vote for a candidate of the
armed resistance?:
yes 14%
no 71%
maybe, 3%
don't know/no answer 12%
May 27-June 6
BORGE AND ASSOCIATES/LA PRENSA
Sample: 1220
"National"
Methodology not disclosed
Key results
United opposition or FSLN for president:
FSLN 25%
opposition 49%
don't know/no answer 24%
Individual for president:
Ortega 13%
other Sandinistas 7%
Chamorro 13%
other opposition leaders 19%
July 28-30
ITZTANI-INOP
Sample: 1500
Urban areas of 28 municipalities (not Atlantic Coast)
Methodology fully disclosed
Key results
Party for president:
FSLN 37%
UNO 3%
other opposition 13%
none 19%
don't know/no answer 28%
Individual for president:
Ortega 26%
other Sandinistas 10%
Chamorro 6%
other opposition 8%
Aug 8-12
MANOLO MORALES FOUNDATION
Sample: 1150
Managua/Le6n/Masaya/Estelf Methodology not disclosed
Key results
Party for president:
FSLN 37%
UNO 10%
other opposition 9%
don't know/no answer 32%
won't vote 10%
Party for president (FSLN or UNO only):
FSLN 38%
UNO 21%
don't know/no answer 28%
won't vote 12%
Individual for president:
Ortega 24%
other Sandinistas 12%
Chamorro 10%
other opposition 11%
Sep 27-Oct 2
ITZTANI-INOP
Sample: 1157
Urban areas of 12 departments (not Atlantic Coast)
Methodology fully disclosed
Key results
Party you most identify with:
FSLN 30%
UNO 11%
other opposition 13%
none 46%
Individual for president:
Ortega 26%
Chamorro 21%
other opposition 2%
none 4%
don't know/no answer 46%
Elections free and honest:
yes 75%
no 9%
Oct 2-7
ECO
Sample: 4114
National, except Atlantic Coast, urban only
Methodology disclosed
Key results
Party for president:
FSLN 36%
UNO 18%
other opposition 2%
undecided 28.1%
abstain 13.7%
Individual for president:
Ortega 29%
Chamorro 21%
other opposition 2%
Oct 12-21
CID-GALLUP
Sample: 1126
National, except Atlantic Coast
Methodology partly disclosed
Key results
Party for president:
FSLN 29%
UNO 38%
Individual for president:
Ortega 29%
Chamorro 40%
Don’t know/no answer 31%
Individual for president, registered voters only:
Ortega 32%
Chamorro 36%
Don’t know/no answer 32%
Honst elections:
Yes 77%
Oct 23-30
BORGE AND ASSOCIATES/LA PRENSA
Sample: 1284
Unknown
Methodology not disclosed
Key results
No information published with respect to candidates
Who can resolve the problems of peace,
Bring stability and economic recuperation,
Decrease unemployment, guarantee
A better future, etc:
UNO 54-56%
FSLN 19-22%
Oct 27-31
UNIVISION
Sample: 1129 registered voters
National, except Atlantic Coast
Methodology partly disclosed
Key results
President/Vice-President / Party (secret ballot):
Ortega/Ramírez/FSLN 40%
Chamorro/Godoy/UNO 39%
Other opposition 5%
Don’t know 16%
Ortega’s job as president:
Positive 57%
Negative 39%
Don’t know 6%
Quality of life better before or after revolution:
Before 61%
After 24%
Don’t know 15%
-57% believe US is responsible for Nicaragua’s problems;
in northern part of country, 73% believe US is responsible.
Nov 14-19
ECO
Sample: 5642 registered voters
National, except Atlantic Coast
Methodology disclosed
Key results
Party for president:
Urban:
FSLN 43%
UNO 20%
Other opposition 3%
Don’t know/no answer 34%
Rural:
FSLN 50%
UNO 11%
Other opposition 1%
Don’t know/no answer 38%
Are UNO and the contra the same thing:
Yes 44%
No 29%
Can the opposition can solve the country’s problems:
Yes 23%
No 46%
Don’t know/no answer 32%
Nov 25-Dec 3
GREENBERG-LAKE
Sample: 971 registered voters
National, including Atlantic Coast
Methodology fully disclosed
Key results
President/Party (secret ballot option)
Ortega/FSLN 44%
Chamorro / UNO 27%
Other opposition 5%
Undecided 23%
Party for National Assembly:
FSLN 43%
UNO 25%
Other opposition 10%
Undecided 22%
Will results of voting be honestly
And accurately reported:
Yes 72%
No 7%
Maybe 11%
Who is responsible for continuation of war:
Government 29%
US/opposition 47%
Who is responsible for economic problems:
Government 36%
US war/embargo/business sector 47%
Dec 18-22
ECO
Sample: 4431 registered voters
National except Atlantic coast, urban only
Methodology disclosed
Key results
President / Vice-President/Party:
Ortega/Ramírez/FSLN 47.7%
Chamorro/Godoy/UNO 14.5%
Other 2.8%
Don’t know/no answer 35%
Can the opposition solve the country’s problems:
Yes 22.7%
No 46.1%
Don’t know 27.5%
No answer 3.6%
|
a monthly magazine of analysis on Central America